)IJI''h<<.!-'li4>HJi4U4aiJ4<ri4iiU*4>J/JI-> ljat<'f-l4n,4<'IU4i4— ...MH-w 
 ^ -...I>444I4 i4JU4iilt<J44 H4 lj44*44.44iiJ4. \-i,'dU»illiMiMUUWl%: 
 
 ^ ^ ^ ^ ^~'|U44444*IHl><tll/Jl*ll*444A444M4U«iW4Jl<Ui4>4-4«ll>4»bli«l 
 
 'Ui>ilU4«*4l4<4M4iil4144ib4i^i£Uiibl*<«iUl4df44>« JlwittU' 
 
 Hitli44M(MI4i4ii4 44444l«4<4lif'H444«44l4Jt44Ml44i4a-'i44l 
 
 <»'4if'l«M4if4444if*H4«^44Hf444i<44V<4*'-'v<<«i»-<*>*^«*{ 
 
 'UWlW<fUiiMM44i444ff44U44iiJi'444<l444U«4i4«4444all«-l4« 
 
 U • i|i(444l444tU4«4i4t4>«t4at4U4U 
 
 W*I lA4<IK(A44iU44U'4'(«iCU444ii44 44 
 
 4 4n lllit 
 
 i 
 
 1 
 
 
 ifnilMf 
 
 ■i Ai^' r-i'.M'" ^•l''-*'> ''.^.^It't' 
 
 
 
 ff^Fff •h *Tl*T ^ TTI^f il^f W*T( Ij 
 
 ■iiiiliiiii 
 
 ■Bail 
 
 - ■*! "I 1 411 ij IfBll 
 
 I||iti||i|l1|i MiHiq niii 
 !fM>i|liilii'll 'II'M |M ■ 
 THi'lf 'Hi!" ""■» M" 
 
 "■-ffl»ll|l||HlfIlH|l|MI 
 
 !!«« 
 
 
 1 1, 
 
 
 Fllf llllll f 4 
 
 
 H MfAH ■! ■■•■•'■••iaMlfll 
 
ENGINEERING LIBRARY 
 
 CORNELL 
 UNIVERSITY 
 LIBRARY ' 
 
^u^ ^: /^ 
 
 
 Cornell University Library 
 TN 705.S881911 
 
 The metallurgy of iron and steel. 
 
 3 1924 004 664 136 
 
 
 DATE DUE 
 
 
 JM)^— c -n&iin^ 
 
 i? 
 
 
 -t^e^T 
 
 'iQnl 
 
 it 7 
 
 
 mr 
 
 ¥^ w^^' ' 
 
 mmm 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 PRINTED IN U.S.A. 
 
The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924004664136 
 
THE METALLUEGY OF IRON A]!0) STEEL 
 
-^^ ^^ 
 
 McGraw-Hill DookCompaijy 
 
 Puj6C£s/iers (^3oo£§/br 
 
 ElGCtiical World TheEngiiiGeiTi^ andMiniiig Journal 
 EngkieGriiig Record Engineering News 
 
 Railway Age Gazette American Machinist 
 
 Signal Lnginper .AraericanEn^tieer 
 
 Electric Railway Journal Coal Age 
 
 Metallurgical and Chemical Ln^neering Power 
 
 ^■111^^^^^^^^^^^^ iiiiyiiiiiiiiiiiiiii^^^^ 
 
 J 
 
, F 
 
 THE METALLURGY OF 
 
 IRON AND STEEL 
 
 BY 
 
 BRADLEY STOUGHTON, Ph.B.,B.S. 
 
 Second Edition 
 Thorottghlt Revised and Entirely Reset 
 
 Fifth Impression, 
 Total Issue, 13,000 
 
 McGRAW-HILL BOOK COMPANY, Inc. 
 
 239 WEST 39TH STREET. NEW YORK 
 
 LONDON: HILL PUBLISHING CO., Ltd^ 
 
 6 & 8 BOUyERIE ST.. E.G. ^ 
 
 1913 
 
Copyright, 1908, by 
 The Hill Publishing Company 
 
 Copyright, 1911, by 
 McGraw-Hill Book Company 
 
 Entered at Stationer's Hall, LondoQ, England, 
 
 Printed and Eleciroiyped by 
 
 The MaPle Press 
 
 York, Pa, 
 
TO 
 PROFKSSOR HENRY MARION HOWE, A.M., LL.D., 
 
 BESSEMER MEDALLIST, KNIGHT OF THE ORDER OF 
 
 ST. STANISLAS, ETC., ETC., 
 
 PRACTITIONER, INVESTIGATOR, AUTHOR, INTERPRETER, 
 
 EDUCATOR AND PHILOSOPHER, WHOM THE WORLD 
 
 OF SCIENCE DELIGHTS TO HONOR, 
 
 ^ THIS VOLUME IS 
 
 AFFECTIONATELY DEDICATED 
 
PREFACE TO SECOND EDITION 
 
 Developments in the Metallurgy of Iron and Steel have been 
 so rapid during the past three years that a more thorough 
 revision of the text of this volume than could be accomplished 
 in conjunction with the usual reprinting for new impressions 
 seemed not only desirable but necessary. Every department of 
 manufacture has been improved in practice, in minor respects 
 at least, and some branches, such as the duplex and special 
 open-hearth processes, electric smelting and refining, use of dry 
 blast, case-hardening, etc., have been advanced sufficiently to 
 warrant an entire rewriting of the sections devoted to them. 
 Opportunity has been taken of these revisions to make more 
 or less extensive changes in all the text, with the result that it 
 has been necessary to reset the type from cover to cover. I am 
 especially indebted to J. E. Johnson, Esq., for valuable sugges- 
 tions and revisions of the chapters on blast furnace, and chemistry 
 and physics. 
 
 In addition to the changes mentioned above, the theory and 
 chemistry of the Bessemer and open-hearth processes have been 
 rewritten and enlarged, as well as the whole chapter on malle- 
 able cast iron. The section on blast furnace fuels has been 
 taken from Chapter II, and this also has been rewritten and 
 enlarged; with the section on open-hearth fuels from Chapter 
 VI, it has been embodied in a new Chapter XIX, on Metallur- 
 gical Fuels. The references to literature have, in accordance 
 with several suggestions, been taken from the end of the indi- 
 vidual chapters and collected in an Appendix at the end of the 
 volume, thus rendering it easier to turn to the index numbers 
 at will. 
 
 In making these changes a constant and serious handicap 
 has^ been the necessity of keeping the number of pages within a 
 small compass in order not to disturb the price at which the 
 book is sold; especially for the sake of student purses. 
 
 It will be noticed that the numbering of the illustrations is 
 not quite consecutive, but a few numbers are left out at the 
 
 vii 
 
PREFACE TO FIRST EDITION 
 
 The purpose of this book is to serve as a text-book, not only 
 for college work, but for civil, mechanical, electrical, metallurgi- 
 cal, mining engineers and architects, and for those engaged in 
 work allied to engineering or metallurgy, America now pro- 
 duces almost as much iron and steel as the rest of the world 
 together, although less than eighteen years ago she held second 
 rank in this industry. It seems fitting that the record of this 
 progress should be brought together into one volume covering 
 every branch of the art of extracting the metal from its ores and 
 of altering its adaptable and ever-varying nature to serve the 
 many requirements of civilized life, 
 
 I take pleasure in acknowledging here, with sincere thanks 
 the assistance of many who have aided in the make-up of the 
 volume, and especially The Adams Co., American Electric Fur- 
 nace Co., American Sheet & Tinplate Co., Bethlehem Steel Co,, 
 Brown Specialty Machinery Co., Connersville Blower Co., 
 Crocker-Wheeler Co., Francis G. Hall, Esq., Holland Linseed 
 Oil Co., Chas. W, Hunt, Esq., Secretary, American Society of 
 Civil Engineers, Professor James F. Kemp, Mackintosh, Hemp- 
 hill & Co., Morgan Construction Co., National Tube Co., 
 S. Obermayer Co., J. W, Paxson Co., Henry E. Pridmore, John 
 A. Rathbone, each of whom have kindly loaned electrotypes. 
 And of Dr, H. C. Boynton, the Brown Hoisting Machinery Co., 
 Buffalo Furnace Works, H. H. Campbell, Esq., Professor William 
 Campbell, Carnegie Steel Co., W. M. Carr, Esq., Central Iron & 
 Steel Co., Crucible Steel Company of America, Fiske & Robinson, 
 The Foundry, Harbiaon- Walker Refractories Co., Joseph Harts- 
 horne, Esq., Professor Henry M. Howe, -Lackawanna Steel Co., 
 Marion Steam Shovel Co., Mesta Machine Co., Morgan Engineer- 
 ing Co. , Professor A. H. Sexton, Wm. Swindell & Bros., United 
 Coke & Gas Co., United Engineering & Foundry Co., Wellman- 
 Seaver-Morgan Co., Whiting Foundry Equipment Co. And of 
 O. S. Doolittle, Esq., for information upon paint, Frank E. Hall, 
 Esq., for the analyses in Table XVIII, and W. J. Keep, Esq., 
 for the figures in Table XXVI. 
 
 ix 
 
m ^'*%^^^ ^^^^ ^-" ^-^^ f^' * »♦ *** > "" ~' 
 
 ||#.>^#« *"^ ^H.«»-«** 
 
TABLE OF CONTENTS 
 
 Pages 
 
 Chapter I. Introduction — Iron and Carbon 3 
 
 Definitions, 6. 
 
 Chapter II. The Manufacture of Pig Iron 8 
 
 Varieties and distribution of iron ores, 9, United States de- 
 posits and transportation, 11. Handling raw material at a 
 modern furnace, 16. The blast furnace and accessories, 17. 
 Smelting practice and products, 24. Calculating a blast-furnace 
 charge, 40. 
 
 Chapter III. The Purification of Pig Iron ... . . 45 
 
 Miscellaneous purification processes, 52. 
 
 Chapter IV. The Manufacture of Wrought Iron and Crucible 
 
 Steel . . . . .57 
 
 The manufacture of wrough iron, 57, The carburization of 
 wrough iron, 68. 
 
 Chapter V. The Bessemer Process . .- 79 
 
 Chapter VI. The Open-Hearth or Siemens-Martin Process. Ill 
 Open-hearth plant, 111. Open-hearth furnace, 1 14. Basic 
 open-hearth practice, 128. Acid open-hearth practice, 138. 
 Special open-hearth processes, 143. 
 
 Chapter VII. Defects in Ingots and Other Castings .... 158 
 
 Chapter VIII. The Mechanical Treatment op Steel .... 172 
 The forging of metals, 174. The reduction of metals in rolls, 
 180. Parts of rolling nlills, 186. Rolling-mill practice, 201. 
 Wire drawing, 210. Pressing, 212. Comparison of mechanical 
 methods, 214. Heating furnaces, 215. 
 
 Chapter IX. Iron and Steel Founding 221 
 
 The making of molds, 222. Design of patterns, 247. Cupola 
 melting of iron for castings, 249. Comparative cupola practice 
 265. Melting steel for castings, 270. 
 
 Chapter X. The Solution Theory of Iron and Steel .... 275 
 The freezing of alloys of lead and tin, 278. The freezing of iron 
 and steel, 286. The complete Roberts-Austen, Roozeboom dia- 
 gram, 293. 
 
 Chapter XI. The Constitution of Steel 299 
 
 The micro-constituents of steel, 299. The strength of steel, 307. 
 Hardness and brittleness of steel, 311. Electric conductivity of 
 steel, 312. Magnetic properties of steel, 312. 
 
 xi 
 

 
 ^%« .««*«™. it .^- ." .*» 
 
 2 t 
 
 '^%¥> *#***' ** 
 
 
 .*N# 
 
 ¥.m ^"*-« 
 
 .^,#*^ 11^ *« ^'•^^ ^♦-' '***^ ^ •^'•^- 
 
 mm 
 
THE METALLUEaY OF lEON AND STEEL 
 
FIG. I.-IRON BLAST FURNACE AND STOVES. 
 
INTRODUCTION— IRON AND CARBON 
 
 This chapter is written for those students of iron and steel — 
 whether they he students engaged at some university, or in an 
 engineering or metallurgical profession — who have previously 
 completed a course in chemistry and physics. For the benefit of 
 those who have not had a technical education, Chapter XX has 
 been especially prepared, and it is hoped that, if they will read 
 that chapter before beginning elsewhere in the book, all the subjects 
 discussed in these pages will he readily intelligible to them. 
 
 The Ferrous Metals. — Iron and steel together form the largest 
 manufactured product in the world, and each of them enters 
 into every branch of industry and is a necessary factor in every 
 phase of our modern civilization. Cast iron, because of the ease 
 with which it can be melted, is produced in final form in almost 
 every city in the United States, and only slightly less widely in 
 other civilized countries. The manufacture of steel is more 
 centralized, for economical reasons, but is several times as great 
 as that of cast iron. Wrought iron is less in amount than 
 either of the others, but has its own importance and uses. These 
 three products — cast iron, steel, and wrought iron — together 
 comprise the whole of the so-called "ferrous group of metals" — 
 that is, the group which we classify together under the name of 
 "iron and steel." They have two characteristics in common: 
 First, that iron is present in all to the extent of at least 92 per 
 cent.; and second, that carbon is their next most important 
 ingredient, and regulates and controls their chief qualities. 
 Their manufacture represents nearly 15 per cent, of all the 
 world's manufacturing wealth, and is far greater than any other 
 like industry, (See Table I.) 
 
 Cast Iron. — Cast iron is impure, weak, and must be brought 
 to its desired size and form by melting and casting in a mold. A 
 typical example would contain about 94 per cent, iron, 4 per 
 cent, carbon, and 2 per cent, of other ingriedients or impurities. 
 
 Pig iron is a raw form of cast iron, and malleable cast iron is 
 a semipurified form. 
 
 3 
 
4 THE METALLURGY OF IRON AND STEEL 
 
 Steel. — Steel is purer than cast iron, much stronger, and may- 
 be produced in the desired size and form either by melting and 
 casting in a mold or by forging at a red heat. It usually contains 
 about 98 per cent, or more of iron, and, in different samples, 
 from 1.50 per cent, down to almost no carbon, together with 
 small amounts of other ingredients or impurities. 
 
 Wrought Iron. — Wrought iron is almost the same as the very 
 low-carbon steels, except that it is never produced by melting 
 and casting in a mold, but is always forged to the desired size 
 and form. It usually contains less than 0.12 per cent, of carbon. 
 Its chief distinction from the low-carbon steels is that it is made 
 by a process which finishes it in a pasty, instead of in a liquid 
 form., and leaves about 1 or 2 per cent, of slag mechanically 
 disseminated through it. 
 
 Iron. — Iron as such — by which I mean pure iron — does not 
 exist as an article of commerce, but appears in service and in the 
 ^market only in the form of cast iron, steel or wrought iron — that 
 is, when contaminated with carbon and other impurities. Some 
 of these impurities are present because they cannot cheaply be 
 gotten rid of, and others, because, like carbon for example, they 
 benefit the metal by giving it strength or some other desirable 
 property. Pure iron is a white metal and one of the chemical 
 elements. It is with one exception the commonest and most 
 abundant metal in the earth, and almost all rocks contain it in 
 greater or less degree, from which we extract it if it is large 
 enough in amount to pay for working. It is almost never 
 found in nature in the form of a metal, but is always united 
 with oxygen to form either a blackish, brownish, reddish or 
 yellowish substance. Indeed, if it should occur in metallic form 
 it would . very soon become oxidized by the action of air and 
 moisture. 
 
 It is the abundance of iron in the earth which is the chief 
 cause of its cheapness, and therefore one 'reason why it is used 
 more than any other manufactured material. The other reason 
 is the ease with which we can confer upon it at will some of the 
 qualities most useful to man, of which the most valuable is prob- 
 ably its strength, and the most wonderful its magnetism, in 
 which it is not even approached by any other substance. What 
 these two properties alone mean in modern structural and elec- 
 trical engineering can scarcely be estimated. 
 
INTRODUCTION— IRON AND CARBON 
 
 
 
 go M M»OS"300r-t«,-4iOiflOr4(MM«ONOOfflOt-CM»-i'^*lflS^rt*ii . 
 
 
 
 eO rH <N CS 
 
 0J00OO00iH00c0iNt*aJOt*t-(D01W5l>C0OQ0c0THt-.c6Sd) . 
 
 d 
 
 3 
 
 2 9 £3 ^ 
 
 , co_ (N CO q m i> q oo. o_ q "3 co_ q (n o oo co_ u5 in »o K oq q ^_ o n S . 
 iH c<) ^* lo oi oo cooroo"b^iocoooc^~cooiC<rcoorTjJ'T(?'Qot-ro'o' 
 
 o 
 
 gi lo c 
 
 ? 
 
 5 
 
 o c^ 
 
 
 
 
 O? <f 
 
 ■«*i « 
 
 2SS5^^^SS^^'§;3?3S"S^"?^'^"5g;S"gS^'SSS : 
 
 o 
 
 
 I— 1 
 
 iH 1-1 
 
 H 
 
 if 
 
 ^ ? 
 
 CO CC 
 
 cooococoosocSooooooooooooooooooo ■ 
 
 COOOeOi-iiOOOOOOOOQOOOOOOOOOOOOO • 
 
 ,K5qqrH-*qqqqqqqqqqqqoooc30c3soooo ■ 
 
 CTiiOU3COOOO"3«3»Q"5b-0>»0-^W(NCOrtC^COU3iOioSiOC^CO ' 
 
 coMcpwcocoepeocoeococo-*ubcocD(OcococO(Oco»ou5»oloio ' 
 
 CO 
 
 CO c< 
 
 l> -^ 
 
 q «: 
 CO cr 
 
 q a 
 
 ,0 
 
 ^ S 
 
 r- 
 
 c 
 
 
 K 
 
 cc 
 
 p: 
 
 ^ 
 
 , 
 
 
 
 ** *******#*******«#*# * tt"^ ; 
 
 M 
 
 o3 
 
 
 
 
 
 'OOcni-HN •b-'^ot>-cD»Ob-oOeje>icDCD^b-i-HQO^cococoQO'* 
 
 13 
 
 
 
 
 
 ■COCSIQOCI ■01wiOcq00C0O'*00rH0>lNC0t*0>rH-*C^(Ni0iO« 
 
 Q 
 
 1 
 
 
 
 
 
 ..HiOt^ifl .COiOI>COiOcO-*Wii3eDOOOCOb-^i1CO-^rHOiOiO"S 
 
 y^ 
 
 
 
 
 
 ; s I5' 2 s" ; s i s «" ^" ^" § s § es ^' s s" s e s § s s ^s" s ?3 
 
 ^\ 
 
 
 
 
 
 P - 
 
 
 
 
 
 .* . * «CO(N«-*»r3iO»OCDt^00C* 
 
 2 
 
 
 
 
 
 
 rHi-iiOOOCOcOOOlOOCJiOQOQOOOOOOOOOOTjiOOO • 
 CJJ01C0C0t--^C0<N«MO05O"5TH0JOoKcoS5NQ5o ■ 
 qC^NiO'*COOSl>OCOiO(MU300WCnOU3NCDeQ'«Ji001QOOO • 
 
 >, 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 ^««*»*-*t*tBkn<h#i»«.a.»»^»#.»tf.«>nK^ ^» ** 
 
 o 
 
 
 
 
 
 »OMlM«CO«*iH(NOOOOtHNWO)eoiOTtt«t>-^ON(NI>»OiO ' 
 
 IH 
 
 
 
 
 
 iHiHi-HiHiHi-ti-lr^ iHt-t.HrHT-li-(NWWWWCOCOCO.-lOrtco 
 
 ^ 
 
 
 ^ 
 
 
 
 "^ 
 
 * r^ M w w ; 
 
 8 
 
 N 
 
 
 
 
 Jp 
 
 !i; ^ 
 
 iOCO-^tOTt((NNO>OOOcOr^e5S»«COOOb-'<J<OOOC30000 . 
 
 
 
 
 1> 
 
 0> CO 
 
 Mcqcoi-ib-oocowu50cocqt»enooootH"^oooooooooo - 
 
 g 
 
 
 
 CO 
 
 ^ OS 
 
 Nr^qi-iooi>'^coTHO'^coi-it-.cob-rHi>'WO<-i'Oo»ooo»o • 
 
 ^— ' 
 
 *3 
 
 
 
 i^ 
 
 ost-rodcir^oodbro'ocDco"Mr^cD"enTfroocDcob."u3cooJi>rM ■ 
 
 i«U500U50SI>-I>-'*COCOO-*OOCOCOOOa>COOOOOOOCOOOOOOCOiO 
 
 
 D. 
 
 
 
 -d 
 
 o 
 
 m 
 
 
 
 h4 
 
 -' 
 
 1-i W(Nr-t.-iW(N««cq(Mw<MiNW«cocococococo-*coeoeo : 
 
 s 
 
 Ol oc 
 
 C3 
 
 CO 
 
 b-l>U31NiOC0C0'*iHCS)OOOOb.00OOU5i0OOOOOOOO 
 C0iflM»0c0OtHCD(Ne0OOQOC<lCDt>-OC<)C^Oir3OOOOOO 
 
 *X"*-^<^'=0'-10* ^'^'^'?i'*°0*>'>'^'^-^oo»flc^_(N^coqcooo a 
 
 H 
 
 r-K M 
 
 CO 
 
 TH 
 
 <D « 
 
 "; 
 
 t>; 
 
 
 V 
 
 o o 
 
 
 ui 
 
 Tf oJ to t>r <3 CO o >ft CO oT uf co" CO w" t-^ co" co~ ■*d^" CD 00 rH c<r CO CO co' ^^ 
 
 1 
 
 (N O 
 
 
 o 
 
 g;^«iO(Nif3aiOO»OU3U3COCOOiCOC^<N(NO(NCOiOOCOTjiOcoO 
 
 M M 
 
 c. 
 
 •« 
 
 '«-*'*-«'«-*-«Tjt'*-«iJ-*-^U3-*-*iOiO»0»OiC"5iOCDW3'*COcDti 
 
 H 
 
 
 CO cc 
 
 CO 
 
 ■^ 
 
 oooocowiHoo'^obocsiooooen-vOOOcoooooooo 
 
 
 O ifl 
 
 U3 
 
 CO 
 
 ^ a- 
 
 o 
 
 c; 
 
 I>qiqM_OOcqOOC^-l^iOOOOOOSOSOOqu3COU300l>00 . 
 
 ^ 
 
 
 
 
 
 ^ bo 
 
 :i £ 
 
 c< 
 
 ■^ 
 
 ^CR'<^<S^^r^cSc^-^in<Oini6nr^<S*n*nc^c^n*6zS^'6^ ; 
 
 
 1 § 
 
 o c 
 
 
 CO 
 
 iHiHOOS»OCONM<OOiOt*COOiONrHOCOiOeOt>.oOW3iOi-HOJ 
 
 5 
 
 i-i ^ 
 
 '^ 
 
 '«l>i>t*l>COC>OSOiOSOO^OiIlMCOCOeOCOCOeOCOTt<r}<cOCJOO '1 
 
 <M 
 
 
 
 
 iHrHWrHr-riHiHiHiHrH^rHrHi-Zr-riHCqeJ" ; 
 
 ;z; 
 
 
 00 rH 
 
 l> 
 
 CO 
 
 O-*'^l>-b-i-iU3'^t<-O00OOO0SC0<M"3Oi0OOOOCqOO ■ 
 t*TtlOOCOiflOO'*^COCOC)i-HO"3Tj<COt^CO.-tMOOC^OCOOO ■ 
 
 .3 
 
 CO P3 
 
 00 
 
 03 
 
 § 
 
 ■^ u; 
 
 r- 
 
 "• 
 
 ^H^-l-<^-o)u^l>co^-I>Moooool>•cooJ■*cocooowoNOO ■ 
 
 
 
 
 
 
 i 
 
 oT 03 
 
 t> 
 
 oc 
 
 »o«:©<ot,oiooii-i«-*cooowt^QOoows'*osrHeoi>i>QOt^oo . 
 
 OJ >0 
 
 N 
 
 ^ 
 
 M 
 
 P^ 
 
 N cc 
 
 •^ 
 
 * 
 
 
 
 iH * iH rH iH i4 iH Oq N W N W N « C^ « N « W W" N ; 
 
 o 
 
 hH 
 
 
 
 
 
 IS- -^ 
 
 10 
 
 Tli 
 
 COt-.HOCOCOCOOO'^O'OOCOOOiOOOO^fflOOOOOOOO 
 
 s 
 
 d 
 
 <0 CO 
 
 o 
 
 00 
 
 t-t--OOiniNPOC»CDCO'*COQOCD'<t»COOOC^i-H.-iOS01CD'*'*U30000S 
 
 «3. «i ^^ 00 ci «i ^- "_ "- oi n •> «?. t^. --1 '-I "4 '^^ «i. =^ ^- '-I ^- ■*- ^ "I ". '^- 
 
 9 
 
 t- w 
 
 oc 
 
 c 
 
 
 
 
 
 
 
 Pq 
 
 S 
 
 O "5 
 
 T- 
 
 oc 
 
 rtOiOCqCOOOOOiO-^^WCOC^QOCoScOOOiO^COWOCOO-*© 
 
 t- CD 
 
 tJI 
 
 c 
 
 c 
 
 m 
 
 ■* »fl 
 
 »« 
 
 coi>i>;i>oo oo^'CDt^^•ooooolqoq. rHt*rH(NcocoTPTjH_M_©oc)^o co| 
 
 "^ 
 
 
 
 
 tH tHW iHiHiHr-liHi-Hr-tiHr-lC^Cq 
 
 
 O CO 
 
 <M 
 
 CO 
 
 oo■^c^osocD^>oo^^c3iC<^co>ooooco^-ool>oc<^oso»oc3JOJW 
 
 o 
 
 o 
 
 r* ^ 
 
 t- 
 
 Ol 
 
 -*b-N^OCnOOOO'*OOOTtils.OeP(NNCOOOOCO'*>OOu3(OOS 
 
 hH 
 
 o 
 
 ^-'-' 
 
 C^ 
 
 <M 
 
 qioqcoqrHcocoi-^qoOb^.-|_q^woo-^qt-;_qqq^r-i-^'#q 
 
 O 
 
 § 
 
 CO M 
 
 OC 
 
 liS 
 
 cortcooococoojioot>oeot^(N-i>^oociNO>t-rH55encococ^b- 
 
 O N 
 
 ■* 
 
 C^ 
 
 fi 
 
 W tH 
 
 ■^ 
 
 t>; 
 
 qioicqi>0)coooqoc0'^ »o lOt-^co-^coqqcoi-iCOcDO ■* oo 
 
 
 
 
 
 
 th" r- 
 
 r-i 
 
 iH 
 
 ■A-r^rA'^'n^r^r^c^G^cic^^c^c^G^c^Gic^eiti^'nfnco'nn-^^'^ 
 
 1 
 
 .H M 
 
 
 00 
 
 ■«*<00C0rHQ0^00THONXAit^t>C0l0t-O-*i-(C0l>O^C0O>b-^ 
 
 Oi o; 
 
 
 CO 
 
 C0U3»O<NX0i0^C0C0MC»0000'-(O»O00cDC0-<*<CJCDCD.Hi«-**C^t» 
 '*COCTi'HU5'*C>jT}tOCOb-OSOt*C^COOOCDCOOSCDOt-»ncDCO WW 
 
 T-t 1-1 
 
 
 q 
 
 
 
 
 
 
 FM 
 
 i 
 
 00 T-1 
 
 d 
 
 (N 
 
 qiOOeo>OCOcOOiOl»Ot>cOOO'*iH"5l>rtfOi-<Oi'<*<OQOOJt- c^x 
 
 00 OS 
 
 (N 
 
 03 
 
 Xtl 
 
 Oi CO 
 
 C 
 
 ^ 
 
 
 
 
 
 % 
 
 o 
 
 r-l 
 
 C) 
 
 N 
 
 coeoTii'Tii'Tirrii'-^Tir-'ji ^^^ocD(0^^^^^^^^odoodcicowM■*"^o^^ 
 
 
 ^ o 
 
 O) 
 
 W 
 
 C^WOOb-COMi-trtfcOHtStOOcDOOOOCOOSCOCOiHiHOOODb-CONO • 
 
 '^THco'#ooioot^i-H'*oeqQOcot*'<**"5i>aocDcqb-cob-^.Hpq • 
 
 -w .3 
 
 r* (N 
 
 iH 
 
 ^ 
 
 o 
 
 en i> 
 
 CS 
 
 ut 
 
 oOqi>0'*OMCMcot-OMOOJi-iO'*ojiHqo»i-<_q'^qe^cD • 
 
 ^ 
 
 1> c 
 
 
 1« 
 
 oO'#eqoJooMiAi>orTjrcJddorNcot-"coc^"orcdo"wodoo"d tjT : 
 
 h ••-• 
 
 3S CD 
 
 
 t- 
 
 C0(M00C<)ii5C0C^'-H00COWOC0i-ib-Ot^'OiCC»'*i-<00C0r-(00ts, , 
 
 ^fS 
 
 00 O 
 
 
 0?. 
 
 C0i-ICO^'*O"3Q0OC0Ob-Oi00»OOC3JCOCnc0t-(Nq'*Qqc0 00 , 
 
 
 
 
 
 
 ^ 
 
 Tji © 
 
 «D 
 
 t* 
 
 t*i>cooooooobrcDi>b^ododQdocJ'oros"t^ooododoroooioidos ■ 
 
 1-^ iH fH 
 
 H 
 
 
 O 00 
 
 o 
 
 Q 
 
 ooocotnoocooocDcoo>t*cot-'*}<oooooocDoocDco'*gjt; oog 
 
 1 
 
 1 1 
 
 "3 -^ 
 
 ^ l> 
 
 us 
 
 '^ 
 
 oco b-cocncqt--cqo-!t<**c3Je<i.-)cob-OTt)oco»oocD<»oioo sS 
 
 CDi-<CJl>»flOiHTjf00C<l'*^rtC0CDC0'*rfl'*OSMf-l_00qi-iqq'O 
 
 1. 
 
 ■^ cr. 
 
 t* 
 
 CO 
 
 f— 1 
 
 CO IN 
 
 <o 
 
 I> 
 
 d d i-I" lo ^ CO CO ■*" oT i> IV w i>" «" oo" oi M CO b^ <3 d «" CO d od (O t; n 
 
 
 Ttt at 
 
 lO 
 
 en 
 
 r-tb-MOJt-iOr-lOCOiOOJCOOCOCOOCOOOJCD'^WOJaiOCO sjo 
 
 
 t5 02 
 
 00 to 
 
 o 
 
 oo_ 
 
 
 i-i 
 
 ■ci 
 
 CO 
 
 ■^U3cDcOt>dcOCnb^COdoodrHCO'*"cDCidcOCOCOU5COCOCOt» tMO 
 
 
 
 
 
 ^r^iHi-iWi-iT-(DlMNiH«(M Oieo 
 
 
 ; 
 
 
 
 
 
 ^ 
 
 
 " 
 
 • 
 
 
 d 
 
 irj 
 
 Q 
 
 iodNiooasd^"(NcO'*ifseDt^oC)oJd^«cO'*»ocot>QOOsd weq 
 
 
 CO t^ 
 
 b- 
 
 co 
 
 000000C000030JOS0>O>0»O»0>050>OOOOOOOOOO'~l ^ y-* 
 
 QOooooa3«SoooooooooooooOooooooc»cnoscnffia>c:)0)0>oso> cno> 
 
 
 
 00 oc 
 
 00 
 
 CO 
 
 
 1-1 "■ 
 
 
 *" 
 
 tH 
 
 i-lTH?-lT-1fHT-1iHrHi-lr-(rHi-lfHT-l?-*i-li-lr-(i-liHiHrHr^iHi-l»H iHiH 
 
6 THE METALLURGY OF IRON AND STEEL 
 
 Carbon. — Carbon is also a chemical element and familiar to 
 everyone; graphite, lamp-black, charcoal, and diamond are the 
 various allotropic forms in which it appears. It is a common 
 substance and present in every form of organic matter, while its 
 oxides — carbon monoxide, CO; and carbon dioxide, COg — are 
 well known gases. Its chemical affinity for iron is very great; 
 iron practically always contains some amount, but if it is de- 
 sired to remove it entirely, the last traces are eliminated only 
 with extreme difficulty^. 
 
 Iron and Carbon. — Carbon has the peculiarity of conferring 
 on iron great strength, which, strange to say, it does not itself 
 possess, and also hardness, which it possesses only in its diamond 
 allotropic form. At the same time it takes away from the iron 
 a part of its ductility, malleability, magnetism and electric con- 
 ductivity. So important is the influence of carbon in regulating 
 and controlling the characteristics of the ferrous metals, that 
 they are individually and collectively classified according to the 
 amount and condition of the carbon in them. The potent effect 
 of carbon must be constantly borne in mind when we come to 
 describe the manufacture of iron and steel and to discuss the 
 methods of regulating the carbon. 
 
 Definitions 
 
 The following definitions are selected from the report of March 
 31, 1906, of the Committee on the Uniform Nomenclature of Iron 
 and Steel of the International Association for Testing Materials, 
 with slight changes: 
 
 Cast Iron. — Generically, iron containing so much carbon or 
 its equivalent that it is not malleable at any temperature. Spe- 
 cifically, cast iron in the form of castings other than pigs, or 
 remelted cast iron suitable for casting into such castings, as dis- 
 tinguished from pig iron, i.e., cast iron in pigs. 
 
 The committee recommends drawing the line between cast 
 iron and steel at 2.20 per cent, carbon for the reason that this 
 appears from the results of Carpenter and Keeling to be the 
 critical percentage of carbon corresponding to the point "a" in 
 the diagrams of Roberts- Austen and Roozeboom. (See page 295). 
 
 Pig Iron. — Cast iron which has been cast into pigs direct 
 from the blast furnace. This name is also applied to molten 
 cast iron which is about to be so cast into pigs, or is in a condition 
 
INTRODUCTION— IRON AND CARBON 7 
 
 in which it could readily be cast into pigs before it has ever been 
 cast into any other form. 
 
 Gray Pig Iron and Gray Cast Iron. — ^Pig iron and cast iron in 
 the fracture of which the iron itself is nearly or quite concealed 
 by graphite, so that the fracture has the gray color of graphite. 
 
 White Pig Iron and White Cast Iron. — Pig iron and cast iron 
 in the fracture of which little or no graphite is visible, so that 
 their fracture is silvery and white. 
 
 Mottled Pig Iron and Mottled Cast Iron. — ^Pig iron and cast 
 iron, the structure of which is mottled, with white parts in which 
 no graphite is seen, and gray parts in which graphite is seen. 
 
 Malleable Cast Iron. — Iron which when first made is cast in 
 the condition of cast iron, and is made malleable by subsequent 
 treatment without fusion. 
 
 Malleable Iron. — The same as wrought iron. ' A name used 
 in Great Britain, but not in the United States, except carelessly 
 as meaning "Malleable cast iron." 
 
 Steel. — Iron which is malleable at least in some one range of 
 temperature, and in addition is either (a) cast into an initially 
 ma-lleable mass; or (b) is capable of hardening greatly by sudden 
 cooling; or (c) is both so cast and so capable of hardening. 
 
 Wrought Iron. — Slag-bearing, malleable iron, which does not 
 harden materially when suddenly cooled. 
 
 In the definition of steel the first sentence ("is malleable at 
 least in some one range of temperature") distinguishes steel 
 from cast iron and pig iron; the second sentence (" is cast into an 
 initially malleable mass") distinguishes it from malleable cast 
 iron, and the third sentence (" is capable of hardening greatly by 
 sudden cooling") distinguishes it from wrought iron. At the 
 best, however, the definition of steel is in a shockingly bad con- 
 dition, and has been brought to it by a series of events which 
 shows the carelessness of the buying public and the greed of men 
 who will appropriate the name for their product that will bring 
 them the best price without regard to whether the name really 
 fits or not.^ 
 
 1 See page 173 of reference No. 1, and page 6 of No. 2. Especially see H. M. Howe, 
 Eng. and Min. Jour., Feb. 11, 1911. Vol. 91, pp. 327-330. 
 
n 
 
 THE MANUFACTURE OF PIG IRON 
 
 Whatever material we are to manufacture — cast iron, 
 wrought iron, or steel — or for whatever purpose the metal is to 
 be used, the first step in the operation is smelting iron ore in a 
 blast furnace with fuel and flux, and obtaining cast iron or pig 
 iron, terms used synonymously in the United States. The pig 
 iron thus produced is an impure grade of iron, containing chiefly 
 carbon, silicon, manganese, sulphur and phosphorus. The 
 amount of pig iron made exceeds that of any other product 
 manufactured by man. Blast furnace fuels, are discussed in 
 Chapter XIX. 
 
 ANALYSIS^ OF VARIOUS GRADES OF PIG IRON. 
 
 r 
 
 Name 
 
 Silicon 
 
 Sulphur 
 
 Phosphorus 
 
 Manganese 
 
 Carbon 
 
 
 Per cent. 
 
 Per cent. 
 
 Per cent. 
 
 Per cent. 
 
 Per cent. 
 
 r No. 1... 
 
 2.75 
 
 0.035 
 
 0.30 to 1.50 
 
 0.20 to 1.60 
 
 3.00 to4 00 
 
 Foundry No. 2. . ; 
 
 2.25 
 
 0.045 
 
 0.30 to 1.50 
 
 0.20 to 1.60 
 
 3.00to4.00 
 
 Irons.i No. 3... 
 
 1.75 
 
 0.055 
 
 0.30 to 1.50 
 
 0.20 to 1.60 
 
 3.00to4.00 
 
 No.' 4... 
 
 1.25 
 
 0.065 
 
 0.30 to 1.50 
 
 0.20 to 1.60 
 
 3.00to4.00 
 
 Forge Iron^ 
 
 0.75 to 1.75 
 
 0.05 to 0.30 
 
 0.30to3.00 
 
 0.20 to 1.50 
 
 3.00to4.00 
 
 Bessemer — Acid 
 
 0.80 to 2.00 
 
 0.03 to 0.08 
 
 under 0.10 
 
 0,30 to 0.50 
 
 3.50to4.00 
 
 Bessemer — Basic. . . , 
 
 under 1.00 
 
 under 0.10 
 
 1.75 to 3.50 
 
 1.00to2.00 
 
 3.50to4.00 
 
 open-hearth — ^Acid. . . 
 
 0.75 to2.50 
 
 under 0.050 
 
 under 0.05 
 
 0.30 to 0.50 
 
 3.50to4.00 
 
 Open-hearth — Basic. 
 
 under 1.00 
 
 under . 100 
 
 0.10 to 2.00 
 
 1.00to2.00 
 
 3.50to4.00 
 
 Ferromanganese^. . . . 
 
 0.50tol.00 
 
 under 0.03 
 
 0.10 to 1.00 
 
 80.00 
 
 5.00to7,00 
 
 Ferromanganese^ 
 
 under 1.00 
 
 under 0.03 
 
 0.10 too. 50 
 
 40.00 
 
 5.00to6.00 
 
 Spiegeleisen^ 
 
 under 1.00 
 
 under 0.05 
 
 under 0.15 
 
 15.00 to 30.00 
 
 5.00to6.00 
 
 Ferrrosilicon* 
 
 8.00 to 15,00 
 
 50.00 
 8.00 to 15.00 
 
 under 0.07 
 under 0.02 
 under 0.01 
 
 0.10 too. 60 
 under 0.08 
 under 0.15 
 
 
 1 00 to2.00 
 
 Ferrosilicon^ 
 
 
 under . 40 
 
 SiUco-Spiegel^ 
 
 15.00 to 20.00 
 
 l.OOtol.50 
 
 Ferro-Chrome^ 
 
 1.00 to 5. 00 
 Low 
 
 about 0.04 
 about 0.04 
 
 0.02 to 0.05 
 0.02 
 
 
 5 50 to 10.0 
 
 Ferro-Chrome^ 
 
 
 under 1.0 
 
 
 
 
 ^ See also Chapter IX. 
 
 ^ Used for Puddling. See Chapter IV. 
 
 * Valued for its mEingEmete. 
 
 * Valued for its silicon 
 
 ^ This grade is made by electric smelting. 
 
 * Valued for its maganese and silicon. 
 
 ' Valued for its chromium which is about 60 to 70 per cent. 
 
 * Special low carbon. 
 
 8 
 
THE MANUFACTURE OF PIG IRON 9 
 
 The gangue of our iron ores consists usually of silica, alumina, 
 etc., and, like the fuel ash, requires the addition of the correct 
 amount of limestone flux to make it into a fusible slag. In the 
 Pittsburg district we charge about 1200 lb. of limestone, 2200 lb, 
 of coke, and 4000 lb. of ore for every long ton of pig iron made. 
 The amount of each is increasing, however, from time to time, as 
 the higher-grade ores are becoming exhausted and there are more 
 impurities to be fluxed and melted. Recently, enriching ores 
 by wet concentration has been practised more extensively, and 
 " washers" at or near the mines decrease the proportion of useless 
 impurities to be transported and to be treated in the furnace, 
 while magnetic concentrators are important in the eastern 
 United States. 
 
 Varieties and Distribution of Iron Ores 
 
 The iron ores used for smelting consist of chemical compounds 
 of iron and oxygen containing more or less water, either in the 
 form of moisture or chemically combined as water of crystalliza- 
 tion. 
 
 Hematite (FegOg). — The best known of these ores is hematite, 
 containing when pure 70 per cent, of iron. The red or brown 
 hematites are the richer varieties (Lake Superior deposits, con- 
 taining, in some cases, as much as 68 per cent, of iron), while the 
 hydrated hematites, or limonites, usually contain a good deal of 
 water of crystallization and are consequently poorer in iron, not 
 often yielding much more than 50 per cent. iron. 
 
 Oolitic hematite is a variety that exists in the form of spherical 
 grains or nodules. It is important because it sometimes contains 
 limestone and is, therefore, valuable not only for the iron but for 
 the fluxing quality of the lime. The Minette ore of Lothringen 
 (formerly Lorraine), Luxemburg, and France is an enormous 
 deposit of this oolitic hematite, running from 30 to 35 per cent, 
 iron and giving a pig iron containing about 2 per cent, of phos- 
 phorus. This ore is the basis of the iron industry of Germany, 
 France, and Belgium, and, upon judicious mixing of varieties 
 when necessary, is self-fluxing. 
 
 Magnetite (Fe304). — Magnetite contains, when pure, enough 
 iron (72.4 per cent.) to attract the magnet. In the United States 
 it is often mixed with other impurities, such as silica, titanium, 
 and phosphorus, so as to render the ore either too poor in iron to be 
 
10 
 
 THE METALLURGY OF IRON AND STEEL 
 
 smelted profitably, or too high in phosphorus to make good steel, 
 or so high in titanium as to unjustly^ prejudice smelters against 
 it through fear of sticky, infusible slags. 
 
 But other magnetite ores — notably those of Sweden — are the 
 purest ores that exist in large quantities anywhere, and form one 
 of the sources of the Swedish iron and steel, which is famous all 
 over the world for its purity, that is, for its freedom from the ob- 
 jectionable elements sulphur and phosphorus. It is these Swedish 
 products which supply the steel industry of Sheffield with pure 
 material for its tool steel and cutlery. 
 
 Siderite (FeCOg) . — Another variety of iron ore is the so-called 
 "spathic" iron ore, or siderite, which is, however, without any 
 importance in the United States. This forms the famous "clay 
 ironstone" of the Cleveland district in England. It is poor in 
 iron and is therefore no longer smelted in any quantity in the 
 United States in competition with the rich hematites. This ore- 
 is almost always calcined before smelting to expell the carbonic 
 acid, in order to save the blast furnace the extra work of this 
 expulsion in its upper levels. 
 
 SUMMARY OF IRON ORE DEPOSITS OF UNITED STATES. 
 (Compiled by James F. Kemp.)^ 
 
 Available 
 metric tons 
 
 Probable addi- 
 tioa metric tons 
 
 Archean magnetites (Appalachian) 
 
 Lump ores 
 
 Concentrate equivalent from low-grade ore 
 
 Adirondack red hematites 
 
 Pennsylvania soft magnetites 
 
 Cambro-ordovician brown hematites (Appalachian) 
 
 Mesozoic and tertiary brown hematites (Ala., Tenn.) 
 
 Alabama gray and red hematites 
 
 Clinton red hematites (largely Ala., Tenn., N.Y. and Wis.) 
 
 Carbonate ores (Ohio, Penn., and Ky.) 
 
 Lake Superior hematites (Minn., Mich., Wis.) 
 
 Mississippi valley specular and red hematites (Mo.) 
 
 Mississippi valley paleozoic brown hematites (Mo.) 
 
 Mississippi valley tertiary brown hematites (La., Tex., Ark.) 
 Cordilleran magnetites and hematites (Utah, Cal., Wyo., 
 
 Nev., Colo.) 
 
 Total 
 
 Titaniferous ores (Adirondacks and Lake Superior) 
 
 20,000,000 
 40,000,000 
 2,000,000 
 40,000,000 
 65,000,000 
 10,000,000 
 27,500,000 
 505,320,000 
 
 30,000,000 
 
 10,000,000 
 
 2,000,000 
 
 3,500,000,000 
 
 15,000,000 
 
 30,000,000 
 
 260,000,000 
 
 63,800,000 
 
 181,000,000 
 
 15,000,000 
 
 27,500,000 
 
 1,368,000,000 
 
 308,000,000 
 
 72,000,000,000 
 
 5,000,000 
 
 45,000,000 
 
 520,000,000 
 
 55,000,000 
 
 4,678,620,000 
 90,000,000 
 
 74,566,500,000 
 128,500,000 
 
 (a) "Iron Ore Resources of the World," Stockholm, 1910. 
 1 See Jour, Ir. and St. Inst., No. 1, 1910, pp. 611-2. 
 
THE MANUFACTURE OF I»IG IRON 
 
 11 
 
 WORLD'S SUMMARY OF IRON ORE RESERVES, (a) 
 (In million metric tons.) 
 
 
 Actual 
 
 reserves 
 
 Potential 
 
 reserves 
 
 • 
 
 Ore 
 
 Iron 
 
 Ore 
 
 Iron 
 
 Europe 
 
 12,032 
 
 9,855 
 136 
 260 
 125 
 
 4,733 
 5,154 
 
 74' 
 
 156 
 
 75 
 
 41,029 + 
 
 81,822 + 
 
 69 + 
 
 457 + 
 
 much 
 
 12,029 + 
 
 America 
 
 40,731 + 
 
 Australia 
 
 37 + 
 
 Asia 
 
 283 + 
 
 Africa 
 
 much 
 
 
 
 Total 
 
 22,408 
 
 10,192 
 
 123,337 + 
 
 53,136 + 
 
 
 
 (a) "Iron Ore Resources of the "World," Stockholm, 1910. 
 
 United States Deposits and Transportation 
 
 In the United States ores of iron are very widely distributed, 
 as will be seen by reference to the map below. The smelting 
 
 FIG. 2.— FROM KEMP'S "ORE DEPOSITS OF THE UNITED STATES." 
 
 of ore also shows a wide distribution. Blast furnaces are in 
 operation in twenty-four states, including Washington, Minne- 
 
12 THE METALLURGY OF IRON AND STEEL 
 
 sota; and New York on the north; California, Texas, and Ala- 
 bama on the west and south. The great pig-iron centers are: (1) 
 The district that includes Western Pennsylvania and Ohio, which 
 produces more than one-half of the pig iron of the country; (2) 
 Illinois, and (3) Alabama. 
 
 It is not to be supposed that all the deposits marked on the 
 map are extensively worked for their iron. The rich hematite 
 deposits of the Lake Superior district furnish annually about 
 40,000,000 tons, which yield more than three-quarters of the pig- 
 iron production of the country. * The only other districts" which 
 produce more than 1,000,000 tons a year are in the states of 
 Alabama and New York. Most of the other deposits, are mined 
 only for local treatment. In addition, a total of over 2,000,000 
 tons per year are imported from Cuba, Spain, and.,other foreign 
 countries, principally for smelting on or near the Atlantic coast, 
 while Pacific coast furnaces receive ores from China. 
 
 Ore Transportation. — A peculiarity of the, Lake Superior 
 deposits is that most of the ore is. transported' a distance of 800 
 miles or more in order to bring it to the coke. Thus, Western 
 Pennsylvania, Ohio and South Chicago receive the bulk of the 
 
 lI?J^J^^^'^-^^?"-''^^^^aBWit-T ,'niF^^S^^'^^^'*^^'^'^^''^'' 
 
 -; ..-— 
 
 'r ■■. ,J 
 
 
 ■s^^^^Silii^Bi 
 
 
 
 ^^igm 
 
 FIG. 3.~A LAKE SUPERIOR ORE MINE. 
 
 ore shipped from the Lake Superior mines. Since the weight of 
 coke used in the blast furnaces is only about one-half the weight 
 of the ore, it might seem uneconomical to carry the latter to the 
 former. But coke is bulky in proportion to its weight; further- 
 more, it suffers a good deal of waste in transportation in conse- 
 quence of its friability and of the fact that so much of it is broken 
 down into pieces less than an inch in diameter (technically 
 known as "breeze") which is not suitable for charging into the 
 blast furnace. The ore, on the other hand, may be handled by 
 the cheapest and most rapid labor-saving devices. Indeed, in 
 many cases, the ore is never touched by shovels in the hands of 
 
THE MANUFACTURE OF PIG IRON 
 
 13 
 
 H 
 
 Hi 
 
 
 
 1 
 
 
14 
 
 THE METALLURGY OF IRON AND STEEL 
 
 man, but is mined, charged, and discharged in units of several 
 tons each by specially designed machinery. 
 
 The mining and transportation of this great amount of mate- 
 rial is in itself a mighty industry, every advance in which has 
 contributed in no small share to the increasing volume and 
 importance of the iron, steel, and other industries of the United 
 States. 
 
 Mining. — Some of the Lake Superior deposits lie near the 
 surface and are therefore cheaply mined. This is especially true 
 of the soft, earthy deposits of the Mesaba range, which are some- 
 times worked in great open 
 cuts, the ore being loaded 
 upon cars by mammoth 
 steam shovels, or some- 
 times by the caving 
 method, the ore falling 
 by gravity into cars situ- 
 ated in underground tun- 
 -nels. The massive, or rock, 
 ores are more costly to ex- 
 tract, and the utmost skill 
 of American blast-furnace 
 men has been exercised to 
 employ as large a portion 
 of the earthy ores as possi- 
 ble without choking up 
 the furnace. In other 
 countries, especially in Germany, the practice of agglomerating 
 fine ores into briquettes is becoming more prevalent. 
 
 Transportation. — On the lake the ore is transported in boats 
 capable of taking a load of 10,000 or 13,000 tons of ore each. 
 The hatches of these great boats are placed such a distance apart 
 that the hinged ore chutes of the bins may be swung down and, 
 when the gates are opened, the ore allowed to flow directly into 
 the hold of the vessel. In a few minutes the vessel has received 
 her full cargo and is ready to start on its long journey down the 
 chain of inland lakes. • 
 
 Unloading. — The unloading of boats is accomplished with 
 almost as great celerity as the loading, and by means of mechan- 
 ical unloading machinery a steamer containing as much as 
 10,300 tons of ore has been completely discharged in 4 hours and 
 
 FIG. 5.— LOADING AN ORE BOAT. 
 
THE MANUFACTURE OF PIG IRON 
 
 15 
 
 30 minutes. Nor is any time wasted in coaling the vessel for a 
 second j ourney up the lakes and back. Great machines pick up 
 whole railroad cars of fuel and empty them bodily into the chute 
 
 FIG. 6.— BROWN-HOIST APPARATUS UNLOADING AN ORE BOAT. 
 
 which connects with the bunkers of the vessel, many of the ore 
 steamers being so constructed that this whplesale loading of coal 
 can go on at the same time that ore is being discharged. 
 
 ■HULETT ELECTRIC UNLOADER. 
 
 Prices. — Iron ores are divided into two great classes, known 
 as " Bessemer''" and "Non-Besseiiier"; in the former class are 
 those ores in which the percentage of iron is more than 1000 
 
16 
 
 THE METALLURGY OF IRON AND STEEL 
 
 times the percentage of phosphorus. This name arises through 
 the circumstance that only when the phosphorus is as low as this 
 can we make from them a steel, with the blast furnace followed 
 by the acid Bessemer process, with phosphorus under 0.100 per 
 cent. The price of Bessemer ores is about 75 cents per ton 
 more than non-Bessemer ores on this account. At present 
 (1911) Mesaba Bessemer ores sell at about $5 per ton, which is 
 twice the price they commanded 12 to 15 years ago. 
 
 Handling Raw Material at a Modern Furnace 
 
 Behind the blast furnace are situated long rows of storage 
 bins, one of which is shown in elevation in Fig. 9. These bins 
 are filled by bottom-dumping railroad cars which bring the ore to 
 the furnaces, or by mechanical apparatus from the great piles of 
 ore stored conveniently near. Between and under these two 
 rows of bins runs a track on which ore larries are transferred back 
 
 FIG. 8— ore-handling MECHANISM AT BLAST FURNACE. 
 
 and forth, being first filled with a weighed amount of ore, lime- 
 stone, or fuel, and then switched into a position from which they 
 can deposit their contents into the loading skip of the blast 
 furnace. 
 
 Loading the Furnace, — The next step in the handling of the 
 raw material is to bring the ore, together with the necessary fuel 
 and flux, into the mouth of the huge furnace that is to convert it 
 into pig iron. In one of the big modern American furnaces, work- 
 ing at top speed, the amount of material which must be dumped 
 into the top during 24 hours will frequently exceed 2000 tons, 
 and the charging must go on for 365 days a year with never a 
 delay of more than a few minutes at a time. 
 
THE MANUFACTURE OF PIG IRON 
 
 17 
 
 In the modern type of furnace this loading is accomplished 
 altogether by mechanism operated and controlled from the 
 ground level, and no men are required to work at the top of the 
 
 FIG. 9.— CROSS-SECTION OF BLAST FURNACE AND SKIP HOIST. 
 
 furnace. In Fig. 9 is a section of such a furnace showing one 
 method of loading — a double, inclined skipway extending above 
 the top of the furnaqe. One skip is seen discharging its load of 
 ore, or fuel and flux, into the hopper, while the second skip is at 
 the bottom of the incline ready to be loaded with its charge. 
 
 2 
 
18 THE METALLURGY OF IRON AND STEEL 
 
 Double Bell and Hopper. — The upper hopper of the furnace is 
 closed at the bottom by an iron cone, known as a "bell." This 
 bell is held against the bottom of the hopper by the counter- 
 weighted lever shown, and is opened by operating the cylinder 
 M, to allow the charge to fall into th^ main hopper, 7, of the 
 blast furnace. In this way the hopper of the furnace is pro- 
 gressively filled with ore, flux, and fuel. This hopper, 7, is also 
 closed at the bottom by a similar bell, A. The operation of this 
 bell is also controlled from the ground, and when a charge has 
 been deposited in the hopper it is opened and the contents 
 allowed to fall in an annular stream, distributing itself on top 
 of the material already in the furnace which is within a few feet 
 of the bottom of the bell. As the upper bell, B, is now held up 
 against the bottom of the upper hopper, there is never a direct 
 opening from the interior of the blast furnace to the outer air, 
 so that the escape of gas, resulting formerly in the long flame 
 rising from the top of the blast furnace whenever material was 
 dropped into the interior, no longer occurs at our modern 
 plants. 
 
 This is not the only means of handling the raw material for the 
 blast furnace. Several varieties of mechanism are extensively 
 used, but the description given heretofore will serve to illustrate 
 the general principles of labor-saving mechanisms in connection 
 with charging the blast furnace. The details of the distribution 
 are of the utmost importance and largely affect the production, 
 the quality of iron, the coke consumption, and, above all, the life 
 of the lining. Cases are on record where apparently trivial 
 changes in this detail have revolutionized the work of the furnace. 
 
 The Blast Furnace and Accessories 
 
 The blast furnace itself consists of a tall cylindrical stack lined 
 with an acid (silicious) refractory fire-brick, the general form and 
 dimensions being shown in Fig. 9. The hearth or crucible is the 
 straight portion occupying the lower 8 ft. of the furnace. Above 
 that extends the widening portion, called the bosh, which reaches 
 to that portion in the furnace having the greatest diameter. The 
 stack extends throughout the remainder of the furnace, from the 
 bosh to the throat. The brickwork o^ the hearth is cooled by 
 causing water to trickle over the outside surface. 
 
 Tuyeres.— ThTough the lining of the furnace, just at the top 
 of the hearth, extend the tuyeres — 8 to 16 pipes having an 
 
THE MANUFACTURE OF PIG IRON 
 
 19 
 
 internal diameter of 4 to 7 in., through which hot blast is driven 
 to burn the coke and furnish the heat for the smelting operation. 
 The "tuyere notches," or openings through which the tuyere 
 pipes enter, as well as the tuyeres themselves, are surrounded by 
 hollow bronze rings, through which cold water is constantly 
 flowing to protect them from meltiug at the inner ends. The 
 number and size of the tuyeres is in proportion to the diameter 
 of the hearth, the volume and pressure of the blast, etc., the 
 blast being given sufficient velocity to carry it, distributed as 
 evenly as. possible, to the very center of the furnace. 
 
 Eye Sight 
 
 FIG. 10.— PARTS OF A BLAST FURNACE TUYERE. 
 
 Discharge Holes. — On the side of the furnace, and 30 to 40 in. 
 below the level of the tuyeres, the ''cinder notch" or "monkey" is 
 situated. This is protected by a water-cooled casting, and the 
 hole is closed by forcing into it an iron plug around which the 
 cinder chills and makes a tight joint. The level of the tuyeres 
 is the point above which the cinder must not be allowed to 
 accumulate. 
 
 In the front, or breast, at the very bottom level of the crucible, 
 is the iron tap-hole, from which all the liquid contents of the 
 furnace can be completely drained. This is a large hole in the 
 brickwork, and is closed with several balls of clay. 
 
 Bosh.- — The hottest part of the furnace is near the tuyeres 
 and a few feet above them. In order to protect the brickwork of 
 
20 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 11.— A BATTERY OF BLAST FURNACE BLOWING ENGINES. 
 
THE MANUFACTURE OF PIG IRON 21 
 
 the bosh from this heat, a number of hollow wedge-shaped cast- 
 ings are placed therein, through which cold water circulates. 
 The brickwork is furthermore protected by a deposition of a 
 layer of carbon, similar to lampblack, on its internal surface, 
 covered by a layer of a sort of slag, or cinder, replacing part of 
 the brickwork. The deposition of carbon comes about through 
 the reaction of the furnace operation itself, in the following 
 manner: For the correct conduct of the smelting operation, and 
 especially for carrying of the sulphur" in the slag, it is necessary 
 that a very powerful reducing influence must exist; this reducing 
 influence is produced by an excess of coke, and one of its results 
 is the precipitation of finely divided carbon on the internal walls 
 of the furnace. It is this thin layer of cinder and carbon which is 
 most effective in protecting the acid lining of the furnace from 
 the corrosive action of the basic furnace slag. Other methods of 
 protecting the bosh walls are in use, one of the most successful 
 being a plain steel jacket thoroughly cooled externally by sprays 
 and lined with a single course of brick. This avoids the benching 
 or stepping which occurs with the construction above described. 
 
 Hot Blast. — The air for smelting is driven into the furnace by 
 immense blowing engines ranging up to 2500 H. P. each, and 
 capable of compressing 50,000 to 65,000 cu. ft. ( = 4875 Ib.^) of 
 free air per minute to a pressure of 15 to 30 lb. per sq. in., which 
 is about *what one furnace requires. It actually requires about 
 4 to 5 tons of air for each ton of iron produced by the furnace. 
 After leaving the engines and before coming to the furnace, the 
 air is heated to a temperature of 425 to 650° C. (800 to 1200° F.) 
 by being made to pass through the hot-blast stoves. 
 
 Hot-blast Stoves. — Each furnace is connected with four or five 
 stoves. These are cylindrical tanks of steel about 110 ft. high 
 and 22 ft. in diameter, containing two or more fire-brick cham- 
 bers. One of these chambers is open, and the others are filled 
 with a number of small flues (see Fig. 12). Gas and air are 
 received in the bottom of the open chamber, B, in which they 
 burn and rise.. They then pass downward through the several 
 flues in the annular chamber surrounding B, and escape at the 
 bottom to the chimney as waste products. In passing through 
 the stove they give up the greater part of their heat to the brick- 
 work. After this phase is ended, the stove is ready to heat the 
 blast. 
 
 * At 70* F. and atmoBpheric pressure, each 1000 cu. ft. of air weighs 75 lb. 
 
22 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 12.— HOT-BLAST STOVE. From Howe, "Iron, Steel and other Alloys." 
 
THE MANUFACTURE OF PIG IRON 23 
 
 The blast from the blowing engine enters at the bottom of the 
 flues,- ^, passes up through the outer chamber, and down through 
 B to the furnace. In this passage it takes up the heat left in the 
 brickwork by the burning gas and air. Sometimes there are 
 three passes, instead of two as described. In a blast-furnace 
 plant, one stove is heating the blast while the other three are 
 simultaneously in the preparation stage, burning gas and air. 
 By changing once an hour a pretty regular blast-temperature is 
 maintained. The gas used for the heating is the waste gas from 
 the blast furnace itself, which amounts to about 90,000 cu. ft. 
 per minute at a temperature of 235° C. (450° F.), and has a 
 calorific power of about 85 to 95 B. T. U. per cubic foot. The 
 latent and available heat of this gas is equivalent to approximately 
 50 per cent, of that of the fuel charged into the furnace. Only 
 about 30,000 cu. ft., or one-third of this gas, is needed for keeping 
 the stoves hot, and the remaining two-thirds is used to produce 
 power. 
 
 Power from Waste Gas. — The waste gas comes down the down- 
 comer Tj Fig. 13, deposits dirt in the dust-catcher TT, is carefully 
 cleaned and scrubbed, and is then led to the stoves or power pro- 
 ducer. This gas varies in composition, but will average about 
 61 per cent, nitrogen, 10 to 17 per cent. COg, and 22 to 27 per 
 cent. CO. The latter can be burned with air to produce heat. 
 C0 + 0-=C02 (generates 68,040 calories). 
 
 If burned under boilers, the available gas will generate enough 
 power to operate the blowing engines, hoisting mechanism, and 
 other machinery used in connection with the furnace. At several 
 plants the gas available for power is cleaned carefully and uti- 
 lized in gas engines, whereby much more power is obtained, the 
 excess being usually converted into electricity and transmitted 
 to more distant points. 
 
 Smelting Practice and Products 
 
 The furnace is filled with alternate layers of fuel, flux, and ore, 
 down to the top of the smelting zone. The exact location of this 
 zone will be dependent upon the volume and pressure of blast, 
 size of furnace, character of slag made, etc., but will extend from 
 the level of the tuyeres to a few feet above them, or about to the 
 top of the bosh. It will require perhaps 15 hours for the material 
 to descend from the top of the furnace to the smelting zone. 
 During this descent, it. is upheld partly by the resistance of the 
 
24 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG, 13, 
 
THE MANUFACTURE OF PIG IRON 
 
 25 
 
26 
 
 THE METALLURGY OF IRON AND STEEL 
 
 uprushing column of hot gases/ partly by its friction on the walls 
 of the furnace, and partly by the loose column of coke which 
 extends through the smelting zone and to the bottom of the fur- 
 nace, and which alone resists melting in the intense heat of this 
 
 Fusion Level 
 
 Legend t ^ Lumps of Coke 
 
 Lumps of Iron Ore 9 
 
 Lumps of Lime O 
 
 Drops o£ Slag iJ 
 
 Drops of Iron # 
 
 Layer of Molten Slag W^^l 
 
 Layer of Molten Iron =:-^±Efi5 
 
 FIG. 15. — ^From Howe, "Iron, Steel, and other Alloys." 
 
 zone. The blast entering the furnace through the tuyeres, con- 
 sists of 23 per cent, by weight of oxygen and 77 per cent, by 
 weight of nitrogen, together with varying amounts of water 
 
 1 The full drop from the blast pressure to practically atmospheric pressure at.the throat 
 represents the resiatence of the stock to the upward passage of the gas and of course this 
 represents also the resistance offered by the gas to the descent of the stock. This re- 
 sistance is in many cases nearly equal to the weight of the stock. 
 
THE MANUFACTURE OF PIG IRON 27 
 
 vapor from moisture in the air (see page 32). The nitrogen is 
 practically inert chemically and performs no function other than 
 that of absorbing heat in the smelting zone and giving it out at 
 higher levels. The oxygen attacks all the coke in the smelting 
 zone and as much of it below the level of the tuyeres as is not 
 covered by accumulations of iron and slag in the hearth, produc- 
 ing a large volume of carbon monoxide gas (CO) and a tempera- 
 ture which may exceed 3000° F.^ The 00 and nitrogen pass up 
 between the particles of solid material, to which they give up the 
 greater part of their heat. The former also performs certain 
 chemical reactions, and thus in both ways the rising column of 
 gases prepares the charge for its final reduction in the smelting 
 zone. 
 
 Chtmical Reactions in the Upper Levels. — As soon as the iron ore 
 enters the top of the furnace, ]two ^reactions begin to take place 
 between it and the gases : 
 
 (1) 2Fe203 + 8CO-7 C02 + 4Fe-l-C; 
 
 (2) 2Fe,03+ CO==2FeO+Fe203 + C02; 
 
 and this continues with increasing rapidity as the material be- 
 comes hotter. The carbon formed by reaction No. 1 deposits in 
 a form similar to that of lampblack on the outside and in the 
 interstices of the ore. This reaction, however, is opposed by two 
 reactions with carbon dioxide gas : 
 
 (3) Fe + C02=Fe0 + C0; 
 
 (4) C +C02 = 2C0. 
 
 Reaction No. 3 begins at a temperature of about 300*^ C. (575° 
 F.), which is met with about 3, or 4 ft. below the top level of the 
 stock; and No. 4 begins at about 535° C. (1000° F.), or 20 ft., 
 below the stock line. Reaction No. 4 is so rapid that the depo- 
 sition of carbon ceases .at a temperature of 590° G. (1100° F.). 
 All the way down the ore is constantly losing a proportion of its 
 oxygen to the gases. At higher temperatures than 590° C, 
 FeO is stable and practically all of the FejOg has been reduced. 
 
 (5) Fe80, + CO = 3FeO + C02. 
 
 The reaction between iron oxide and solid carbon begins at 400° 
 
 0. (750° F.). 
 
 (6) Fe203 + 3C=2Fe + 3CO. 
 
 1 It is of minor importance whether the CO gas is formed directly or as a result of the 
 two following reactions: 
 
 C+03=C02 
 
 CO2 + C = 2C0 
 
28 THE METALLURGY OF IRON AND STEEL 
 
 At 700° C. (1300° F.) solid carbon begins to reduce even FeO: 
 
 (7) FeO + C=Fe + CO. 
 
 Practically all the iron is reduced to a spongy metallic form by 
 the time the temperature of 800° C. (1475° F.) is reached. This 
 is about 45 ft. from the stock line and less than 30 ft. above the 
 tuyeres. At 800° C. the limestone begins to be decomposed by 
 the heat; and only CaO comes to the smelting zone: 
 
 (8) CaC03 = CaO + C02. 
 
 The foregoing facts are summarized in Fig. 16, which is adapted 
 from H. H. Campbell, with certain changes.^ It is not to be sup- 
 posed that these figures are exactly correct for the . different 
 levels, and it is probable that they change from day to day and 
 from furnace to furnace, but a general idea may be obtained from 
 this sketch. It will be seen that the upper 15 or 20 ft. of stock 
 is a region of FcgOg and Fe3Ci4, gradually being converted to 
 FeO by CO gas, and forming quantities of COg gas. If these 
 reactions were the only ones, the top gases would contain no CO 
 and would have jio calorific power, but reaction No. 1 produces 
 both metallic iron and carbon, both of which reduce COg and 
 thus waste much energy, as far as the blast furnace is concerned: 
 
 Reaction No. 3, Fe + C02=Fe0 + C0, absorbs 2340 calories, but 
 wastes 68,040 calories. 
 
 Reaction No. 4, C + C02 = 2C0, absorbs 38,880 calories. 
 
 From 20 to 35 ft. below the stock line is the region of FeO, 
 gradually being converted to metallic iron sponge by carbon. 
 At the lower level of this zone the limestone loses its COj, which 
 joins the other furnace gases. From 35 ft. down to the smelting 
 zone is the region of metallic iron. This spongy iron is impreg- 
 nated with deposited carbon which probably to some extent 
 soaks into it and dissolves, in a manner like in nature but not in 
 degree to the way ink soaks into blotting-paper. This carburi- 
 zation of the iron reduces its melting-point and causes it to be- 
 come liquid at a higher point above the tuyeres than it otherwise 
 would. 
 
 On reaching the smelting zone the iron riielts and trickles 
 
 ^ quickly down over the column of coke, from which it completes 
 
 its saturation with carbon. At a corresponding point the lime 
 
 unites with the coke ash and impurities in the iron ore, forming a 
 
 fusible slag which also trickles down and collects in the hearth. 
 
 1 See pp. 54 and 62 of Book No. 2. 
 
THE MANUFACTURE OF PIG IRON 
 
 29 
 
 It is during this transit that the different impurities are reduced 
 by the carbon, and the extent of this reduction determines the 
 characteristics of the pig iron, for in this operation, as in all 
 smelting, reduced elements are dissolved by the metal, while 
 those in the oxidized form are dissolved by the slag. Only one 
 exception occurs, namely, that iron will dissolve its own sulphide 
 (FeS) and, to a less extent, that of manganese (MnS), but not 
 that of other metals, as, for instance, CaS. 
 
 (1) 2Fe20a+8CO = 7C02+4Fe+C (begins) 
 
 (2) 2Fe203+CO = 2FeO+C02+Fe203 (begins) 
 
 575"' (3) Fe + C02=FeO+CO (begins) 
 750° (6) Fe203+3C=2Fe+3CO 
 
 1025° (4) + 002 = 200 (rapid) 
 1100° Deposition of carbon ceases 
 
 1300° (7) FeO + C=Fe+CO (begins) 
 
 1475° 
 
 (7) FeO + C=Fe+C0 (complete) 
 
 (8) CaCo3 =CaO+C02 
 
 1830° (4) C + CO2 = 2C0 (prevails) 
 
 CO2 cannot exist below this level 
 
 Smelt- (9) Si02 + 2C=S1+2C0 
 
 ing (10) FeS+CaO+0=CaS+Fe+CO 
 „ (11) Mn02+2C = Mn+2CO 
 
 ^one (12) P20b+5C=2P+5CO 
 
 FIG. 16,— DIAGRAM SHOWING CHEMICAL ACTION IN BLAST FURNACE. 
 
 Chemical Reactions in the Smelting Zone. — -There is always a 
 large amount of silica present in the coke ash, and some of this is 
 reduced according to the reaction: 
 
 (9) Si02 + 2C = Si + 2CO. 
 The extent of this reaction will depend on the length of time the 
 iron takes to drop through the smelting zone, the relative inten- 
 sity of the reducing influence, and the avidity with which the slag 
 
30 THE METALLURGY OF IRON AND STEEL 
 
 takes up silica. A slag with a high melting-point will trickle 
 sluggishly through the smelting zone and cause the iron to do 
 the same, to some extent, thus giving it more chance to take up 
 silicon. A higher temperature in the smelting zone, which in- 
 creases disproportionately the avidity of carbon for oxygen, will 
 promote reaction No. 9. We can produce this higher tempera- 
 ture by supplying hotter blast. A larger proportion of coke to 
 burden^ will further promote this reaction, because this not only 
 increases the amount of the reducing agent, but also raises the 
 temperature and, therefore, the chemical activity of this agent. 
 Thus the coke has both a physical and a chemical influence in 
 increasing the intensity of the reduction in the smelting zone. 
 A basic ^lag, because of its avidity for silica, will oppose reaction 
 No. 9; it is one of the principal means of making low-silicon pig 
 iron. This is in spite of the fact that the basic slags are sluggish, 
 and therefore trickle slowly through the smelting zone, thus 
 exposing the silica longer to reducing influence, and also in- 
 creasing the temperature of "the materials in this zone (1) by 
 causing them to pass through it more slowly and absorb more heat, 
 and (2) by reducing the level of ihe smelting zone nearer to the tu- 
 yeres, which confines the intense temperature to the smaller 
 area, or, in other words, diminishes the passage of heat upward. 
 Sulphur comes into the furnace chiefly in the coke. It is 
 partly in the form of iron sulphide (FeS), and partly in the form 
 of iron pyrites (FeSj), which loses one atom of sulphur near the 
 top of the stock>nd becomes FeS, which will dissolve in the iron 
 unless converted to sulphide of calcium (CaS). This is brought 
 about, according to the explanation of Professor Howe, by the 
 following reaction: 
 
 (10) FeS + CaO + C = CaS+Fe + CO. 
 
 CaS passes into the slag, and the odor of sulphur is very strong 
 when the slag is running from the furnace. It is evident from 
 reaction No. 10 that intense reduction, which increases the silicon 
 in the iron, has the contrary effect on the sulphur, and this 
 explains the common observation that iron high in silicon is 
 liable to be low in sulphur. Indeed, this relation is so constant 
 that it is almost a rule. There are two exceptions, however: 
 (1) Increasing the proportion of coke has a doubly strong influ- 
 ence in putting silicon in the iron. As regards sulphur, on the 
 
 1 The burden is the amount of material that the coke has to melt. We lighten the bur- 
 den by increasing the amount of coke, and vice versa. 
 
THE MANUFACTURE OF PIG IRON 
 
 31 
 
 other hand, it has a self-contradictory effect; by increasing the 
 amount of sulphur in the charge it tends to increase it in the 
 iron, which is partly or wholly counteracted by its effect in 
 reaction No. 10. (2) A basic slag may hold silicon from the iron, 
 and it also holds sulphur from the iron by dissolving CaS more 
 readily. In other respects the conditions which make for high 
 
 TABLE II.— COMPOSITION OF BLAST-FURNACE SLAGS. 
 From H. H. Campbell. P. 50 of No. 2. 
 
 
 Slag 
 
 Iron 
 
 
 
 
 SIO2 
 
 A1,0, 
 
 CaO 
 
 MgO 
 
 FeO 
 
 s 
 
 Total 
 
 not in 
 elud- 
 ing S 
 
 Si 
 
 s 
 
 Remarks 
 
 1 
 
 33.10 
 32.27 
 24.26 
 32.68 
 32.28 
 34.50 
 34.98 
 34.70 
 33.68 
 29.86 
 28.95 
 30.63 
 32.55 
 30.08 
 31.46 
 36.08 
 37.19 
 36.86 
 ■32.06 
 33.57 
 35.38 
 36.35 
 33.70 
 35.11 
 35.10 
 35.84 
 
 A^verage 
 
 33.21 
 34,84 
 31.77 
 35.55 
 
 4.verage 
 
 33.15 
 30.73 
 34,75 
 35.35 
 
 14.92 
 14.57 
 11,^3 
 13.50 
 9.38 
 7.94 
 12.05 
 11.44 
 11.93 
 12.04 
 12.04 
 10.47 
 11.13 
 11,44 
 11.50 
 12.85 
 12.65 
 10.74 
 11.97 
 10.65 
 11.76 
 10.21 
 12,56 
 14.21 
 14.75 
 14.34 
 
 3 for he 
 
 13.67 
 11.75 
 11.98 
 12.05 
 
 B for m 
 
 10.27 
 11.32 
 11,30 
 14.43 
 
 40.76 
 41.02 
 40.25 
 43.28 
 46.95 
 46.47 
 41.33 
 41.27 
 45.96 
 45.20 
 49,30 
 49.13 
 47,16 
 46.36 
 44.85 
 41.69 
 35.47 
 42.46 
 42.46 
 44.11 
 38.19 
 40.10 
 38.12 
 22.48 
 27.95 
 32.71 
 
 t fuma 
 
 40.68 
 41.30 
 45.58 
 40.52 
 
 oderate 
 
 45.57 
 47.36 
 40.12 
 29.69 
 
 9.67 
 
 10.30 
 
 13.28 
 
 9.44 
 
 9.52 
 
 10.47 
 
 9.62 
 
 9.96 
 
 6.69 
 
 11.41 
 
 8.46 
 
 7.49 
 
 6.61 
 
 8.76 
 
 10.41 
 
 7,25 
 
 11.32 
 
 6.62 
 
 10.25 
 
 8.55 
 
 12.32 
 
 10.95 
 
 11.60 
 
 22.38 
 
 22.28 
 
 17.46 
 
 Qes: 
 
 11.08 
 9.79 
 9.05 
 8.86 
 
 or coo 
 
 9.81 
 
 8.36 
 
 10.86 
 
 20.71 
 
 , 
 
 
 98.45 
 98.16 
 98.32 
 98.90 
 98,13 
 99.38 
 97.98 
 97.37 
 98.26 
 98.51 
 98.75 
 97.71 
 97.45 
 96.64 
 98.22 
 98.41 
 97.53 
 97.31 
 97,37 
 97.69 
 98.53 
 98.60 
 98.30 
 100.12 
 100.08 
 100.35 
 
 98.64 
 97.68 
 98.38 
 97.66 
 
 98.80 
 
 97.75 
 
 98.29 
 
 100.18 
 
 3.37 
 3.18 
 4.81 
 1.26 
 0.70 
 0.69 
 2.60 
 2.32 
 1.27 
 1.27 
 0.57 
 0.26 
 0.15 
 0.68 
 0.20 
 2,15 
 1.92 
 1.50 
 1.59 
 0.94 
 1.13 
 0.66 
 0.50 
 1.37 
 1.85 
 1.60 
 
 3.79 
 2.46 
 1.27 
 1.79 
 
 0.88 
 0.35 
 0.81 
 1,61 
 
 tr. 
 
 tr. 
 .01 
 .06 
 .11 
 .05 
 .03 
 .02 
 .02 
 .02 
 
 tr. 
 ,02 
 .03 
 .03 
 .07 
 .020 
 .029 
 .028 
 .032 
 .017 
 ,040 
 .095 
 .101 
 .048 
 .038 
 .034 
 
 tr. 
 .025 
 .020 
 .027 
 
 .07 
 ,03 
 .063 
 ,040 
 
 Cuban ore, Hot furnace. 
 
 9, 
 
 
 
 Cuban ore, Hot furnace. 
 
 3 
 
 
 
 Cuban ore, Hot furnace. 
 
 4 
 
 
 
 Cuban ore, Warm furnace. 
 
 5 
 
 
 
 Cuban ore, Cool furnace- 
 
 6 
 
 
 
 Cuban ore, Cool furnace. 
 
 7 
 
 
 
 Spanish ore, Hot furnace. 
 
 8 
 
 
 
 
 q 
 
 
 
 Spanish ore, Hot furnace. 
 
 10 
 
 
 
 Spanish ore, Hot furnace. 
 Spanish ore. Cool furnace. 
 Spanish ore, Cool furnace. 
 
 11 
 
 
 
 12 
 
 
 
 13 
 
 
 
 Spanish ore, Cool furnace. 
 
 14 
 
 
 
 Spanish ore, Cool furnace. 
 
 15 
 
 
 
 Spanish ore, Onol fumnnp. 
 
 16 
 17 
 18 
 19 
 20 
 21 
 22 
 23 
 24 
 
 0.54 
 0.90 
 0.63 
 0.63 
 0,81 
 0.90 
 0,99 
 0.32 
 
 1.62 
 1.70 
 1.54 
 1.76 
 1.74 
 1.60 
 l.i?S 
 0.96 
 
 Lake ore 
 and part 
 anthracite 
 
 ' Coal; 
 mostly 
 Connells- 
 ville coke 
 Lake ore 
 
 ■ and Con- ■ 
 nelsville 
 coke. 
 
 Cuban ore. 
 
 Hot furnace 
 
 Fairly hot 
 
 Fairly hot 
 
 Fairly hot 
 
 Normal 
 
 Normal 
 
 Cool 
 
 Cool' 
 
 Av. of 8 wks 
 
 25 
 
 
 
 Av.of 7 wks. 
 
 26 
 
 
 
 Av.of 7 wks. 
 
 
 
 
 
 
 
 
 Spanish ore. 
 
 
 
 
 Spanish ore. 
 
 t 
 
 0.68 
 1 funi 
 
 1.66 
 tace: 
 
 Lake ore. 
 Cuban ore. 
 
 
 
 
 Spanish ore. 
 
 
 1.26 
 
 1.40 
 
 Lake ore. 
 Lake ore. 
 
 
 
 
 
 
32 THE METALLURGY OF IRON AND STEEL 
 
 silicon make also for low sulphur. Particularly is this true of a 
 high temperature in the smelting zone, and the term "hot iron" 
 has come to be synonymous in the minds of blast-furnace fore- 
 men with iron high in silicon and low in sulphur. 
 Manganese is reduced by the following reaction: 
 
 (11) Mn02 + 2C = Mn + 2CO. 
 
 The amount of manganese in the iron is dependent, to a certain 
 extent, upon the character of the ores charged, but it may be 
 controlled somewhat by the character of slag made, because an 
 acid slag will carry a large amount of manganese away in the 
 form of silicate of manganese (MnSiOg). 
 
 With a certain unimportant qualification, the amount of 
 phosphorus in the iron is controlled by the character of raw 
 materials charged, and districts or countries having high-phos- 
 phorus ores must make high-phosphorus irons. This is not an 
 insuperable objection, because the presence of phosphorus, even 
 up to 1.5 per cent., is desired in certain irons for foundry use, 
 and the basic processes for making steel can remove this element. 
 
 The chemical influence of the blast furnace is a strongly 
 reducing one, and this is produced in, order, first, to reduce the 
 iron from the ore; second, to get rid of the sulphur, and third, to 
 saturate the iron with carbon. Many attempts have been made 
 to provide a process wherein the reducing influence was not so 
 strong, and thus to produce a purer material than pig iron, 
 because it is the intensity of the reduction which contaminates 
 the iron with carbon and silicon. The great weakness of all such 
 processes, however, is that they do not get rid of the sulphur, 
 which is the most objectionable impurity that iron is liable to 
 contain and which is not cheaply removed by any process 
 after once it makes its way into the iron. Finally, to saturate 
 the iron with carbon rentiers the blast-furnace operation very 
 much cheaper, because pure iron melts at a temperature much 
 higher than can readily be obtained in the furnace, a,nd iron is 
 handled much more cheaply molten than in the solid or pasty 
 state. Even the presence 6f silicon is an advantage, as we shall 
 see in Chapter XII. 
 
 Drying the Blast. — The water vapor blown into the furnace 
 (derived from the moisture of the air) is equivalent to from 1/3 
 to 2 gal. of water per 10,000 cu. ft. of blast, or 1 1/3 to 8 gal. per 
 minute, depending on the humidity of the atmosphere. Though 
 
THE MANUFACTURE OF PIG IRON 33 
 
 this steam is as hot as the blast, it materially cools the smelting 
 zone of the furnace by dissociating there: 
 
 H20 + C==2H + CO (absorbs 28,900 gram calories); 
 
 or 1 lb. of steam absorbs 728,280 calories. The hydrogen and 
 oxygen reunite in a cooler part of the furnace and return the 
 same amount of heat, but this does not compensate for that 
 taken away from the smelting zone, where it is most needed. 
 For this reason several American and European plants have 
 adopted James Gayley's process of .drying the air by refrigeration 
 before it is drawn into the blowing engine. This results in 
 valuable saving in fuel and greater regularity of furnace working. 
 In fact, so great is the economy shown in this respect that there 
 was a tendency at first to receive the results with skepticism. 
 
 The same condition prevailed upon the introduction of the 
 hot blast, early in the last century, for the same reasons: That 
 no explanation was forthcoming of the observation that th6 fuel 
 saved in practice was so much greater in proportion than the 
 heat saved or restored to the crucible. But it was pointed out 
 by J. E. Johnson, Jr., in the discussion of Mr. Gayley's results, 
 that certain functions of the smelting process can only be carried 
 out above a certain " critical temperature," which was in fact the 
 free-running temperature of the slag, and that the heat available 
 above this temperature, and not the total heat developed in the 
 furnace, was the measure of the economy that can be attained. 
 Approximately speaking, the interval between the "critical 
 temperature'' and the theoretical combustion temperature, is 
 the measure of this available heat, and as the critical temperature 
 is, roughly speaking, four-fifths of the theoretical combustion 
 temperature, any change which increased or decreased the latter 
 by a given proportion, will increase or decrease the available 
 heat by a proportion five times as great. This simple explana- 
 tion enables us to calculate with accuracy the effect of any given 
 change in known conditions.^ 
 
 Johnson also pointed out that the theoretical combustion 
 temperature might be increased by enriching the air in oxygen, 
 instead of reducing the moisture in it, since this enrichment 
 would incre,ase the calorific efficiency. A Belgian blast furnace 
 
 1 If further explanaiion of this argument is needed, it may be found in the following 
 simile: Water boils at 212° F. If the temperature of a boiler is 262°, there is a certain 
 pressure of steam; if we increase the temperature only 50°, we greatly increase the pressure; 
 yet 50° appears small in comt)arison to 262°. 
 3 
 
34 THE METALLURGY OF IRON AND STEEL 
 
 has recently adopted this type of practice, and the results will be 
 watched with great interest, as well on practical as on theoretical 
 grounds. 
 
 Slag Disposal. — The slag, on account of its lower specific gravity, 
 floats in the bath of iron in the hearth and accumulates until it 
 reaches nearly to the level of the tuyeres. The cinder-notch is 
 then opened by withdrawing the iron plug and piercing the skull 
 of chilled cinder with a steel bar, and the cinder drawn off down 
 to the lever of the notch. This is done four or five times be- 
 tween each iron tap, i.e., each six hours. It flows down an 
 inclined iron runner for a distance of 15 to 30 ft. and pours into 
 an iron ladle on a standard-gage railroad track, whence it is 
 drawn away by a locomotive and poured out on the slag dump." 
 Slag varies in composition according to the will of the blast- 
 furnace manager, and some typical analyses are given in Table 
 11. Slags high in lime are sometimes treated with additional 
 lime to make a good grade of Portland cement, known as "Puzzo- 
 lini." The amount of cinder made will depend on the amount 
 of silica, alumina, etc., in the ore, the amount of coke ash, and 
 the amount of flux, which will also depend on the desired slag 
 analysis. Under favorable circumstances, the slag may weigh 
 slightly less than half the iron; under other conditions it may 
 weigh nearly twice the iron, depending on the amount of im- 
 purity in the charge. 
 
 Weight of Slag, — The amount of slag may be calculated from 
 the amount of lime (CaO) in the furnace, which may be calcu- 
 lated from the percentage of lime in the limestone and other 
 materials charged into the furnace. Since all the lime charged 
 goes into the slag, the amount of the latter will be equal to the 
 weight of lime divided by the percentage of the lime in the slag. 
 Thus, if we use per ton of iron 1300 Ib.^f limestone, containing 
 50 per cent, of lime, there will be 650 lb. of lime charged for every 
 ton of iron made. If the slag made contains 40 per cent, lime, 
 then the weight of slag will be 650/ .40 = 1625 lb. per ton of iron 
 made. 
 
 Iron Disposal. — Immediately after the last "flushing," i.e., 
 removal of cinder, the tap hole or iron notch is opened by several 
 men drilling a hole in it with a heavy, pointed steel bar. Out of 
 this notch flows 100 to 150 tons of hquid pig iron, with which is 
 carried along 30 tons or so of slag. The "skimmer^ is situated 
 about a dozen feet from the front of the furnace. It is an iron 
 
THE MANUFACTURE OF PIG IRON 35 
 
 plate extending down almost to the bottom of the runner. The 
 slag is deflected by this plate into a runner of its own, which leads 
 it off to a slag ladle such as described before. The heavier pig 
 iron flows under the skimmer and is distributed to six or seven 
 brick-lined ladles on a standard-gage railroad track. It is then 
 drawn away to the steel works, or, if not wanted there, is poured 
 into iron molds at the pig-casting machine. 
 
 Mechanical Pig-Molding Machine, — There are several types of 
 molding machine, but a common form is illustrated in Figs. 18-9, 
 and consists of a long continuous series of hollow metallic molds 
 carried on an endless chain. 'D is the pig-iron ladle pouring 
 
 FIG. 17. 
 
 metal into the spout, from whence it overflows into the molds as 
 they travel slowly past. The pig iron chills quickly against the 
 metallic molds, and by the time it reaches the other end of the 
 machine, it consists of a solid pig of iron which drops into the 
 waiting railroad car as the chain passes over the sheave. The 
 pig iron is now in^ form convenient for transportation or for 
 storing until needed. The molds travel back toward the spout, 
 underneath the machine and hollow side down. At the point C 
 they are sprayed with whitewash, the water of which is quickly 
 dried ofl^ by the heat of the mold, leaving a coating of lime inside 
 to which the melted iron will not stick. This mechanical casting 
 
36 
 
 THE METALLURGY OF IRON AND STEEL 
 
 is a great improvement over the former method of cooling iron 
 in front of the blast furnace, because of the severity of the work 
 which the former method involved and which, in hot weather, 
 was well-nigh intolerable to human beings. It also gives pigs 
 which are cleaner, i.e., freer from adhering sand. This silicious 
 sand is objectionable, especially in the basic open-hearth furnace. 
 
 FIG. 18.— UEHLING PIG-CASTING MACHINE. 
 
 Sand-casting, — This method is still used at some furnaces, 
 because of the capital needed to install machines and their high 
 cost for repairs. Moreover, foundrymen often prefer the sand- 
 cast pig because they are able to tell by the appearance of its 
 fracture what grade of castings it will make, which they cannot 
 well do with iron cast in metal moulds (see pages 318, 336). In 
 
 FIG. 19.— HEYL AND PATTERSON PIG-CASTING MACHINE. 
 
 the sand method, the cast house extends in front of the furnace 
 and its floor is composed of silica sand, in which the molds or 
 impressions to receive the liquid iron are made. The main runner 
 extends from the taphole down the middle of the floor, and the 
 space on either side of it is used alternately for alternate castings. 
 The plan of the arrangement is shown in Fig. 20. 
 
THE MANUFACTURE OF PIG IRON 
 
 37 
 
 After cooling the iron, the pigs are broken away from the sows, 
 which are also broken into pieces with a sledge, and then all is 
 carried over and thrown into a railroad car. In making "basic 
 iron" — i.e., iron for the basic open-hearth steel process — the 
 moulds for the sows and pigs are permanently made of metal, so 
 that the iron will not carry acid sand into the basic hearth. 
 
 Irregvlarities in Blast-furnace Working. — Although the man- 
 agement and control of the operation is in general as I have 
 described it, the blast furnace is 
 by no means a perfect machine,/ 
 and great difficulties arise in the\ 
 working of the furnace and in 
 maintaining a uniform grade of 
 product. The chief of these 
 difficulties result from localized 
 chilling of the semi-molten 
 charge. This is most liable to 
 happen in the upper part of the 
 smelting zone, where a lump of 
 pasty material may attach itself 
 to the walls of the furnace. This 
 has the effect of hindering the 
 descent of that part of the charge 
 above it and of deflecting the hot 
 gases to other parts of the fur- 
 nace. The result of the first 
 action is to disarrange the order 
 and evenness with which origi- 
 nally horizontal rings of stock 
 come down into the hearth. The 
 obstruction is also liable to re- 
 ceive chilled materials from 
 
 above and to build itself out toward the center. When the 
 furnace is working badly, these scaffolds may occur at two or 
 more places at the same time and cantilever out toward the 
 middle. This will cause a "hanging" of the charge, and may 
 become so bad as to cause a complete arch over the smelting 
 zone, through which it is impossible to drive the blast. Some- 
 times the scaffold may be broken down by suddenly cutting off 
 the blast pressure and allowing the full weight of material in the 
 furnace to come upon the obstruction; but sometimes it is neces- 
 
 T ^^^=^=;:=>^ 
 
 " 'sklBUaBT ^^ Eunnej 
 
 ^ JlJUULII 
 
 ^iffliiiffl 
 
 nniiimi 
 
 mmm 
 
 nnnnnnjinnnnnnnn 
 
 'UUUUUuUUUUUiUUU 
 
 nnnnf'lnnnnrr""" 
 
 ^l.... 
 
 FIG. 20— SAND-CASTING BED. 
 
38 
 
 THE METALLURGY OF IRON AND STEEL 
 
 sary to cut a hole inthe wall of the furnace and melt it out with 
 a blow-pipe burning oil or gas, or with some other form of heat. 
 The "scaffolding" of a furnace and hanging of the charge is 
 more liable to happen when large percentages of the earthy 
 
 FIG. 21.— PIG BEDS. 
 
THE MANUFACTURE OF PIG IRON 39 
 
 Mesabi ores are used, and in this type of practice localized hang- 
 ing and slips are not infrequent. When the slip is extensive in 
 character and a large amount of material is suddenly precipitated 
 into the hearth, the upward rush of gases resembles an explosion 
 inside the furnace and may do damage to the charging apparatus 
 and throw a part of the stock out of the top of the furnace. 
 Some furnaces are provided with explosion doors, which fly open 
 under pressure and relieve the strain; while the practice in other 
 instances is to fasten everything down lis tight as possible and 
 prevent the rapid escape of the gases. 
 
 There is also a large amount of hanging due to the action of the 
 blast in tending to drive the stock before it up into the stack of 
 the furnace and thus compress it. This action is more liable to 
 take place with fine ores. 
 
 Cooling of the charge also results in some cases in the freezing 
 of material over the mouths of the tuyeres. The solid layer may 
 sometimes be broken away with a bar, and the blow thus allowed 
 to continue until more heat can be brought down into the hearth. 
 Sometimes it is necessary to melt out the frozen material with a 
 blow-pipe, and in extreme cases it may even be necessary to 
 break through it with explosives. 
 
 Another difficulty sometimes met with is the freezing up of the 
 metal in the lower part of the hearth, so that it is impossible to 
 open the tap-hole. Then a new tap-hole must be mad^ by boring 
 through the front of the furnace at a higher level, from which the 
 iron is drained, and then the heat gradually worked down until 
 the whole hearth is melted out and normal conditions reestab- 
 lished. The bad work of a furnace is often cumulative in its 
 effects, because irregularities in the smelting zone have an effect 
 upon the top gases, which, in turn, derange the work of the 
 stoves and hence impair the hot blast. 
 
 These irregularities in the smelting have a disturbing effect 
 upon the character of the iron made, and the changes sometimes 
 come suddenly and without warning. For instance, a sudden pre- 
 cipitation of cold material into the hearth will chill the smelting 
 zone and cause the silicon in the iron to be low and the sulphur 
 high. The same effect will be produced by the leakage of several 
 gallons of water into the hearth through the burning out of a 
 tuyere or the cooling-ring of one of the tuyeres.^ 
 
 1 In my early days at the blast furnace I was once informed by the assistant manager 
 that, on one occasion, he tapped several tons of water from the tap-hole with the iron. 
 
40 THE METALLURGY OF IRON AND STEEL 
 
 Dimensions of Blast Furnace, — The size of a modern blast 
 furnace is limited by the conditions of its work; the hearth may 
 not be much more than 15 ft. in diameter, else the blast from the 
 tuyeres will not be distributed evenly to the center; the batter of 
 the bosh walls cannot be much more nor less than a certain 
 amount, because they must give support to the charge above 
 them, and yet allow the solid coke to slip down; the height of 
 bosh is limited, becase its top must be practically the same as the 
 top of the smelting zone — that is, no solid material except coke 
 should descend into the bosh. These conditions therefore limit 
 the diameter of the top of the bosh to not much more than 22 ft. 
 From the bosh the stack walls must decrease in diameter upward 
 in order that the descending charge, which swells in the reactions 
 that take place from the throat downward, shall not become 
 wedged in the stack; as the throat must have a sufficiently large 
 diameter to properly charge the materials, this limits the height 
 of the stack. Modern furnaces are therefore usually built about 
 90 ft. in height, and exceeding that limit has resulted in some 
 cases in a decrease rather than an increase of fuel economy. 
 Moreover, increasing the height of the furnace increases the blast 
 pressure necessary much faster than the weight of the charge 
 column so the height is soon reached at which the furnace can 
 work only irregularly if at all. 
 
 Calculating a Blast-furnace Charge 
 
 This subject is of prime importance to young metallurgists, 
 because the ability to calculate a charge is sometimes a cause of 
 advancement, and the knowledge of the way to do so is not always 
 obtainable from one's superior. 
 
 Assumptions. — Let us assume that we desire to produce a slag 
 containing 55 per cent, lime, 15 per cent, alumina, and 30 per 
 cent, silica, these proportions being determined by the experi- 
 ence of the manager, and that the materials from which the 
 charge is. to be made analyze according to Table III. Assume 
 furthermore that the coke ash is equal to 10 per cent, of the coke, 
 and that the iron we are going to make will contain about 1 per 
 cent, silicon. 
 
 Whether he waa himself deceived or whether he was merely trying to test me, I have never 
 been able to decide; but the fact is worthy of mention in an elemental^ treatise to illustrate 
 the character of the tales to which even the educated men around a plant will treat a novice. 
 
THE MANUFACTURE OF PIG IRON 
 TABLE III 
 
 41 
 
 Material 
 
 Per cent. 
 CaO 
 
 Per cent. 
 MgO 
 
 Per cent. 
 Al,03 
 
 Per cent. 
 SiO^ 
 
 Per cent. 
 Fe^Oa 
 
 Per cent. 
 Fe 
 
 Ore A 
 
 5 
 
 2 
 
 20 
 
 46 
 
 4 
 4 
 
 2 
 12 
 18 
 
 2 
 
 n 
 
 16 
 
 50 
 
 4 
 
 
 60 
 
 Ore B 
 
 50 
 
 Coke ash 
 
 Limestone 
 
 10 
 2 
 
 Silicon in the Iron, — This last assumption necessitates our 
 allowing a corresponding amount of silica, because the silica 
 reduced and absorbed by the iron will not be available for slag- 
 making purposes. One per cent, of silicon is roughly equal to 2 
 per cent, of silica; we may therefore make the requisite allowance 
 by subtracting from the silica in each material an amount equiv- 
 alent to 2 per cent, of its iron content. Thus we begin to make 
 up Table IV. . 
 
 Magnesia. — In considering slags, magnesia is classified under 
 the head of lime, thus obtaining column 2 in Table IV. 
 
 TABLE IV 
 
 
 
 Material 
 
 Per cent. 
 CaO 
 
 Per cent. 
 Al,03 
 
 Per cent. 
 SiO^ 
 
 Ore A 
 
 9 
 
 2 
 
 20 
 
 50 
 
 2 
 12 
 18 
 
 2 
 
 10 
 
 Ore B 
 
 15 
 
 Coke ash 
 
 50 
 
 Limestone 
 
 4 
 
 
 
 Self -fluxing of Materials. — It is evident that in so far as each 
 of the materials in Table IV contains all the components of the 
 slag, they will partially flux themselves. For example, the 2 
 per cent, of alumina in ore A will theoretically combine with 4 
 per cent, of the silica (2 per cent. X 30/15 =4 per cent.) and 
 7 per cent, of the lime (2 per cent. X 55/15 = 7.3 per cent.) to 
 make a slag of the desired proportions, leaving unfluxed per- 
 centages as per the first line of Table V. In the same manner we 
 may use up all of the lime in ore B by uniting it with weights of 
 
42 
 
 THE METALLURGY OF IRON AND STEEL 
 
 alumina and silica in proportion to the percentages of these com- 
 ponents in the slag. Similar simplifications in the analyses of 
 coke ash and limestone may then be calculated, and Table V will 
 be completed. 
 
 TABLE V 
 
 Material 
 
 Per cent. 
 CaO 
 
 Per cent. 
 Al^O^ 
 
 Per cent. 
 SiO, 
 
 Ore A 
 
 2 
 
 
 6 
 
 Ore B 
 
 11.5 
 13.0 
 
 11 
 
 Coke ash 
 
 
 39 
 
 Limestone 
 
 43 
 
 
 
 
 
 Weight of Charge. — Let us assume that we are going to make 
 one *ton of pig iron for every ton of coke used in the charge, and 
 that the coke will be put in in charges weighing 11,000 lb. each. 
 This weight includes about 10 per cent, of moisture, dust, etc.,* 
 so we calculate with it as if it weighed only 10,000 lb. Now let us 
 determine how much ore will be put in each charge: The ores 
 average 55 per cent, of iron; therefore, gspgrVent ^ 18,000 lb., the 
 amount of ore that must be in each charge, according to the 
 assumption of this paragraph. 
 
 Adjusting the Alumina and Silica. — Next adjust the different 
 materials so that the weight of alijmina shall be 15/30 of the 
 weight of silica. In the first rough approximation of this we may 
 ineglect the coke ash, because the weight of this ash is so small in 
 relation to the other materials. Therefore only the two ores 
 need be apportioned, and we quickly find by trying a few mix- 
 tures at random that 60 per cent, of ore A mixed with 40 per 
 cent, of ore B will give the desired relation:^ 60 per cent. X 6 -MO 
 Xper cent. 14 = 920 parts of silica; 60 per cent. X 0+40 per 
 cent. X 11.5=460 parts alumina; 460/920 = 15/30. Now draw 
 Table VI, and enter 10,800 lb. of ore A( = 60 per cent, of 18,- 
 000), 7200 lb. of ore B, 1000 lb. of coke ash ( = 10 per cent, of 
 10,000), and the percentages from Table V. All the weights in 
 
 iThis assumes that the breeze (or "braize") is removed after weighing the coke. If 
 it is removed before, allow only 3 per cent, excess, which makes coke charges =10,300 lb. 
 
 ^Try first 50 per cent, of each, and we see that there is too much alumina; therefore 
 try less than 60 per cent, of the ore having the most alumina, and correspondingly more 
 of the other, and we have it. 
 
THE MANUFACTURE OF PIG IRON 
 
 43 
 
 this table m9,y then be filled in except those of the limestone and 
 total CaO; ' 
 
 To obtain the total number of pounds of lime: 
 
 AI2O3: 958 X|| = 3513 1b. 
 SiO^: 2046 X|| = 3751 lb. 
 Average of 3515 and 3751 is 3632. 
 
 TABLE VI 
 
 Material 
 
 CaO 
 
 AI.O, 
 
 SiO^ 
 
 
 Weight 
 
 Per 
 
 Cent. 
 
 Xb. 
 
 Per 
 Cent. 
 
 Lb. 
 
 Per 
 
 Cent. 
 
 Lb. 
 
 Ore A 
 
 10,800 
 7,200 
 1,000 
 
 c 
 
 7,940 
 
 2 
 
 216 
 
 
 
 6 
 14 
 39 
 
 648 
 
 Ore B 
 
 11.5 
 13.0 
 
 828 
 130 
 
 1008 
 
 Coke a^h 
 
 
 
 390 
 
 Limestone 
 
 43 
 
 B 
 3416 
 
 A 
 
 3632 
 
 
 Total lb... . . . 
 
 
 
 958 
 
 
 2046 
 
 
 
 
 
 Adjusting for Lime. — It is now only necessary to determine the 
 amount of total lime that shall bear the correct relation to the 
 alumina and silica calculated. This we do by means of the 
 method shown in the figures above Table VI. We enter this in 
 the square "A" of Table VI. The figures at the square "B'' 
 are then obtained (3632 — 216 = 3416), and thence the weight of 
 limestone to be used — (3416-^43 per cent. =7940).^ 
 
 Checking the Calculations. — We now check up all the calcula- 
 tions by making up Table VII, in which we go back to the original 
 percentages found by chemical analysis and given in Table III. 
 In making up this final table, however, we use our experience in 
 making slag calculations and estimate Slight changes: For ex- 
 
 ^ In practice this is often much simplified by taking the AI2O3 simply as fixed, within wide 
 Umits, and figuring merely the ratio of CaO and MgO to ^102; disregard AI2O3, unless it is 
 abnormally high, and therefore bad for physical reasons. If a constant limestone is used 
 all such calculations are much simplified by figuring the net CaO and MgO in the stone after 
 the SiOz in the atone is satisfied. 
 
 If a given ratio of CaO+MgO to SiOa be desired it is only necessary to add up the 
 Si02 in ore and fuel, multiply by the desired ratio and divide by the efficiency of the 
 stone. The net Si02 in the ores should be taken, that is, the Si02 fluxed by the bases 
 in the ore should be deducted. 
 
44 
 
 THE METALLURGY OF IRON AND STEEL 
 
 ample, Table VI shows us that the alumina comes a little low in 
 relation to silica; therefore we increase ore B, say, by 400 lb. and 
 decrease ore A correspondingly. But ore A is high in lime; 
 therefore we use a little more limestone to offset this reduction. 
 
 TABLE VII 
 
 Material 
 
 CaO + MgO 
 
 Al,03 
 
 SiO^ 
 
 Fe 
 
 
 Weight 
 
 Per 
 cent. 
 
 Lb. 
 
 Per 
 
 cent. 
 
 Lb. 
 
 Per 
 cent. 
 
 Lb. 
 
 Per 
 cent. 
 
 Lb. 
 
 Ore A 
 
 Ore B 
 
 Coke ash. . . 
 Limestone. . 
 
 10400 
 7600 
 1000 
 8200 
 
 9 
 
 2 
 
 20 
 
 50 
 
 936 
 
 152 
 
 200 
 
 4100 
 
 2 
 12 
 
 18 
 
 2 
 
 208 
 912 
 180 
 164 
 
 11 
 
 16 
 
 50 
 
 4 
 
 1144 
 
 1216 
 
 500 
 
 328 
 
 60 
 
 50 
 
 10 
 
 2 
 
 6240 
 
 3800 
 
 100 
 
 164 
 
 Total weights 
 
 5388 
 
 = 54.8 per 
 cent. 
 
 1464 
 
 = 14.9 per 
 cent. 
 
 ....3188 10,304 
 
 -206(=2%10,X304) 
 
 2982=^30.3 percent. 
 
 These figures are much closer to those desired than the limit of 
 accuracy in furnace operation. The chief difference is that we 
 are making a little more iron with 10,000 lb. of coke than we 
 intended. If any change seems necessary it is then well to reduce 
 the weight of ore A to 10,000, leaving everything else the same. 
 This will lighten the burden and bring the calculated lime, 
 alumina, and silica even closer to the desired figures. 
 
 Phosphorus and Manganese. — No account of the phosphorus 
 has been taken in the calculation above. This is necessary some- 
 times. For example, if ore A happened to be very high in phos- 
 phorus we could not use so large a proportion of it. It would 
 then be necessary either to secure another ore low in both phos- 
 phorus and alumina, or else to make a slag with less alumina. 
 The same line of reasoning applies to manganese, but to less 
 extent because the manganese in the iron can be controlled by the 
 work of the furnace to a considerable extent, and in no event 
 does more than two-thirds of that in the charge go into the iron. 
 
Ill, 
 
 THE PURIFICATION OF PIG IRON 
 
 The large amount of carbon in pig iron makes it both weak and 
 brittle, so that it is unfit for most engineering purposes. It is 
 used for castings that are to be subjected only to compression, or 
 transverse or very slight tensile strains, as, for example, support- 
 
 I Annealed I 
 
 Malleable 
 Cast Iron 
 
 Skeleton of American Iron and Steel Manufacture 
 1906 
 
 50,032,279 Tons of Iron Ore 
 
 
 
 
 
 
 25,307,192 Tons 
 
 f Pig Iron 
 
 
 
 
 
 
 
 3 
 
 t 
 
 
 SOjji 
 
 
 52 
 
 % 
 
 
 20|5S ' 
 
 r 
 
 1 
 
 56 
 
 
 3 
 
 % 
 
 Remelted 
 
 
 Bemelted 
 
 
 62 
 Bessemer 
 Converter! 
 
 
 465 Basic 
 Open- 
 Hearth 
 Furnaces 
 
 
 195 Acid 
 Open- 
 Hearth 
 Furnaces 
 
 
 8,000 
 Puddling 
 Furnaces 
 
 Cast 
 
 lite 
 IroE 
 
 
 i 
 
 Gray 
 
 Cast Iron 
 
 B 
 
 essem 
 (12,2' 
 
 erSt 
 5,250 
 
 361 1 
 
 I 
 
 1 
 
 ^asic Open 
 learth Stee 
 (9,649,400; 
 
 i ] 
 
 Acid 
 ieart 
 (1,82 
 
 Open 
 iSte 
 L,613) 
 
 3l 
 
 ■ 
 
 
 Wrought Iron 
 
 • 6|r 
 
 I fiemelteT] 
 
 Crucible Steel 
 (118,000) 
 
 "9^ 
 
 Used 
 as Such 
 
 All Tons are 2^240 lbs. each 
 
 FIG. 25. 
 
 ing columns, engine bed-plates, railroad car wheels, water mains, 
 etc., but the relatively increasing amount of steel used shows the 
 preference of engineers for the stronger and more ductile material. 
 To-day three-fourths of the pig iron made in the United States is 
 subsequently purified by either the Bessemer, open-hearth, or 
 
 45 
 
46 
 
 THE METALLURGY OF, IRON AND STEEL 
 
 puddling process. Each of these will reduce the carbon to any 
 desired point, while the silicon and manganese are eliminated as a 
 necessary accompaniment of the reactions — ^indeed, we might 
 almost say as a condition precedent to carbon reduction. Phos- 
 phorus and sulphur are reduced by the puddling process, 'and by 
 a special form of open-hearth process known as "the basic open- 
 hearth process/'^ The complete scheme of American iron and 
 steel manufacture is given in Fig. 25. 
 
 Explanation of Fig. 25. — ^Practically all the iron ore mined is 
 smelted in blast furnaces to produce pig iron. About 25 per cent, 
 of the pig iron is used in the form of castings without purification, 
 and 75 per cent, is purified and converted into another form. In 
 all cases of purification the impurities are removed by oxidizing 
 them, and we must again emphasize the rule that unoxidized 
 elements dissolve in the metal, while those in the oxidized con- 
 dition pass into the slag, or, if there is no slag, form a slag for and 
 of themselves. In considering the Bessemer, open-hearth, and 
 puddling processes then, we have to do with oxidizing conditions, 
 whereas the opposite was the case in the blast furnace. The 
 oxidation is effected by means of the oxygen of the air or that of 
 iron ore, FegOg, or its equivalent, or of both air and oxide of iron. 
 
 There is not an exact relation between the amounts of pig iron 
 used for the different purposes and the amounts of the resulting 
 materials. In 1910 the following production was made: 
 
 PRODUCTION FOR 1910 (Long Tons) 
 
 t 
 
 Pig Iron Used 
 
 Made 
 
 Malleable cast iron^ 
 
 800,000* 
 5,500,000* 
 10,614,000 
 9,084,520 
 500,000 
 800,000* 
 
 1,000,000* 
 6,750,000* 
 9,412,772 
 15,392,357 
 1,212,152 
 2,000,000* 
 177,638 
 55,325 
 
 Gray cast iron 
 
 Bessemer steel 
 
 Basic open-hearth steel .... 
 
 Acid open-hearth steel 
 
 Wrought iron 
 
 Crucible steel 
 
 Electric steel 
 
 
 
 
 * Estimated. 
 
 1 The basic Bessemer process is not in operation in America. 
 
 ^ It is true that the annealing process for malleable cast iron purifies the outer layers of 
 the castings from carbon, and, if the castings are very thin, this purification may extend 
 to the center; but this is not primarily a purification process and will be treated at length 
 in another section . 
 
THE PURIFICATION OF PIG IRON 
 
 47 
 
 The reason for this discrepancy is found in the scrap iron or 
 steel mixed with the pig iron in the manufacture of gray-iron 
 castings and open-hearth steel. Perhaps an average of 25 per 
 cent, of old scrap will be mixed with 75 per cent, of new pig iron 
 for making iron castings, and 50 per cent, or so of steel scrap will 
 be mixed with 50 per cent, or so of pig iron in the basic open- 
 hearth process, while wrought iron is often made by the piling and 
 rerolling of old wrought-iron scrap. 
 
 FIG. 26.— SECTION THROUGH BESSEMER CONVERTER WHILE BLOWING. 
 
 Bessemer Process. — In the Bessemer process, perhaps 15 tons of 
 melted pig irOn is poured into a hollow pear-shaped converter 
 lined with silicious material. Through the molten material is 
 then forced 30,000 cu. ft. of cold air per minute. In about four 
 minutes the silicon and manganese are all oxidized by the oxygen 
 of the air and have formed a slag. The carbon then begins to 
 
48 THE METALLURGY OF IRON AND STEEL 
 
 oxidize to carbon monoxide, CO, and this boils up through the 
 metal and pours out of the mouth of the vessel in a long brilliant 
 flame. After another six minutes the flame shortens or " drops "; 
 the operator knows that the carbon has been eliminated to the 
 lowest practicable limit (say 0.04 per cent.) and the o|)eration is 
 stopped. So great has been the heat evolved by the oxidation 
 of the impurities that the temperature is now higher than it was 
 at the start, and we have a white-hot liquid mass 6f relatively 
 pure metal. To this is added a carefully calculated amount of 
 carbon to produce the desired degree of strength or hardness, or 
 
 FIG. 27.— BLOWHOLES OR GAS-BUBBLES IN STEEL. 
 
 both; also about 1.0 per cent, of 'manganese and 0.15 per cent, of 
 silicon.^ The manganese is added to remove from the bath the 
 oxygen with which it has become charged during the operation 
 and which would render the steel unfit for use. The silicon is added 
 to get rid of the gases which are contained in the bath. After 
 adding these materials, or '* recarburizing," as it is called, the metal 
 is poured into ingots, which are allowed to solidify and then 
 rolled, while hot, into the desired size and form. The character- 
 istics of the Bessemer process are: (a) Great rapidity of purifica- 
 tion (say ten minutes per "heat") ; (b) no extraneous fuel is used; 
 and (c) the metal is not melted in the furnace where the purifica- 
 tion takes place. 
 
 1 In the case of making rail steel. 
 
THE PURIFICATION OF PIG IRON 
 
 49 
 
 Acid Open-hearth Process. — The acid open-hearth furnace is 
 heated by burning within it gas and air, each of which has been 
 highly preheated before it enters the combustion chamber. A 
 section of the furnace 'is shown in Fig. 28. The metal lies in a 
 shallow pool on the long hearth, which is lined with silicious 
 
 Tz^wa^^^ 
 
 FIG. 28.— DIAGRAM OF REGENERATIVE OPEN-HEARTH FURNACE. 
 The four regenerative chambers below this furnace are filled with checkerwork of brick 
 around which the gas and air may pass. Before the furnace is started these bricks are 
 heated up by means of wood fires. The gas enters the furnace through the inner regenera- 
 tive chamber on one side and the air enters through the outer one on the same side. They 
 meet and unite, passing through the furnace and thence passing to the chimney through 
 the two regenerative chambers at the opposite end. In this way the brickwork in the out- 
 going chambers is heated still hotter by the waste heat of the furnace. The current of gas, 
 air, and products of combustion is changed every twenty minutes, whereby all four regenera- 
 tors are always kept hot. The gas and air enter the furnace in a highly preheated condi- 
 tion and thus give a greater temperature of combustion, while the products of combustion 
 go out to the chimney at a relatively low heat and thus fuel economy is secured . 
 
 material, and is heated by radiation from the intense flame. 
 The impurities are oxidized by the slag lying on top of the metal. 
 This action is so slow, however, that the carbon in the pig iron 
 takes a long time for combustion. The operation is therefore 
 
 4 
 
50 THE METALLURGY OF IRON AND STEEL 
 
 hastened in two ways: (a) iron ore is added to the bath to pro- 
 duce the reaction. 
 
 Fe^Oa + 3C = SCO tH 2Fe, 
 
 and (b) the carbon is diluted by adding with the furnace charge 
 large proportions of steel scrap^ often as .much as 75 or 90 per cent. 
 It takes about 6 to 10 hours to purify a charge. The manganese 
 and silicon go into the slag first; then the carbon boils off as a gas. 
 When this has proceeded to the desired point we recarburize and 
 cast the metal into ingots. The characteristics of the acid open'- 
 hearth process are: (a) A long time occupied in purification; (6) 
 large charges are treated in the furnace (the modern practice is 
 usually 15 to 100 tons to a furnace) ; (c) at least a part of the charge 
 is melted in the purification furnace; and (d) the furnace is heated 
 with preheated gas and air. 
 
 Basic Open-hearth Process. — ^The basic open-hearth process is 
 similar to the acid open-hearth process, with the difference that 
 we add to the bath a sufficient amount of lime to form a very 
 basic slag. This slag will dissolve all the phosphorus that is 
 oxidized, which an acid slag will not do.^ The characteristics of 
 the basic open-hearth process are the same as those of the acid 
 open-hearth, with the addition of: (e) Lime is added to produce a 
 basic slag; (/) the hearth is lined with basic, instead of silicious, 
 material, in order that it may not be eaten away by this slag; and 
 (g) impure iron and scrap may be used, because phosphorus and, 
 to a limited extent, sulphur can be removed in the operation. 
 
 Puddling Process. — Almost all the wrought iron today is made 
 by the puddling process, invented by Henry Cort about 1780, 
 with certain valuable improvements made by Joseph Hall fifty 
 years later. In this process the pig iron is melted in a reverbera- 
 tory furnace the hearth of which is lined with oxide of iron. 
 During the melting there is an elimination of silicon and manga- 
 nese and the formation of a slag which automatically adjusts 
 itself to a very high content of iron oxide by absorption from the 
 lining. After melting, the heat is reduced and a reaction set up 
 between the iron oxide of the slag and the remaining silicon, 
 manganese, carbon, phosphorus, and sulphur of the bath, whereby 
 the impurities are oxidized and all removed to a greater or less 
 extent. The slag, because of its basicity (by iron oxide), will 
 
 1 We can oxidize the phosphorus in any of these processes, but in the acid Bessemer and 
 the acid open-hearth furnaces the highly silicious slag rejects the phosphorus, and 'it is 
 immediately deoxidized again and returns to the iron. 
 
THE PURIFICATION OF PIG IRON 
 
 51 
 
 retain all the phosphorus- oxidized, and therefore the greater part 
 of this element may be removed. The oxidation of all the im- 
 purities is produced chiefly by the iron oxide in the slag and the 
 lining of the furnace, although it is probable that excess oxygen 
 in ^he furnace gases assists the slag by acting as a carrier of 
 oxygen from it to the impurities. 
 
 FIG. 29 —PUDDLING FURNACE. 
 
 The purification finally reaches that stage at which the utmost 
 heat of the furnace is not sufficient to keep the charge molten, 
 because iron, like almost every other metal, melts at a higher 
 temperature the purer it is. The metal therefore "comes to 
 nature," as it is called, that is to say, it assumes a pasty state. 
 
 FIG, 30.— METHOD OF PILING MUCK BAR. 
 
 The iron is rolled up into several balls, weighing 125 to 180 lb. 
 apiece, which are removed from the furnace, dripping with slag, 
 and carried over to an apparatus, where they are squeezed into a 
 much smaller size and a large amount of sleg ejected from them. 
 The squeezed ball is then rolled between grooved rolls to a bar, 
 
52 THE METALLURGY OF IRON AND STEEL 
 
 whereby the slag is still further reduced, so that the bar contains 
 at the end usually about 1 or 2 per cent. This puddled bar, or 
 "muck bar," is cut into strips and piled up, as shown in Fig. 30, 
 into a bundle of bars which are bound together by wire, raised to 
 a welding heat, and again rolled into a smaller size. This rolled 
 material is then known as "merchant bar," and all wrought iron, 
 except that which is to be used for manufacture into crucible 
 steel, is treated in this way before «ale. The effect of the further 
 rolling is to eject more slag, and also to make a cross network of 
 fibers, instead of a line of fibers all running in the same direction, 
 i.e., lengthwise of the bar. The fibers are produced by the action 
 in rolling of drawing out the slag into strings, long fibers of metal 
 also being produced, each of which is surrounded by an envelope 
 of slag. 
 
 Miscellaneous Purification Processes. 
 
 Bell-Krupp Process. — The late Sir I. Lowthian Bell devised a 
 process in which liquid pig iron is violently stirred up with iron 
 oxide, producing a slag which carries away in the course of from 
 7 to 10 minutes more than 90 per cent, of the silicon and phos- 
 phorus in the metal. As soon as carbon begins to burn the proc- 
 ess is stopped, and therefore there is almost no reduction in 
 this element. The operation is conducted on the revolving 
 hearth of a mechanical puddling furnace, into which the melted 
 iron is poured while the hearth is rotating at about 11 revolutions 
 per minute. The temperature is lower than that of the open- 
 hearth process, in order that the elimination of phosphorus may 
 be rapid. (See p. 59 and 146.) 
 
 The purified metal is used to some extent in the manufacture of 
 crucible steel. During recent years, when the low phosphorus 
 ores of America have become more scarce and the price of Besse- 
 mer pig iron is consequently increased, the metal has been bought 
 to a limited extent by foundrymen using the acid open-hearth 
 process or the baby Bessemer process for the production of steel 
 castings. 
 
 Finery Fire, — This furnace is known under various names, 
 such as "refinery hearth," "running-out fire," "finery fire," etc. It 
 consists of a shallow, rectangular hearth, surrounded on the sides 
 by water-cooled, hollow blocks of iron about 3 to 3 1 /2 ft. long by 2 
 ft. wide and 24 to 30 in. deep. In and above this hearth is built a 
 
THE PURIFICATION OF PIG IRON 53 
 
 ■fire of coke upon which is placed 500 or 600 lb. of pig iron. The 
 coke is burned by a blast at 2 to 3 lb. pressure from 2 to 3 tuyeres 
 on each side, and the pig iron gradually melts and sinks below it. 
 When this takes place, more coke and pig iron is placed upon the 
 top, and the operation repeated. A bath of pig iron forms in the 
 hearth, and upon this the blast from the tuyeres impinges. This 
 oxidizes the silicon in the metal, and also a large amount of iron, 
 phosphorus, and sulphur. A slag high in iron oxide, and there- 
 fore very basic, is formed. As the temperature is low, phosphorus 
 is eliminated without burning much carbon, and the result is the 
 production of a purified iron still high in the latter element. 
 
 It takes about two hours to perform this purification, and then 
 the metal is tapped out from the tap-hole in the front. It usuaUy 
 runs into a long, shallow trough, whence the name " running-out 
 fire"; but sometimes the refined metal is not allowed to cool, but 
 is run directly into the furnace in which the purification is to be 
 completed. The consumption of coke in this first furnace is 
 about one-eighth of the metal produced, and the loss from 5 to 20 
 per cent., depending upon the purity of the iron treated. 
 
 The running-out fire is frequently used in connection with the 
 charcoal finery, known as the "knobbling fire," to produce knob- 
 bled charcoal-iron, which is employed especially for boiler tubes 
 and to a less extent for boiler plate, wire, rivets, etc. Running- 
 out fires for this purpose are, however, often known as " melting 
 £neries," because in them the pig iron is melted before it goes 
 to the knobbling fire. 
 
 Melting Finery, — In the melting finery there are usually two 
 tuyeres in the back, and the melted metal, after an operation 
 similar to that described above, is run directly into two charcoal 
 fineries. ' ' 
 
 Charcoal Fineries, — The charcoal fineries, or ''knobbling fires," 
 are similar in general to Fig. 31, but have only one tuyere, 
 situated in the back. They take a charge of 250 lb. apiece. 
 During the transfer from the melting finery the slag is separated 
 from the metal as well as possible, but some gets into the char- 
 coal fineries. It is allowed to solidify and then is removed. 
 Upon the metal is now charged some damp charcoal. The cold 
 blast is turned on and the metal constantly agitated and raised 
 up from the bottom so as to bring it in contact with the blast. 
 Charcoal is added from time to time and is kept damp to avoid 
 loss, and the slag is removed at intervals, but there must always 
 
54 
 
 THE METALLURGY OF IRON AND STEEL 
 
 be a layer of slag between the metal and charcoal. As the 
 metal comes to nature, it is pressed together with the pointed 
 bar, like a crowbar, which is used for the agitation and raising. 
 At the end of about an hour and a half, the ball is withdrawn 
 and hammered. The cinder from the knobbling fire is usually 
 charged into the melting finery. 
 
 The great advantages of the knobbled iron as compared to 
 puddled iron are its softness and relative freedom from slag. 
 
 ^r- 
 
 Blast 
 
 r<S5iigs^"_-_~^ 
 
 Pipe 
 
 :ji 
 
 FIG. 31.— MELTING FINERY. 
 
 Also it is often lower in phosphorus and sulphur. Tubes made 
 of this material may be flanged out very extensively without 
 showing any cracks, and rivets will flow easily when hammered 
 cold. 
 
 The Lancashire Process. — In the Lancashire process, which 
 some believe is a decendant in Sweden of the purification in the 
 
THE PURIFICATION OF PIG IRON 
 
 55 
 
 two fires last described, the same operations are performed in one 
 hearth, which is made of iron plates, sometimes cooled with water, 
 with a tap-hole in the front. The process consists of three stages; 
 
 1. The melting down, which is somewhat similar to the opera- 
 tion in the running-out fire or melting finery. 
 
 2. A purification period, during which the metal is nearly puri- 
 fied and comes to nature. 
 
 3. A remelting above the tuyere for further purification. 
 
 In America, pig iron heated red-hot in the chamber H-C (Fig. 
 32), during the working of a previous charge, is placed between 
 two layers of charcoal and a little above the level of the tuyere. 
 It is quickly melted, the liquid drops being forced to trickle down 
 through the blast and thus be exposed to strongly oxidizing 
 conditions. 
 
 FIG. 32.— AMERICAN LANCASHIRE FURNACE. Tu=Tuyere. From Howe, 
 "The Metallurgy of Steel." 
 
 The second period begins at the end of about 15 minutes, when 
 the melted met^l is all collected in the bottom. It becomes pasty 
 in contact with the cold hearth, and is raised by a pointed bar and 
 charcoal allowed to fall under it. Some slag at the same time is 
 tapped off and some is mixed with the pasty lump, which pro- 
 duces a reaction between the two that assists in the purification. 
 Toward the end of this period, which lasts 20 or 25 miniltes, car- 
 bonic oxide comes off very rapidly, and when the metal becomes 
 so stiff that great pressure is needed to raise it, and the slag has 
 become thinner and whiter, the third period begins. 
 
 The metal is now broken into pieces and raised to its original 
 position, the action of the first period being substantially repeated. 
 During this period the workman is careful not to touch the mass 
 collecting in the bottom of the hearth lest he mix slag with it. 
 By the time about two-thirds of the metal is melted, some rich 
 
56 
 
 THE METALLURGY OF IRON AND STEEL 
 
 iron oxide slag is aclded in order to keep enough in the bottom of 
 the hearth to protect the metal from being carburized by the 
 charcoal. When all the metal has melted and dropped down in 
 front of the tuyere, the pasty ball is pried out of the hearth and 
 hammered. The third period takes about 25 to 30 minutes. 
 
 Walloon Process, — In the Swedish Lancashire or Walloon proc- 
 ess, the long pigs of metal are fed slowly down into the fire, so 
 that it is not necessary to constantly pry them up with bars. 
 The charges are smaller and the product is more liable to be 
 
 FIG. 33. — From Howe, "The Metallurgy of Steel." 
 
 heterogeneous, because the first melted metal is decarburized 
 more than the last, and this heterogeneousness is not removed 
 by the remelting. (See Fig. 33.) 
 
 The " Walloon process " is used in Sweden for making bar iron 
 from the very pure pig iron reduced from Dannemora iron 
 ore. This bar iron is used in Sheffield, England, for conversion 
 into blister-steel, and some of the steelmakers pay a large price 
 for it, in the belief that it has a certain intangible "body" not 
 contained in wrought iron from any other process, and which 
 makes a superior quality of steel. It is probable, however, that 
 this body is wholly imaginary. 
 
IV 
 
 THE MANUFACTURE OF WROUGHT IRON AND 
 CRUCIBLE STEEL 
 
 The Manufacture of Wrought Iron 
 
 Pig Iron Used. — The pig iron employed is of the grade known 
 as " forge iron " or " mill iron." In the United States we prefer to 
 have this contain about 1 per cent, of silicon, because the higher 
 the silicon the larger will be the amount of slag made, while if it 
 is too low the iron will be oxidized excessively. Manganese is 
 usually about 0.50 per cent., although it varies anywhere from 
 0.25 per cent, to over 1 per cent., depending on what the blast 
 furnace puts in the pig. Phosphorus is preferred to be less than 
 1 per cent., and sulphur not more than 0.10 per cent., because 
 neither of these elements is entirely eliminated during the proc- 
 ess. A large amount of phosphorus in wrought iron is not, how- 
 ever, as objectionable as it is in steel, because the slag mechanic- 
 ally mingled with the wrought iron hinders it from being brittle 
 under shock, which is the chief damage caused by phosphorus. 
 Pig iron containing 2.50 per cent, and even 3 per cent, of phos- 
 phorus, and as much as 0.35 per cent, of sulphur, is sometimes 
 used. The larger the amount of impurities the larger the loss of 
 metal in the process, both because they are oxidized and lost and 
 because, in the case of silicon, the oxidation increases the amount 
 of slag made, which, in turn, increases the amount of iron slagged 
 off. 
 
 Puddling Furnaces, — There are many different varieties of pud- 
 dling furnace, varying in capacity from 300 to 1500 lb. and even 
 more, but the commonest is probably the 500-lb. furnace, built 
 either single or in pairs, back to back, the latter arrangement 
 having the advantage of reducing loss of heat by radiation, which 
 is always a very large factor. Puddling furnaces are heated by 
 gas or bituminous coal. The commonest method is a deep bi- 
 tuminous-coal fire, giving a long flame, and with a large area of 
 grate in relation to the area of the hearth in order that a high 
 temperature may be maintained. 
 
 57 
 
58 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Fettling. — The hearth is lined or "fettled" with oxide of iron 
 in the form of roll scale, or high-grade iron ore, or "bulldog," 
 i.e., roasted puddle cinder, and this oxide, together with the 
 
 ^^^ 
 
 ' ^ff^/;^^^f;t^;^k&^^^^Mm 
 
 !Si««fi5«lfi«55555««i55555Sfi5(55Kfi5l5!^Z«»^ 
 
 ^^^y^yy^^^^yy 
 
 ''V/yx/'yi^yy 
 
 FIG. 34.-— 500-LB. PUDDLING FURNACE. 
 
 metal oxidized during the melting, supplies the'base which auto- 
 matically maintains a very basic slag and also serves as the prin- 
 cipal oxidizing agent of the impurities. The fettling is repaired 
 
 Floor Line" 
 
 riG. 35.— 1500-LB. PUDDLING FURNACE. 
 (Two work doors ) 
 
 between melts as often as is necessary, and suffers wear with 
 each operation. 
 
 Puddling, — The pig iron is .usually charged by hand through the 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 59 
 
 working doors of the furnace, and the puddler^s assistant fires 
 vigorously in order to melt it down as f asl^ as possible, which usu- 
 ally takes about 30 to 35 minutes. As soon as melted, there fol- 
 lows a short stage of 7 to 10 minutes, during which iron oxide in 
 the form of roll scale or very high-grade iron ore is added, in 
 order to make a very basic slag, the charge being thoroughly 
 mixed and cooled, for which purpose the damper is put on and 
 
 FIG. 36.— CHARGING THE PUDDLING FURNACE. 
 
 sometimes even water is thrown on to the bath. The object is to 
 reduce the temperature to the point where the slag will commence 
 to oxidize the impurities and especially the phosphorus and sul- 
 phur ahead of the carbon. At low temperatures, phosphorus 
 oxidizes ahead of carbon; at higher temperatures, carbon in 
 preference to phosphorus. See Table XI for slow removal of 
 carbon at first. As soon as the carbon reaction is started, light 
 flames begin to break through the covering of slag, produced by 
 burning carbon monoxide: ' 
 
 L Fe203 + 3C = 3CO + 2Fe; 
 2. CO + = CO„. 
 
60 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 37.— PUDDLING. 
 
 FIG. 38.— THE BOIL. 
 CA. removable iron shield dowa which water flows protects the puddler from the heat.) 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 61 
 
 The slag must be very basic at this time lest the CO reduce phos- 
 phorus and sulphur and cause them to return to the metal. 
 As the carbon monoxide forms more and more abundantly, the 
 charge is more violently agitated by its escape, and the ''boil" is 
 in progress. The formation of gas in its interior causes the 
 charge to swell greatly, and it thus rises in the furnace and a 
 large amount of slag pours out of the slag-hole and into a waiting 
 buggy. About one-half of all the slag produced during the proc- 
 ess is removed at this time. The boil continues from 20 to 25 
 
 FIG. 39.— GETTING A HOLD OF THE BALL. 
 
 minutes, and during this time the puddler stirs or "rabbles" the 
 charge vigorously with a long iron rabble, shaped like a hoe. 
 Toward the end of the boil the metal begins to come to nature, 
 and points of solid metal project through the cover of slag, while 
 other pasty masses form on the bottom of the furnace. Both of 
 these things must be corrected immediately by the puddler, lest 
 (1) the iron that is exposed to the furnace gases become too 
 much oxidized, (2) lest the iron sticking to the cold bottom be- 
 come too much chilled, or (3) lest the charge be not uniform 
 
62 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 40.— TAKING THE BALL OUT. 
 
 ( ''^^^:z 1 '. — '■ '■ — r-r- 
 
 £, : ^, 
 
 
 K BHIH 
 
 
 
 FIG. 41.— THE BALL ENTERING THE SQUEEZER. 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 63 
 
 in composition. Finally, all the charge comes to nature and 
 the "balling" period begins and occupies about 15 to 20 minutes. 
 During this period the bath is divided into three or four portions, 
 which are each rolled up into a ball, consisting of a large number 
 of particles partially welded together. The balls are rolled up 
 near the fire-bridge in order, first, to protect them from direct 
 contact with the flame, and second, to keep them as hot as 
 possible until the puddler can draw them, so that the slag may 
 be fluid and thus more easily squeezed out of the metal. The 
 balls are then squeezed in turn, and the furnace hearth repaired 
 for another charge. The total time between operations is usually 
 from one hour and ten minutes to one hour and 40 minutes- 
 
 FIG. 42.— ROTARY SQtJEEZER. 
 
 Squeezers. — A very common form of squeezer is that shown in 
 Fig. -42, the distance between the inner and outer circle being 
 greater on the entering side than on the outgoing side. As the 
 inner circle revolves, the corrugations on the surface carry the 
 ball around, giving it at the same time a movement of rotation. 
 By the time the ball exits on the opposite side, it has been 
 squeezed and kneaded sufficiently to get rid of a large amount 
 of slag. In European countries the squeezer is rarely used and 
 the ball is "shingled" — reduced under a hammer — to weld its 
 particles together and force out the slag. 
 
 Chemistry of the Process, — The removal of the impurities dur- 
 ing the puddling process are shown in Table XI, which is quoted 
 because it records probably the first successful attempt ever 
 made to study in this way the chemistry of an iron or steel process. 
 A similar study is graphically represented in Fig. 43. 
 
64 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XL— REMOVALS IN HAND PUDDLING 
 
 By Calvert and Johnson, Phil. Mag.j 1857 
 
 Time after 
 charging 
 
 Si 
 
 Sample No. 1. 
 Sample No. 2. 
 Sample No. 3 . 
 Sample No. 4. 
 Sample No. 5. 
 Sample No. 6. 
 Sample No. 7. 
 Sample No. 8. 
 Sample No. 9. 
 Puddled bar. . 
 
 Hrs. Min. 
 
 
 40 
 00 
 5 
 20 
 35 
 40 
 45 
 50 
 
 Per cent. 
 2.275 
 2.726 
 2.905 
 2.444 
 2.305 
 1.647 
 1.206 
 0.963 
 0.772 
 0.296 
 
 Per cent. 
 2.720 
 0.915 
 0.197 
 0.194 
 0.182 
 0.183 
 0.163 
 0.163 
 0.168 
 0.120 
 
 Per cent. 
 0.301 
 
 0.134 
 
 Per cent. 
 0.645 
 
 0.139 
 
 100 
 
 90 
 
 80 
 
 -^70 
 
 S60 
 
 |50 
 
 I 
 
 ^40 
 
 I 
 
 & 
 
 30 
 
 SO 
 
 10 
 
 
 
 
 
 
 J52— - 
 
 , ■ 
 
 
 . 
 
 — 
 
 
 
 ^ 
 
 
 
 Si 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 // 
 
 
 
 
 
 
 
 
 
 ^- 
 
 / 
 
 
 
 
 
 J^ 
 
 ^ 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 / 
 
 ^ 
 
 
 
 J 
 
 / 
 
 
 
 
 / 
 
 
 
 
 / 
 
 U 
 
 
 
 
 / 
 
 / 
 
 
 
 
 ^ 
 
 
 
 
 
 / 
 
 
 
 ^^'^^^ 
 
 
 
 
 
 
 
 10 
 
 so 
 
 70 
 
 40 50 60 
 
 Time (Minutes) 
 FIG. 43.-REMOVALS IN HAND PUDDLING. 
 
 80 
 
 90 
 
 100 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 65 
 
 During the melting-down stage, the silicon and manganese in 
 the puddling charge are almost entirely eliminated, and these re- 
 actions are as complete as they will be by the end of the "clear- 
 
 -PUDDLE ROLLS. 
 
 ing" stage which follows it. Much phosphorus and sulphur are 
 also removed. The boil period is, of course, the period during 
 which the carbon escapes, together with some phosphorus and 
 sulphur which was not removed during the first two periods. 
 
 FIG. 45 —ROLLING PUDDLE BAR. 
 
 Fuel. — The temperature of the puddling process is as high as 
 can be obtained in furnaces of this type without preheating the 
 air.^ The result is a very large waste of heat up the chimney, 
 
 1 Indeed, in some cases the air is preheated by the regenerative process, although this is 
 not the usual practice. 
 
 5 
 
66 THE METALLURGY OF IRON AND STEEL 
 
 although sometimes economy in this respect is obtained by plac- 
 ing boilers, or else furnaces to heat metal for the rolls, where 
 they will receive the waste heat of the puddling furnaces. The 
 two greatest items of expense in the puddling process are the fuel 
 used and the excessive labor, which, on account of the strength 
 and endurance demanded, receives a high price. The amount of 
 fuel burned per ton of iron produced will usually be about one ton 
 of a soft bituminous coal, or a little more, although better figures 
 than this are obtained in some cases. 
 
 FIG. 46.— WROUGHT IRON SHOWING STRINGS OF SLAG MAGNIFIED 
 50 DIAMETERS. 
 
 CUnetched.) 
 
 Losses. — The loss in the puddling process usually averages 
 from 4 to 6 per cent, of the weight of the metal charged. The 
 following table gives a typical example of loss : 
 
 TABLE OF LOSSES IN HAND PUDDLING 
 
 Percentage of 
 loss 
 
 Silicon burned 1 . 00 
 
 Carbon burned 3 . 50 
 
 Sulphur burned 20 
 
 Phosphorus bifrned 50 
 
 Manganese burned .30 
 
 Total 5 . 50 
 
 Iron reduced from oxide* = 1.00 per cent. gain. 
 Net loss 4 . 50 
 
 Slag. — The slag which runs from the tap-hole during the boil 
 is known as "boilings," while that tapped out at the end of the 
 
 1 There is much iron oxidized and carried off in the slag, but there 18 also much reduced 
 by impurities. The figure here given represents the excess of reduction over oxidation; 
 in some cases it runs as high as 6 per cent, or more. 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 67 
 
 process is known as "tappings" or "tap-cinder." The character- 
 istics of the boilings are that they contain a larger amount of 
 phosphorus than the tappings, and also that globules of metallic 
 iron are carried off in the violent agitation of the boil. An anal- 
 ysis of the two varieties, giving a mean composition from seven 
 heats, is as follows:^ 
 
 
 ' Boilings 
 
 Tappings 
 
 Ferric oxide (FeoOq) ... 
 
 6.94 
 
 62.61 
 
 19.45 
 
 6.32 
 
 4.68 
 
 12.90 
 
 Ferrous oxide (FeO) 
 
 64.62 
 
 Silica (SiOj) - . 
 
 15.47 
 
 Phosphoric anhydride (PgOg) 
 
 3.91 
 
 Not determined (MnO, S, CaO, etc.) 
 
 3.10 
 
 
 100.00 
 
 100.00 
 
 Total iron 
 
 53.55 
 
 59.29 
 
 
 
 The amount of slag will depend chiefly upon the amount of 
 silicon in the pig iron. It will average in weight about one-half 
 the weight of the charge, where the silicon is high, as in English 
 practice (say 1.70 to 2 per cent.)-, and about one-quarter to one- 
 third in American practice, where the silicon is about 1 per cent. 
 
 Mechanical Furnaces. — The labor in puddling is very severe on 
 the men, and many attempts have been made to remedy this by 
 mechanical furnaces which will work the charge without so much 
 manual labor. Several forms of mechanical appliances and of 
 mechanical furnaces have been invented, but without any per- 
 manent success. However, the mechanical furnace shown in 
 Fig. 47, devised by James P. Roe, of Pottstown, Pa., has given 
 some satisfactory results. It is suspended on trunnions, and 
 the water-cooled bottom and sides are lined with magnesite brick. 
 The oil and blast for combustion enter through the two trunnions 
 and the products of combustion escape through a stack at each 
 end, which meet above the top of the furnace and discharge into 
 the atmosphere as shown. The furnace is made to oscillate 65° 
 each way from the vertical, which keeps the slag and bath uni- 
 formly mixed and avoids the hand-rabbling of the ordinary 
 
 1 Page 297 of Number 40. 
 
68 
 
 THE METALLURGY OF IRON AND STEEL 
 
 puddling process. The -^hole charge, weighing about 4000 lb., 
 is discharged in one ball by sliding it down the hearth of the fur- 
 nace toward the end, and then out into a hydraulic squeezer of 
 special design, in which it is compressed in three dimensions until 
 it is a slab and ready for rolling. 
 
 Bundled Scrap. — In the United States a good deal of wrought 
 iron is made every year by bundling scrap up into a pile roughly 
 resembling Fig. 30. This is then tied up with wire, heated to a 
 welding heat and rerolled into "puddle bar." This is generally 
 an inferior grade of product. 
 
 FIG. 47.— ROE MECHANICAL PUDDLING FURNACE. 
 
 The Carburization of Wrought Iron 
 
 Wrought iron is converted into steel by the operation of car- 
 burizing, or the adding of carbon to it. This is to-day accom- 
 plished in two ways: (1) By the cementation, or steel conversion, 
 process, in which carbon is allowed to soak into red-hot steel in a 
 manner like in nature to the absorption of ink by blotting-paper; 
 and (2) by the crucible process, in which wrought iron is melted 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 69 
 
 in a crucible with carbon, or with iron containing carbon, e.g., 
 cast iron, washed metal, etc. 
 
 Cementation Process, — We have already observed in describ- 
 ing the blast-furnace process that iron at a bright-red heat will 
 absorb carbon very slowly. The action appears to be a traveling 
 of solid carbon into the interior of solid iron, forming with it a 
 chemical solution. When the steel is at a proper heat, the rate 
 of travel is approximately 3/8 
 of an inch per 24 hours. 
 
 Steel Converting Furnace, — ^A 
 section of the type of cementa- 
 tion-furnace used in Sheffield, 
 England, is shown in Fig. 48. 
 The superstructure, e, is a mere 
 chimney for the purpose of 
 carrying off the products of 
 combustion from the fire at c, 
 and for reducing loss of heat by 
 radiation. The real furnace is 
 the part underneath this super- 
 structure or stack, and it has 
 several small chimneys of its 
 own. The two converting pots 
 are shown underneath the 
 points a a. In Sheffield they 
 are built of stone and are 21/2 
 to 4 ft. wide and deep, and 8 to 
 15 ft. long. On the bottom of 
 the pot is first placed a layer 
 of charcoal in small pieces freed 
 from dust. On this is laid a 
 layer of the wrought-iron bars 
 to be converted, which are 2 to 5 in, wide, 1/2 to 3/4 in. thick, 
 and nearly as long as the pot. *A little space is left (about 1/2 in.) 
 between each pair of bars in order that they may be completely 
 surrounded by the charcoal. On top of the layer of bars is then 
 placed another layer of charcoal, and then a layer of bars, and 
 so on until the pots are filled, each one containing from 10 to 
 30 tons of iron. The top of the pots is then luted air-tight and 
 the fire in the fireplace, c, is lighted. 
 
 In about two days the temperature has reached a full red-heat 
 
 FIG. 48.— CEMENTATION FURNACE. 
 
70 THE METALLURGY OF IRON AND STEEL 
 
 of say 650° to 700° C. (1200° to 1300° F.), and this is maintained 
 from 7 to 11 days longer, depending upon the grade of steel to be 
 made. The names of the different grades of steel made in Shef- 
 field are as follows: 
 
 No. 1. Spring heat . 50 per cent, carbon 
 
 No. 2. Country heat 0. 60 per cent, carbon 
 
 No. 3. Single-shear heat 0. 75 per cent, carbon 
 
 No. 4. Double-shear heat 1 . 00 per cent, carbon 
 
 No. 5. Steel-through heat 1 . 25 per cent, carbon 
 
 No. 6. Melting heat 1 .50 per cent, carbon 
 
 The product is controlled by a series of trial bars which' are so 
 placed that, beginning about seven days from the full-red heat, 
 they can be withdrawn from the furnace from time to time, 
 broken, and examined. The appearance of the fracture teUs the 
 extent of the cementation. When the cementation has pro- 
 ceeded to the desired point, the fire is withdrawn and the furnace 
 is allowed to cool for about a week, when the pots are opened and 
 the bars withdrawn through the door b. 
 
 Blister-steel. — The product of the cementation process is 
 known as "blister-steel" because its surface is covered with 
 blisters, due to the formation of gas by a reaction between the 
 carbon absorbed and the slag contained in the wrought iron: 
 
 C+FeO = CO+Fe. 
 
 The blister-steel has gained about 1 per cent, in weight over 
 the wrought iron due to added carbon, and the appearance of 
 the fracture is entirely different, as the broken surface now shows 
 large bright crystals of cementite, — i.e, carbide of iron. 
 
 Shear-steel. — The bars of blister-steel are sometimes forged to 
 a smaller size, piled up, and then the pile forged down again into 
 a bar, which makes what is known as "single shear-steel." 
 Single shear-steel may be again piled and forged into double 
 shear-steel. It is now more common to melt the blister-steel in 
 crucibles, which separates the metal from the slag it contains. 
 
 Crucible or Cast Steel. — The cementation process, on account 
 of the length of time and the very large amount of fuel required, 
 has now been largely superseded by the crucible process. In this 
 process the wrought iron is cut up into small pieces and melted 
 in covered crucibles, the desired amount, of carbon being placed 
 on top of the charge before the melting, together with any other 
 alloying element desired, such as chromium, tungsten, mangan- 
 ese, etc. 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 71 
 
 Furnaces. — In Sheffield, England, coke-furnaces, or melting- 
 holes, containing each two crucibles, are almost universally used, 
 while in America gas-furnaces, containing about six crucibles 
 each, are the common type. In the gas-furnace it is necessary 
 
 m m 
 
 FIG. 49.— REGENERATIVE GAS CRUCIBLE FURNACE OR "MELTING-HOLE." 
 
 that the gas and air for combustion shall be preheated, in order 
 that we may (1) obtain fuel economy and (2) reach the desired 
 temperature for melting quickly. 
 
 Regenerative Furnace. — The operation of the regenerative 
 furnace is shown in Fig. 49.^ 
 
 ^ For description of the operation see page 49. 
 
72 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Crucibles. — In England, the crucibles are made of fire-clay. 
 They are usually made by hand at the steel melting plants and 
 are dried for one or two months, on a shelf next to the chimney 
 of the melting-furnace. Before being used they are heated to 
 redness in an annealing furnace and are then ready to receive a 
 charge of 50 lb. of metal. The clay is deeply cut by the slag, and 
 therefore the charge must be reduced to 44 lb. for the second melt 
 and 38 lb. for the third, in order that the slag-line may be lower 
 each time. After the third melt they are thrown away. The 
 advantages of clay crucibles are that the first cost is lower, and 
 
 FIG. 50.— SECTION THROUGH SHEFFIELD MELTING HOUSE. 
 
 that they do not give up any carbon to the metal, so that the 
 composition of the final product may be regulated with greater 
 exactness and a lower carbon steel may be made, if desired.- 
 
 In America, the crucibles are usually made of a mixture 
 approximating 50 per cent, graphite and 50 per cent, fire-clay. 
 They are made and tempered by factories outside of the steel- 
 works and are received by the latter ready for use. They last 
 about six heats, after which the bottom is sawed off and used for 
 the top of a new crucible. They hold almost 100 lb. of metal 
 because at high heat, they are stronger than clay and can there- 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 73 
 
 fore stand greater strains. Grasping a crucible with the high- 
 leverage tongs used and raising it from the furnace puts it under 
 a severe strain. 
 
 Metal Used. — Although crucible steel is supposed to be made 
 by the melting of pure wrought iron with charcoal, washed metal, 
 ferromanganese and other "physic," it is not at all uncommon 
 for the wrought iron to be diluted with varying amounts of 
 cheaper scrap steel, which lowers the quality of the product. The 
 pieces of iron are put into the crucible first, and on top of that 
 is placed tfie charcoal or pig iron, ferromanganese or spiegel- 
 eisen, and various physics, such as salt, potassium ferrocyanide, 
 oxide of manganese, etc. The purpose of these physics is not 
 entirely clear. Probably the salt and oxide of manganese make 
 a more fluid slag; the ferromanganese puts a little manganese 
 in the steel; and the ferrocyanide may, perhaps, favor the ab- 
 sorption of carbon by the steel. 
 
 In some cases the material is charged directly into a hot cruci- 
 ble from the previous melt, but when graphite crucibles are used, 
 these are sometimes allowed to cool, in order to be examined for 
 cracks, because the breaking of a crucible in the furnace, allowing 
 the liquid mass to flow out upon the floor, is very objectionable. 
 There is a hole in the middle of the floor of the furnace, so that if 
 such an accident happens the metal may run down into a pit 
 underneath; and the floor is also covered with a few inches of 
 coke-dust to absorb the metal in such a contingency. 
 
 FIG. 51.— STAGES IN CRUCIBLE STEEL MELTING. 
 
 Melting. — In the case of a gas-furnace the crucibles are placed 
 in the furnace and the gas and air turned on. In the case of a 
 coke-furnace, the crucibles are placed upon their fire-clay stands 
 and the coke packed around them upon the fire left from the last 
 melt. At the end of about an hour, the coke fire has burned 
 down so low that it has to be poked down and more coke added. 
 After another hour or so the coke fire requires further attention 
 and at this time, care must be taken that no coke falls into the 
 
74 
 
 THE METALLURGY OF IRON AND STEEL 
 
 crucible when the cover is off. The melting takes on the average 
 from two to three hours in both cases, and the charge is examined 
 by removing the crucible cover, to make sure that it is entirely 
 molten. 
 
 Killing. — When the charge is entirely molten, it is kept in the 
 furnace for one half to an hour longer in order that it may teem 
 "dead," that is, pour quietly without the evolution of gas, and 
 yield solid ingots. If the "killing" time is too long, the ingots 
 will be solid, but the steel will be hard, brittle^ and weak, probably 
 as a result of the absorption of too much silicon from the walls 
 of the crucibles. Graphite crucibles probably yield more silicon 
 
 FIG. 52.— POURING. 
 
 than clay crucibles. Just what takes place during the killing 
 of steel is not definitely known. Some have suggested that the 
 gas contained in the steel is ehminated from it during this time, 
 but the alternate suggestion that the principal effect of the killing 
 is to cause the steel to absorb silicon, becoming sound on this 
 account, is the more generally accepted one. The amount of 
 silicon in the final steel will vary greatly, but will average per- 
 haps from 0.10 to 0.50 per cent. 
 
 Pulling. — When, in the judgment of the melter, the steel has 
 been properly killed, the crucibles are removed from the fire by 
 the puller-out, who straddles the top of the furnace and grasps 
 the crucible with a pair of tongs, his legs and arms being swathed 
 in wet cloths to. protect him from the heat, and his eyes frequently 
 being protected by heavy blue glasses. The puller-out then 
 passes the crucible to the pourer, who pours it as shown in Fig. 
 
MANUFACTURE OF WROUGHT IRON AND CRUCIBLE STEEL 75 
 
 52, the slag first being swabbed off with a ball of cold slag on the 
 end of an iron rod. The total time of . operations is 3 1/2 to 
 5 1/2 hours in England, and 3 1/2 to 4 hours in America, 
 
 Ingot Molds, — Ingot molds are shown in Fig. 53, and as a usual 
 thing each mold has a capacity to take the charge from one cru- 
 cible. The metal must be teemed into this with 
 great care, so that the stream shall not touch the 
 sides during pouring. In case a large ingot is to 
 be poured, several crucibles are poured at once 
 into the same mold, care being taken that the 
 metal shall be liquid as long as the pouring con- 
 tinues., or else a "cold-shut" will result; that is, 
 two parts of the ingot not fully melted together. 
 
 Grading. — The composition of the final metal is 
 a matter of some uncertainty, especially as re- 
 gards the carbon and silicon. The former is 
 more easily adjusted when clay crucibles are used, 
 because the amount of carbon dissolved from a 
 graphite crucible will depend to a large extent 
 upon the time and temperature of the operation, 
 etc. The ingots are therefore always graded after 
 they have cooled, by breaking off the upper part 
 of them, which contains the pipe and is therefore 
 useless except when remelted, and examining the 
 fracture with the eye. The skilled steel man can 
 thus estimate the carbon within 0.10 per cent., and 
 the ingots are put away in the pile with others of 
 like analysis. At large American works, however, 
 this grading by eye is always supplemented by 
 chemical analysis. 
 
 Chemistry, — The chemistry of the crucible 
 process is very simple, and consists principally in 
 the elimination of the slag in the wrought iron 
 and the absorption by the metal of carbon, silicon, 
 and manganese. There is also a very slight in- 
 crease in sulphur, which perhaps comes from the pyrite in the 
 clay or graphite, or from sulphurous gases which find their way 
 under the cover of the crucible. Phosphorus also increases 
 slightly, perhaps from the slag of the wrought iron. 
 
 Loss. — The loss is due to the elimination of the slag and to 
 some slight oxidation of metal by oxygen inside the crucible. 
 
 r 
 
 ^ 
 
 
 \ 
 
 ; 1 
 
 \ 
 
 n 
 
 1 
 
 1 \ 
 
 1 
 
 1 
 
 1 
 1 
 
 L* 
 
 
 
 lU 
 
 T 
 
 1 
 
 FIG. 53.— INGOT 
 MOLD FOR 
 CRUCIBLE 
 
 STEEL. From 
 Howe, "The 
 Metallurgy of 
 
 Steel." 
 
76 THE METALLURGY OF IRON AND STEEL 
 
 It is counteracted to some extent by the absorption of carbon, 
 silicon, and manganese, and will average slightly more than 2 
 per cent, in clay crucibles and somewhat less than 2 per cent, 
 in graphite crucibles, the difference being doubtless due to less 
 oxidation in the presence of the graphite. 
 
 Fuel. — In coke fires, the amount of fuel used will be three to 
 four times the weight of steel produced. In gas-fired furnaces 
 the amount of fuel used to make producer gas will be equal to, or 
 slightly less in weight than, the amount of steel produced. The 
 high cost in making crucible steel is due to the cost of crucibles, 
 fuel, labor, and raw material. , 
 
FIG. 60.— A BESSEMER BLOW. 
 
 78 
 
V 
 
 THE BESSEMER PROCESS 
 
 Pig Iron Used. — In the large American works, pig iron for the 
 Bessemer process is preferred to have about 1 per cent, of silicon. 
 Silicon is the chief slag producer and also the chief heat producer. 
 To keep it at a low figure limits the amount of slag made, which 
 limits one of the sources of iron loss. Furthermore, the lower 
 the silicon the shorter will be the time of blow; but it is usually 
 risky to allow it to fall below 1 per cent., or the blow will be cold, 
 and it is only by very rapid working and permitting the least 
 possible delay between operations, so that the converter and 
 ladles are kept very hot, that we are able to get along with as 
 little as this. The manganese is below 0.8 per cent. This also 
 furnishes heat*; but it is now an expensive ingredient of pig iron, 
 and also has the effect of making very liquid slags, which cause a 
 good deal of slopping or "spitting" from the converter (i.e., ejec- 
 tion of the material by the blast), and also make the steel ingots 
 dirty and spotted with oxide spots, due to slag carried over with 
 the steel. Pig iron containing manganese 1.50 per cent, and 
 silicon 1.00 to 0.90 per cent., gives a very '^wet" slag, which 
 follows the metal into the ladle and boils up through it, oxidizing 
 the manganese in the steel: 
 
 (1) FeO + Mn = MnO+Fe. 
 
 The phosphorus and sulphur must be below 0.10 and 0.08 per 
 cent., respectively, in order that the steel may be salable, as 
 neither of these elements is reduced in the acid Bessemer process. 
 
 Mixer. — It takes about two blast furnaces to supply one con- 
 verter with metal, so that a modern plant of two to four convert- 
 ers wiU be operated in conjunction with a large blast-furnace 
 plant. The product of each of these furnaces, if not too different 
 from the desired analysis, will be poured into a huge reservoir, or 
 "mixer," capable of holding 150 to 500 tons, which is then used 
 as a source of supply for the converters. 
 
 1 Indeed, fonnerly, in the Swedish Bessemer practice, the pig iron contained 2 per cent. 
 of manganese, and this element was relied upon as the chief source of heat, because silicon 
 was necessarily low in the Swedish charcoal pig iron. 
 
 79 
 
80 
 
 THE METALLURGY OF IRON AND STEEL 
 
 The mixer serves several very useful purposes: (1) It equalizes 
 the irregularities of pig iron composition by mixing the product 
 of several furnaces, and also brings the composition somewhat 
 under the control of the metallurgist of the Bessemer plant, be- 
 cause he not only can pick and choose from the different furnaces, 
 but he has a few large cupolas under his dominion in which he 
 can melt iron of any desired analysis to pour into the mixer and 
 help regulate its contents. 
 
 (2) Because of its large size, and the fact that it is continually 
 in receipt of new fresh metal, the mixer can keep its contents 
 
 FIG. 61.— MIXER. 
 
 molten for an indefinite length of time, whereas a ladle containing 
 15 tons of pig iron would chill up in a very few hours. Mixers are 
 supplied with blowpipes which can contribute a small amount of 
 heat to the charge, but it is not often necessary to use them. 
 Sometimes they are heated by gas and the regenerative process. 
 
 (3) The capacity of the mixer is so large that a delay either at 
 the blast furnace or at the steel-works will not discommode it 
 greatly, and thus each operation is independent of the other. 
 
 (4) Pig iron in the mixer suffers a slight loss in sulphur, be- 
 cause manganese sulphide forms and, not being very soluble in 
 iron, slowly rises into the slag. 
 
THE BESSEMER PROCESS 
 
 81 
 
 Construction of the Converter,— The coustructibn and dimen- 
 fiions of the converter are shown in Figs. 62 and 63. It consists of 
 a steel shell, riveted together and supported by two trunnions 
 upon which it can be made to rotate. One of these trunnions is 
 hollow, and serves as a wind-pipe to connect the blast from the 
 blowing-engine with the wind-box at the bottom of the vessel. 
 On the other trunnion is fastened a pinion^ which engages with a 
 
 FIG. 62.— FIFTEEN-TON CONVERTER SHELL. From Howe, "The Metallurgy of 
 
 Steel." 
 
 rack joined to a hydraulic piston and of such a length that its 
 movement can rotate the converter through an angle of at least 
 270°, The lining of the bottom is pierced with about 250 half- 
 inch holes, which connect the wind-box with the inside of the con- 
 verter and serve for the passage of the blast. The shape of the 
 converter is such that, when it is lying on its side, the metal will 
 not cover any of these tuyere-holes. This is necessary, or the 
 
 6 
 
82 
 
 THE METALLURGY OF IRON AND STEEL 
 
 blast could never be turned off without having molten metal run 
 down into the wind-box. The converter may have either an 
 eccentric or a concentric shape. The advantage of the eccentric 
 shape is that less heat can escape from the nose; the advantage 
 of the concentric shape is that the vessel may contain its charge 
 when turned on either of the sides, or may receive and pour from 
 either side. 
 
 Lining. — The lining is made of highly refractory acid material 
 composed principally of silica. In England, a ganister rock is 
 
 S^^^ 
 
 ^2^^ 
 
 ^ao 
 
 FIG. 63.- 
 
 -FIFTEEN-TON CONVERTER SECTION, 
 of steel." 
 
 From Howe, "The Metallui^y 
 
 used, or sometimes the lining is rammed up around a pattern and 
 is composed of silicious material held together by a small amount 
 of fire-clay. In America, it consists usually of blocks of ganister 
 or of mica-schist (a silicious rock consisting of pseudostrata, or 
 laminse, formed by tiny plates of mica) laid with a thin layer of 
 refractory fire-clay between, and in such a manner that the edges 
 
THE BESSEMER PROCESS 
 
 83 
 
 of the laminse will be exposed to the wear to which it is subjected.^ 
 After a new lining is put in, it is carefully dried, and every 
 Sunday afternoon, before the converter begins its operation for 
 the week, a wood fire is kept in it for several hours in order to 
 heat the lining to a red heat. Between the heats the lining is re- 
 paired, if necessary, with balls of silicious material and clay. On 
 Sundays, and with an occasional lay-off for which one extra shell 
 
 FIG. 64. — ^From Howe, "The Metal 
 lurgy of Steel" 
 
 PIG. 65.— ECCENTRIC CONVERTER. 
 From Howe, "The Metallurgy of Steel." 
 
 is provided, more extensive repairs are made, and in this way the 
 lining is made to last several months — say 10,000 to 20,000 heats. 
 The converter slags are always high in silica and corrode the lin- 
 ing only slightly. If, however, any uncombined oxide of iron 
 comes in contact with it, it is attacked very rapidly. For this 
 reason the mouths of the tuyeres are rapidly eaten away, and this 
 part of the converter lasts only about 20 to 25 blows. The 
 bottom is therefore fastened to the body with links and keys, so 
 that it may be readily detached and replaced by a new one. 
 
 1 If we represent the blocks of mica-schist by big books, and lay these books in a hori- 
 zontal position with the edges of tlje leaves exposed, it will illustrate the method employed. 
 
84 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Indeed, in some works bottoms are changed with an average delay 
 to the operation of only about 20 minutes for each replacement. 
 Bottoms. — The lining of the bottom is made by placing the 
 tuyere-bricks (see Fig. 66) in position and then filling in around 
 them with refractory material consisting of damp silicious ma- 
 terial held together with clay and containing usually some coke 
 breeze, which seems to lessen the chemical activity of the corro- 
 sion. The details of lining vary so greatly that no general rules 
 can be given. The number of tuyeres is from 18 to 30, the number 
 of holes in each from 12 to 18, and the size of the holes 3/8 in, 
 (England) or 3/8 to 5/8 in. (America).^ The correct lining is of the 
 
 FIG. 66.— COREOSION OF THE BOTTOM LINING OF BESSEMER CONVERTER. 
 
 greatest importance and is the most influential factor in deter- 
 mining the life of the bottom, which furthermore depends upon the 
 care in drying, the temperature of blowing, the pressure of blast, 
 and the composition of pig iron. A bottom should dry 36 hours 
 or more. Its life is shortened by (1) hotter blows, (2) longer 
 blows, (3) lower blast pressure (because the blast holds the metal 
 away from the mouths of the tuyeres), and (4) more manganese 
 in the pig iron (because a wet slag is more corrosive). Between 
 heats, when the vessel is on its side receiving the recarbiu'izer, 
 
 ^ The acid Bessemer process finds its greatest importance in America, and next to that 
 in England. Germany is the leader in basic Bessemer practice. 
 
THE BESSEMER PROCESS 85 
 
 pouring into a ladle, or receiving a new charge, the lining of the 
 bottom can be repaired. For instance, if one tuyere eats away 
 faster than its fellows, the excessive corrosion can be prevented 
 by stopping it up with mud, because, if no air passes through the 
 holes, no oxide of iron is formed at their mouths. Or a worn 
 tuyere may be replaced by a new one, eic, etc. These repairs 
 are chiefly made through the wind-box,, the back plate of which 
 is removable. 
 
 When a bottom is worn out, it is taken away and a new one 
 brought on a car and placed under the converter, which is in the 
 vertical position. Around the 
 
 top is piled a ring of thick wet ^^C^^^^^^^^^^ 
 
 mud, and, as the bottom is forced /y^^P^ottonT^^^bL . 
 
 up against the body by hydraulic ^^^^ /^mereast 
 
 pressure, the mud is squeezed «!, .^ W / ^^^^^ 
 
 7 . . Shoulder ^^ / NokI 
 
 into a firm joint. Lack of space ^^ ^^^^^4^^^;gs^aw 4 
 
 prevents an account of some of ^^^^^^^^^^^^ 
 
 the interesting expedients that ^^ii^^^^T 
 
 are resorted to to stop an occa- fig. 67.— converter paints. 
 sional leak in this joint without 
 
 delaying the first heat, which must be avoided if possible, as the 
 first heat on a new bottom is already too liable to be a cold one. 
 
 Operation of the Converter. — ^Figures 68 and 69 are sections 
 through a converter plant at different points. In Fig. 68, A is 
 the- mixer, capable of containing, say, 300 tons of metal; B is the 
 ladle carriage from the blast furnace, from which the ladle C has 
 been raised to pour the metal into the mixer. Immediately 
 above D is the ladle that is to take the metal from the mixer to 
 the converter, for which it is transferred along a level track until 
 it comes to E (Fig. 69), where it is poured into the vessel, now in 
 the horizontal position, as shown by the dotted line. When the 
 metal has been poured in, the wind is turned on and the vessel ele- 
 vated into the vertical position. The blast now pours through the 
 18 inches or so of metal in the bottom of the converter in a wide 
 spray of tiny bubbles until the impurities are oxidized, when it is 
 turned again into the horizontal position and the wind cut off. 
 In anticipation of this a predetermined quantity of spiegeleisen 
 has been tapped from the spiegel cupola into the ladle at H (Fig. 
 68). (For soft steel ferromanganese, not spiegeleisen, is used.) 
 
 Spiegeleisen is pig iron very high in manganese. Some analy- 
 ses are given on page 8. It is melted in the spiegel cupola to- 
 
86 
 
 THE METALLURGY OF IRON AND STEEL 
 
 gether with a predetermined amount of high silicon pig iron, and 
 is then used to recarburize the bath in the vessel, for which pur- 
 
 llCT.-t-5 88d 
 
 FIG. 68. 
 
 pose the ladle is now run into the position E (Fig. 69) and its 
 contents poured into the bath. 
 
 After the " spiegel reaction " is completed, the steel is poured 
 
 FIG. 69. 
 
 from the converter into a ladle held at the point by the jib-crane 
 /, or by one of the traveling cranes K or L, The ladle is then car- 
 ried over to a position above the ingot molds into which the 
 
THE BESSEMER PROCESS 
 
 87 
 
 steel is to be teemed In pouring the steel into the ladle, the 
 slag is held back in the vessel as much as possible, because this 
 not only furnishes heat for the next operation, but alsp makes 
 it shorter by as much as 20 or 25 per cent., because the slag, 
 being an oxidized substance, assists in oxidizing the impurities 
 in the next charge. 
 
 The turning on and off of the blast and the rotation of the con- 
 verter is all executed by the blower, who stands upon the pulpit 
 at W and operates the various valves, and also judges by eye the 
 
 Converter, Cupola. 
 
 Ladle. 
 \ Runner. 
 
 FIG. 70.~POURING METAL INTO THE CONVERTER. 
 
 progress of the purification from the appearance of the flame 
 which issues from the mouth of the converter. 
 
 Every effort is made by him to so arrange the different opera- 
 tions of the converter, cranes, and ladles and to bring in both 
 spiegeleisen and iron, that each step shall fit into the others with- 
 out delay to the operation in any of the converters. He also has 
 under his control means for lowering the temperature of the bath, 
 if necessary: (a) By ordering an amount of cold steel scrap to be 
 thrown into the mouth of the vessel during the blow^ and (&) by 
 admitting live steam into the converter with the blast. The de- 
 
 1 Scrap is so valuable for use in the open-hearth process that (6) is now used much more 
 than (a). 
 
88 
 
 THE METALLURGY OF IRON AND STEEL 
 
 composition of the steam very quickly reduces the heat of the 
 blow. Attempts have been made to dispense with the blower, 
 who, on account of the long training and experience necessary, is 
 the most highly paid man in the plant next to the foreman,^ by 
 judging of the progress of the operation with the spectroscope; 
 but when two or three vessels are blowing at the same time the 
 reflected light from one interferes with the spetroscopic in- 
 dications of the others. Furthermore, the spectroscope gives no 
 indications of the temperature of the blow, and until the past few 
 years no reliable pyrometer existed suitable for this use. 
 
 PIG. 71.— FIFTEEN-TON STEEL TEEMING LADLE. 
 
 Speel Ladles. — The ladles to receive the steel and teem it into 
 molds are steel shells lined with a cheaper grade of acid refractory 
 material, because the life of these ladles is. limited to about six 
 heats by the wearing out of the nozzle, and therefore a more ex- 
 pensive lining would be wasted. The arrangements of nozzles, 
 stoppers, and handles shown in Fig. 71 are provided in order that a 
 
THE BESSEMER PROCESS 
 
 89 
 
 thin stream of steel may be poured into the molds and may be 
 interrupted when a new mold is being brought into place. The slag 
 lies on top of the metal, and when this begins to come out of the 
 nozzle the stopper is let down and the ladle carried over a slag car 
 and turned upside down to dump out the slag. ' At this time the 
 blower observes the lining of the ladle in order to tell whether the 
 steel was too hot or too cold. If there is a skull of metal frozen 
 inside the ladle, the steel was too cold; if there is no frozen metal, 
 it was too hot; but if there is a spot of metal here and there on 
 the bottom, it was just right. 
 
 FIG. 72.— TEEMING INTO MOLDS. 
 
 Ingot -Molds, Stools and Cars. — The arrangement of the molds 
 into which the ingots are to be cast is shown in Fig. 73, which also 
 gives the dimensions of molds commonly used for railroad rails, 
 wire, and pipe. Molds last about 100 heats, after which they are 
 so cracked inside that they are with difficulty lifted off the solidi- 
 fied ingot of steel, and are also very liable to tear it, producing 
 cracks which are not easily welded up later because they become 
 oxidized on the interior. A continuous series of mold cars are 
 
90 
 
 THE METALLURGY OF IRON AND STEEL 
 
 fed into the steel-mill at one end and drawn out, with the ingot 
 inside, at the other end. They should be heated so hot that the 
 palm of the hand will not bear the heat on the outside, and are 
 washed inside with a thin clay wash which prevents the liquid 
 metal sticking to the cast iron. At the pouring platform they 
 are moved forward under the ladle by means of a little finger, 
 which enters a notch in the side of the car and is itself carried on 
 the rod of a hydraulic piston. 
 
 FIG. 73.— INGOT MOLDS, STOOL AND CAR. 
 
 5000-lb. molds: 7 ft. high, 15 3/4 in. square at top, 19 1/4 in. at bottom; about 2 1/2 in. 
 
 thickness of cast iron. 
 
 Stripping. — As soon as they can be run out to the stripping 
 house, the ingots are solidified sufficiently on the outside for the 
 mold to be removed, leaving the ingot standing on the car ready 
 to be drawn to the rolling-mill. In the majority of cases it is only 
 necessary to place the jaws of the stripping machine under the 
 lugs on the mold, raise it up to a sufficient height, and then trans- 
 fer it to a stool on a car in the second track of the stripping house. 
 
THE BESSEMER PROCESS 
 
 91 
 
 Sometimes the ingots do not come out so easily, however; and 
 then the plunger is rested on top of the ingot, which is held down 
 by the hydraulic pressure as the mold is drawn upward. Elec- 
 trically operated stripping machines are now preferred to the 
 older hydraulic type. The empty molds are stored in the yard 
 until they are sufficiently cool to be drawn back to the steel-mill 
 for another charge, and during the wait they are washed inside 
 with clay water, which is quickly dried by the heat of the mold. 
 
 FIG. 74.— STRIPPING. 
 
 Chemistry of the Acid Bessemer Process. — If a stream of air be 
 made to impinge upon a melted bath of iron, the metal and im- 
 purities will be immediately oxidized, about in proportion to the 
 relative amounts of each present. We may therefore consider 
 the chemistry of the Bessemer process as a union of the oxygen 
 from the tuyeres with the first element it meets, irrespective of 
 relative affinity, and a subsequent attack upon these oxidized 
 elements by unoxidized ones. 
 
 Practically no oxygen of the air escapes from the bath uncom- 
 bined, even though it has but 18 in. of metal to pass through and 
 amounts to more than 5000 cu. ft. per minute,^ because the blast 
 pressure is made high in order that the air may be broken up 
 into a fine spray of bubbles as soon as it strikes the metal and 
 thus offer a large surface of contact for chemical reaction. As 
 
 1 Air is composed of 20.8 parts, by volume, of oxygea and 79.2 parts of nitrogen, and the 
 amount of blast per minute is 15,000 to 30,000 cu. ft. See analyses of gases in Table XV. 
 
92 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 75.— SOAKING PIT HEATING FURNACES AT ROLLING MILL WHICH 
 RECEIVE THE RED HOT, NEWLY STRIPPED INGOT. 
 
 FIG. 76— LINE OF COOLING INGOT MOLDS OUTSIDE BESSEMER MILL. 
 
THE BESSEMER PROCESS 93 
 
 iron oxides are easily reduced, and as. iron is the predominant 
 element, it especially serves as a carrier of oxygen from the blast 
 to the other elements: 
 
 1. FeO +Mn=Fe +MnO. 
 
 2. FeO + C=Fe +C0. 
 
 3. 2FeO + Si = 2Fe + Si02. 
 
 In the first part of the blow, the oxidefe of carbon also suffer 
 reduction: 
 
 4. 2CO + Si = Si02 + 2C. 
 
 5. CO, +Si = Si02 + C. 
 
 6. C0 + Mii = Mn0 + C. 
 
 At a later period' this condition is reversed as described on pages 
 94 to 95. 
 
 Equilibrium is therefore established by the elements which 
 have the greatest chemical affinity for oxygen, getting practically 
 all of that agent, either directly or by robbing their neighbors. 
 Some iron oxide survives, however, because it is so predominant 
 in amount that its neighbors cannot quite rob it before a part 
 has united with silica, which is either formed by the oxidation of 
 silicon or else won from the lining: 
 
 7. FeO + Si02=FeSi03. 
 
 The ferrous silicate forms a slag with manganese silicate, and 
 this slag will dissolve oxides of iron. Even after the oxide of iron 
 has been so absorbed, it may be reduced by manganese and 
 carbon, although not to the same extent. 
 
 Slag Formation. — Manganese oxide unites with silica: 
 
 8. Mn0 + Si02 = MnSi03; 
 
 and it is perhaps for this reason that manganese, unless very 
 high (say over 2 per cent.); is removed early in the blow. The 
 silicates of iron and manganese dissolve in each other and form a 
 slag, and this slag will dissolve large amounts of iron oxide, 
 manganese oxide, silica and alumina, the latter coming from the 
 vessel lining. At all times the bath is so violently agitated that 
 the metal and slag are intimately mixed, and the reducing effect 
 of manganese, sihcon and carbon on the iron oxide dissolved 
 in the slag limits it in amount. The slag itself therefore serves 
 as a carrier of oxgyen to the impurities. 
 
 Critical/ Temperature. — As the temperature rises, chiefly due 
 to the oxidation of silicon, the chemical afiinity of carbon for 
 
94 
 
 THE METALLURGY OF IRON AND STEEL 
 
 oxygen increases relatively more than that of the other im- 
 purities, and reaction No. 4 ceases and then reverses: 
 
 9. 810^+20 = 2CO + Si. 
 What the exact temperature of this reversal in has never been 
 determined, but we may estimate it as being somewhere between 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Acid Bessemer Blow 
 American Practice 
 
 
 
 
 
 4.00^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Cai 
 
 bon 
 
 
 
 
 
 
 
 
 
 
 
 
 8.00^ 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 Me 
 
 tal 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 '' 
 
 
 
 3.00^ 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 L003i 
 
 
 
 
 
 
 
 
 
 \ 
 
 1 
 
 
 
 
 fe 
 
 V 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 / 
 
 03 
 
 
 \ 
 
 \fA 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 / 
 
 1 
 
 
 '~d 
 
 :%°^ 
 ?"- \ 
 
 
 
 
 
 
 
 \ 
 
 / 
 
 V 
 
 
 1' 
 
 
 Hi. 
 
 ^^- 
 
 "^ •^ 
 
 
 
 
 
 \ 
 
 I- 
 
 
 '- 
 
 
 133456789 10 
 
 Hinutes of Blov^g lEnd of Blow 
 
 FIG. 77.— ANALYSES OF METAL. 
 
 1450^ and 1550° C. (2642° to 2822° F.), and unless the silicon is 
 all oxidized before this temperature is reached, we shall have 
 "residual silicon" in the steel. In English Bessemer practice, 
 where silicon is often above 2 per cent, of the pig iron, this is not 
 rare, because the high silicon takes longer to go and also increases 
 the temperature : 
 
 10. Si + 20 = Si02 (generates 180,000 calories). ^ 
 
 1 It is immaterial from the hfeat standpoint whether oxidation takes place directly or as 
 a result of two reactions: 
 
 2 Fe +26 =2 FeO (generates 131,400 calories) 
 2 FeO + Si = SiOz + 2 Fe (generate a 48.60 calories) 
 
 180,000 
 
THE BESSKMER PROCESS 
 
 95 
 
 There is also a critical temperature for manganese oxidation, 
 above which reaction No. 6 is reversed and residual manganese is 
 left in the steel; but this happens only when the manganese in the 
 pig iron is very high. No warning is given that the temperature 
 is approaching these critical points, because no flame — nothing 
 but sparks — comes from the mouth of the converter until carbon 
 
 
 
 
 
 
 
 ( 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 x^ 
 
 ' 
 
 
 
 
 1701 
 
 lbs. 
 
 eo?£ 
 
 
 
 
 
 A^'^ 
 
 
 
 
 
 ■\ 
 
 \^ 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 /" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 S] 
 
 ag 
 
 
 
 
 
 >r 
 
 
 4Qi 
 
 
 
 
 
 
 
 
 
 
 
 /' 
 
 5001 
 
 JB. 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 ^lifH 
 
 lbs. 
 
 
 
 
 
 
 
 
 
 
 
 /r 
 
 
 
 
 
 
 
 
 
 
 
 
 ¥' 
 
 won 
 
 s. 
 
 
 
 ,''' 
 
 30 :s 
 
 
 
 
 
 
 < 
 
 '/Manga 
 
 leae 
 
 Dxid 
 
 ! ^' 
 
 ^ 
 
 
 
 
 
 «••' 
 
 , •■•^ 
 
 
 / 
 
 
 
 
 
 s 
 
 
 
 
 
 
 
 
 / 
 
 /in 
 
 OlbE 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 / 
 
 
 r> 
 
 
 
 
 
 , 
 
 \ 
 
 
 20^ 
 
 
 
 /lO 
 
 50 lb! 
 
 
 ^ 
 
 ^J 
 
 
 
 / 
 
 
 S 
 
 \ 
 
 
 
 
 
 
 
 
 
 K 
 
 
 / 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 inY 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 4 5 6 7 
 
 Minutes of BlowjuK 
 
 9 10 
 
 End of Blow 
 
 FIG. 78.— ANALYSES OF SLAG. 
 
 begins to burn, and therefore no indication is given to the opera- 
 tor of the degree of heat until too late. 
 
 Second Period, — The second period begins when the carbon 
 commences to burn, which happens in America only after the 
 silicon and manganese have been almost eliminated, as shown 
 by Tables XIII and XIV and Figs. 77 and 78. During this 
 period the reactions consist principally of the oxidation of carbon, 
 although a little silicon passes off at the same time. Further 
 details will be learned from the tables and figures. It is interest- 
 
TABLE XIII.— REMOVAL OF IMPURITIES I l^j THiiJ bJLbb.bivxiLi 
 
 ACCOMPANYING SLAG-ANALYSES 
 
 .x.^v, WITH 
 
 Time after 
 
 Removal of metalloids—, 
 per cent. 
 
 1 
 
 Analysis of corresponding sIh^h — per cent. 
 
 commence- 
 ment of 
 blow 
 
 C 
 
 Si 
 
 Mn 
 
 P 
 
 S 
 
 SiO 
 
 AhO 
 
 FeO 
 
 
 a 
 
 MnO 
 
 CaO 
 
 MgO 
 
 P 
 
 S 
 
 1- 
 
 < 
 
 Min. Sec. 
 Pig iron. 
 
 2.98 
 2.94 
 2.71 
 1.72 
 0.53 
 0.04 
 
 6.45 
 
 3.55 
 
 3.21 
 
 1.25 
 
 0.207 
 
 0.034 
 
 0.370 
 
 3.93 
 2.465 
 0.949 
 0.0S7 
 
 4.00 
 4.30 
 0.90 
 0.10 
 
 3.94 
 4.20 
 1.10 
 0.05 
 
 4.49 
 3.87 
 1.30 
 0.33 
 
 4.35 
 
 4.10 
 1.00 
 0.08 
 
 4.22 
 4.20 
 1.30 
 0.55 
 
 3.5 
 
 3.6 
 
 3.4 
 
 2.4 
 
 0.09 
 
 0.075 
 
 q.o 
 
 0.94 
 0.63 
 0.33 
 0.03 
 0.03 
 0.02 
 
 0.038 
 
 2.39 
 1.08 
 0.11 
 0.06 
 0.04 
 0.06 
 
 1.96 
 
 0.443 
 0.112 
 0.028 
 
 1.02 
 0.03 
 0.03 
 0.03 
 
 1.14 
 0.04 
 0.03 
 0.01 
 
 1.08 
 0.03 
 0.03 
 0.02 
 
 0.88 
 0.10 
 0.05 
 
 0.04 
 
 1.06 
 0.43 
 0.12 
 0.07 
 
 1.70 
 
 0.80 
 
 0.28 
 
 0.05 
 
 0.01 
 
 0.0 
 
 0.0 
 
 1.43 
 0.09 
 0.04 
 0.03 
 0.01 
 0.01 
 
 1.15 
 
 0.49 
 O.IC 
 0.13 
 0.13 
 0.10 
 1.17 
 
 3.46 
 
 1.645 
 0.429 
 0.113 
 
 1.83 
 0.22 
 
 0.12 
 0.09 
 
 0.64 
 0.12 
 0.12 
 0.06 
 
 0.83 
 
 0.11 
 0.09 
 0.07 
 
 1.15 
 O.IC 
 0.15 
 0.08 
 
 5.12 
 3.26 
 0.85 
 0.43 
 
 
 
 0.100 
 0.104 
 0.106 
 0.106 
 0.107 
 0.108 
 
 0.109 
 0.09 
 
 0.06 
 0.06 
 0.06 
 0.06 
 0.06 
 0.06 
 
 0.059 
 
 ^1 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 3 20 
 6 3 
 
 8 8 
 
 9 10 
 
 42.40 
 50.26 
 62.54 
 63.56 
 
 5.63 
 5.13 
 4.06 
 3.01 
 
 40.29 
 34.24 
 21.26 
 21.39 
 
 4.36 
 0.96 
 1.93 
 2.63 
 
 
 6.54 
 7.90 
 
 8.79 
 8.88 
 
 1.22 
 0.91 
 0.88 
 0.90 
 
 0.36 
 0.34 
 0.34 
 0.36 
 
 0.008 
 0.008 
 0.010 
 0.014 
 
 0.009 
 0.009 
 0.014 
 0.008 
 
 
 Steel. 
 9 20 
 
 Pig iron. 
 
 62.20 
 
 .276 
 
 AI2O3 
 
 & 
 P2O5 
 
 17.44 
 
 2.90 
 
 
 13.72 
 
 0.87 
 
 0.29 
 
 0.010 
 
 0.011 
 
 
 8 
 
 6 
 
 1 
 
 H 
 
 o 
 
 d 
 
 o 
 
 11 
 
 m 
 
 ft ^ 
 
 i . 
 § . 
 
 H 
 
 1 
 
 62.65 
 73.24 
 75.63 
 61.30 
 64.15 
 
 40.95 
 
 46.78 
 51.75 
 46.75 
 
 50.20 
 55.26 
 
 47.20 
 40.50 
 
 46.50 
 
 48.76 
 59.82 
 
 48.48 
 
 51.00 
 53.44 
 57.80 
 55.76 
 
 47.16 
 53.26 
 62.34 
 
 44.52 
 
 46.72 
 45.87 
 39.07 
 37.63 
 
 7.98 
 4.51 
 5.19 
 
 4.24 
 5.71 
 AhOa 
 
 8.70 
 4.65 
 2.98 
 
 2.80 
 
 3.86 
 
 2.86 
 2.70 
 
 2.24 
 
 12.90 
 0.78 
 0.98 
 0.72 
 
 2.56 
 1.84 
 1.94 
 1.68 
 
 5.83 
 
 2.28 
 3.90 
 
 2.14 
 
 4.36 
 3.08 
 2.49 
 2.94 
 
 1.93 
 
 0.00 
 
 0.00 
 
 13.47 
 
 13.95 
 
 0.60 
 
 6.78 
 
 5.50 
 
 16.86 
 
 1.80 
 14.20 
 18.52 
 31.19 
 
 0.90 
 34.72 
 21.08 
 35.82 
 
 0.90 
 20.24 
 
 17.04 
 18.48 
 
 0.77 
 13.50 
 
 9.54 
 30.60 
 
 0.70 
 4.20 
 6.24 
 9.45 
 
 . ._. . 
 
 10.52 
 
 8.72 
 7.70 
 9.12 
 
 2.39 
 
 15.78 
 11.83 
 10.92 
 10.82 
 
 12.81 
 
 ■*> 
 
 2.18 
 37.00 
 37.90 
 32.23 
 
 5.44 
 26.31 
 31.01 
 25.43 
 
 0.58 
 13.95 
 
 15.48 
 12.29 
 
 3.40 
 23.90 
 22.80 
 22.23 
 
 2.14 
 29.76 
 23.70 
 21.39 
 
 11.16 
 46.38 
 52.26 
 48.92 
 
 0.65 
 1.11 
 0.96 
 0.75 
 0.75 
 
 30.35 
 2.98 
 1.76 
 1.19 
 
 27.22 
 0.62 
 0.38 
 0.32 
 
 32.07 
 2.60 
 3.25 
 2.35 
 
 31.80 
 0.44 
 0.46 
 0.36 
 
 21.79 
 0.42 
 0.60 
 0.38 
 
 19.10 
 1.26 
 0.70 
 1.00 
 
 0.52 
 0.64 
 0.38 
 0.29 
 0.24 
 
 16.31 
 1.53 
 0.45 
 0.52 
 
 10.88 
 0.22 
 0.14 
 0.11 
 
 6.75 
 0.24 
 0.30 
 0.21 
 
 9.03 
 
 trace 
 
 
 
 
 15 
 
 0.09 
 0.08 
 
 
 
 
 
 
 17 
 
 
 
 
 18 
 
 
 
 
 Steel. 
 
 0.09 
 
 0.040 
 0.040 
 0.045 
 0.045 
 
 0.018 
 
 trace 
 trace 
 trace 
 
 
 
 
 Pig iron. 
 
 0.01 
 0.03 
 0.02 
 0.01 
 
 0.34 
 0.04 
 trace 
 
 trace 
 
 32 
 
 After slagging 
 
 
 
 trace 
 
 End of boil. 
 
 
 
 trace 
 
 Knd of blow. 
 
 
 
 
 Pig iron. 
 
 
 
 
 3' 
 
 
 
 
 
 
 
 
 4 45 
 
 
 
 
 
 
 
 
 5 45 
 
 
 
 
 
 
 
 
 Pig iron. 
 
 
 
 
 
 
 
 
 2 15 
 
 
 
 
 
 
 
 
 4 30 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 Pig iron. 
 
 
 
 
 
 t 
 
 
 
 3 
 
 
 
 
 
 
 
 
 5 
 
 
 
 
 
 
 
 
 5 45 
 
 
 
 
 
 
 
 
 
 . Pig iron. 
 
 
 
 
 
 22.18 
 0.23 
 0.28 
 0.21 
 
 18.37 
 0.54 
 0.29 
 0.46 
 
 
 
 
 2 30 
 
 
 
 
 
 
 
 
 
 5 30 
 
 
 
 
 
 
 6 30 
 
 
 
 
 
 
 Pig iron. 
 
 
 
 
 
 
 
 
 4 15 
 
 
 
 
 
 
 
 
 8 35 
 
 
 
 
 
 
 
 
 9 20 
 
 
 
 
 
 
 
 
 
 1.50 
 1.50 
 1.63 
 1.43 
 1.42 
 1.20 
 0.08 
 
 0.05 
 0.05 
 0.05 
 0.05 
 0.05 
 0.05 
 0.05 
 
 
 
 
 
 
 
 Pig iron. 
 
 
 P2O5 
 
 
 3 
 
 32.6 
 42.6 
 36.0 
 35.6 
 33.0 
 15.6 
 
 
 
 7.26 
 2.57 
 5.91 
 6.17 
 7.89 
 13.43 
 
 
 
 
 
 0.60 
 0.15 
 1.60 
 2.61 
 5.66 
 15.06 
 
 
 
 6 
 
 
 
 
 
 
 
 
 9 
 
 
 
 
 
 
 12 
 
 
 
 
 
 
 
 ,. 
 
 14 30 
 
 
 
 
 
 ...'.1 
 
 16 30 
 
 
 
 
 
 
 ■ " '^ 
 
 
 
 1 
 
 
 
 
 
 
 
 1 See No. 52. 
 
THE BESSEMER PROCESS 
 
 97 
 
 TASLE XIV.— removal of impurities in the BESSEMER 
 
 PROCESS 
 
 Time 
 after com- 
 mence- 
 ment 
 of blow 
 
 Removal of metalloids — per cent. 
 
 Si 
 
 Mn 
 
 References and remarks 
 
 Min. Sec. 
 
 Pig iron. 
 
 5 
 
 10 
 15 
 20 
 25 
 
 Pig iron. 
 5 
 
 10 
 15 
 20 
 25 
 
 Pig iron. 
 
 4 30 
 
 13 
 
 16 
 
 Steel 
 
 Pig iron. 
 
 6 
 
 12 
 
 18 
 
 Steel 
 
 Pig iron. 
 
 6 
 
 9 
 
 13 
 
 Steel 
 
 Pig iron. 
 5 
 
 10 
 15 
 20 
 25 
 
 3.52 
 
 3.6 
 
 3.3 
 
 2.5 
 
 1.0 
 
 trace 
 
 3.5 
 3.6 
 3.3 
 2.5 
 1.0 
 trace 
 
 3.52 
 2.78 
 0.43 
 0.05 
 0.23 
 
 3.57 
 3.95 
 1.64 
 0.19 
 0.37 
 
 3.270 
 2.170 
 1.550 
 0.097 
 0.519 
 
 3.5 
 
 3.6 
 
 3.3 
 
 3.25 
 
 2.0 
 
 trace 
 
 3.00 
 
 2.0 
 
 1.25 
 
 0.75 
 
 0.65 
 
 0.35 
 
 3.0 
 
 1.75 
 
 1.25 
 
 0.9 
 
 0.7 
 
 0.5 
 
 1.85 
 1.21 
 0.93 
 0.28 
 0.27 
 
 2.26 
 
 0.95 
 
 6.47 
 
 trace 
 
 trace 
 
 1.952 
 0.795 
 0.635 
 0.020 
 0.033 
 
 2.25 
 
 1.0 
 
 0.5 
 
 0.2 
 
 0.1 
 
 trace 
 
 1.25 
 0.60 
 0.20 
 0.10 
 trace 
 
 0.75 
 0.25 
 trace 
 
 1.93 
 1.69 
 1.00 
 0.37 
 0.62 
 
 0.04 
 trace 
 trace 
 trace 
 0.54 
 
 0.086 
 
 trace 
 
 trace 
 
 trace 
 
 0.309 
 
 1.0 
 0.35 
 0.2 
 trace 
 
 0.048 
 0.051 
 0.064 
 0.067 
 0.053 
 
 0.014 
 
 trace 
 
 trace 
 
 trace 
 
 trace 
 
 German method. 
 ''Leaving Silicon in 
 the Bath," Stahl 
 und Eisen, vol. iii, 
 pp. 262-264. 
 
 German method. 
 Same reference. 
 
 German method. 
 Carl Rott. 31 
 
 English method. 
 Carl Rott. 31 
 
 English method. Sne- 
 
 ^ lus. See Encyclo- 
 
 pcsdia Britdnnicay 
 
 American ed., vol. 
 
 xiii, p. 334. 
 
 Stahl und Eisen, vol. 
 iii, pp. 262-264. 
 
98 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XIV.— REMOVAL OF IMPURITIES IN THE BESSEMER 
 VROCE&S.— Continued. 
 
 Time 
 after com- 
 
 Removal of metalloids — per cent. 
 
 
 
 
 
 
 
 References and remarks 
 
 mence- 
 
 
 
 
 
 
 ment 
 of blow 
 
 C 
 
 Si 
 
 Mn 
 
 P 
 
 S 
 
 1 
 
 Pig iron. 
 
 3.50 
 
 1.50 
 
 0.71 
 
 1.57 
 
 0.16 
 
 » 
 
 5 
 
 3.55 
 
 0.50 
 
 0.56 
 
 1.60 
 
 0.14 
 
 " Basic Process," Sta?a 
 
 10 
 
 2.35 
 
 0.09 
 
 0.27 
 
 1.43 
 
 0.13 
 
 und EiseUj vol. iii, 
 
 15 
 
 0.07 
 
 trace 
 
 0.12 
 
 1.22 
 
 0.12 
 
 pp. 262-264. 
 
 18 
 
 trace 
 
 
 
 trace 
 
 q.08 
 
 0.10 
 
 . 
 
 Pig iron. 
 
 2.97 
 
 0.53 
 
 0.61 
 
 1.22 
 
 0.15 
 
 . 
 
 4 30 
 
 2.480 
 
 0.009 
 
 0.247 
 
 1.250 
 
 0.206 
 
 "Basic Process," by 
 
 9 15 
 
 0.811 
 
 0.0 
 
 0.0 
 
 1.320 
 
 0.262 
 
 Miiller, at Horde, 
 
 11 15 
 
 0.049 
 
 0.0 
 
 0.0 
 
 0.786 
 
 0.262 
 
 Encyc, Brit., p. 346. 
 
 13 
 
 0.0 
 
 0.0 
 
 0.123 
 
 0.021 
 
 0.206 
 
 ■ 
 
 ing to note that the phosphorus and sulphur in the metal are not 
 eliminated; because the acid slag will not dissolve them even if 
 they become oxidized. For this reason the percentage of these 
 impurities increases slightly during the blow, because their 
 actual weight remains the same, while the weight of the bath 
 decreases. 
 
 Gases. — In connection with Table XV it is interesting to note 
 what a large proportion of the carbon is oxidized to the monoxide, 
 and this is the more important because this formation generates 
 only 29,160 calories, while the higher oxidation generates more 
 than three times as much: 
 
 Ch-20 = C02 (generates 97,200 calories). 
 
 Thus a large amount of heat is wasted, and is generated only when 
 the flame passes out of the mouth of the converter and unites with 
 more oxygen. When the heat of a charge is too low, it is custom- 
 ary at some plants to tip the converter forward or backward, so 
 that a few tuyere holes will be above the level of the bath and will 
 blow free air into the interior of the converter. This results in a 
 portion of the carbon monoxide being oxidized to carbon dioxide 
 inside the converter, and is a limited means of making the blow 
 hotter. Oxidation of additional iron also increases the heat 
 generated. 
 
THE BESSEMER PROCESS 99 
 
 TABLE XV.— ANALYSES OF B0TT0M;-BL0WN CONVERTER-GASES 
 
 Time after 
 
 starting 
 
 blow 
 
 
 
 Per cent. 
 
 
 
 CO 
 
 CO^ 
 
 
 
 H 
 
 N 
 
 Reference^ 
 
 2 min 
 
 
 10.71 
 8.59 
 8.20 
 3.58 
 2.30 
 1.34 
 
 blow. 
 
 9.127 
 5.998 
 4.856 
 1.853 
 
 6.608 
 5.613 
 4.144 
 2.995 
 
 0.92 
 
 4.762 
 1.699 
 0.967 
 0.550 
 
 7.256 
 1.296 
 0.980 
 1.318 
 
 0.88 
 2.00 
 2.00 
 2.16 
 2.00 
 
 0.908 
 1.120 
 1.699 
 
 1.112 
 1.040 
 1.120 
 
 88.37 
 86.58 
 85.28 
 74.83 
 66.24 
 65.55 
 
 86.111 
 73.840 
 73.735 
 81.587 
 
 86.137 
 76.400 
 68.256 
 68.961 
 
 ^ 
 
 4 min 
 
 6 min 
 
 10 min 
 
 12 min 
 
 14 min 
 
 18 min 
 
 3 to 5 min. 
 
 3.95 
 
 4.52 
 19.59 
 29.30 
 31.11 
 
 End of 
 
 Sir Loth- 
 ian 
 Bell. 37 
 
 9 to 10 min. 
 21 to 23 min. 
 26 to 27 min. 
 
 2 to 3 min. 
 
 17.555 
 
 19.322- 
 
 14.311 
 
 38 
 
 8 to 10 min. 
 12 to 15 min. 
 17 to 19 min. 
 
 15.579 
 25.580 
 25.606 
 
 38 
 
 1 See No. 52. 
 
 Slag, — In Table XIII the lime and magnesia in the slag come 
 from a small aiftount of blast-furnace slag which finds its way to 
 the converter through the mixer, in spite of efforts made to hold it 
 back at all points when pouring. Because the blast-furnace slag 
 is basic, it has the effect in the converter of making the slags wet 
 and sloppy, and therefore increasing the loss. Although the iron 
 as shot is only shown in one case, this is not because it is absent 
 ^ in the other cases, but merely because it was not determined. At 
 all times the slag carries a great many pellets of iron, which 
 should be added to the combined iron, since they represent a loss 
 in the process. It is interesting to note how closely the amount 
 of iron in the slag, after the spiegel addition, approximates 
 15 per cent.,^ and there seems to be a chemical balance which 
 fixes this amount as a condition of equilibrium. When the 
 silicon in the pig iron is higher, and therefore the amount of slag 
 made is larger, there is a slightly lower percentage of iron oxide 
 
 * Not including pellets, which average 6 to 8 per cent. more. 
 
100 
 
 THE METALLURGY OF IRON AND STEEL 
 
 in it. The practice of "side blowing for heat/' described above, 
 has the effect of increasing the amount of iron oxide in the slag 
 by increasing the oxidizing influences in the interior of the con- 
 verter. The rise in manganous oxide in the slag during recarbur- 
 izing is, of course, due to the formation of MnO by the action of 
 the manganese on the oxygen of the bath, while the iron oxide in 
 the slag is reduced at the same time by the action of manganese 
 and carbon. 
 
 The weight of slag at different periods of the Bessemer process 
 has been calculated by H. H. Campbell^ from its analyses, with 
 the following average results: 
 
 TABLE OF SLAG WEIGHTS IN BESSEMER PROCESS ' 
 
 Percentage of blow 
 
 finished 
 
 
 Pounds of slag 
 
 Percentage 
 of charge 
 
 20 
 36 
 66 
 
 89 
 
 Silicon flame 
 Brightening 
 Carbon flame 
 Full carbon flame 
 
 1035 
 1146 
 1255 
 
 1385 
 
 1 
 
 4.5 
 5.1 
 5.5 
 6.1 
 
 The final amount of slag made will probably average about 71/2 
 to 8 per cent, of the weight of the metal produced, or, roughly, 
 7 per cent, of the weight of the pig iron charged. 
 
 Flame. — A flame is the result of burning gas, and as practically 
 the only gas in this process is carbon monoxide, there is no flame 
 except during the period when the carbon is burning. In the 
 first part of the blow ^ large number of small sparks issue from 
 the mouth of the converter, consisting mainly of pellets of iron 
 and slag ejected by the blast. At the end of the first two or 
 three minutes, a small tongue of reddish-yellow flame begins to 
 pour from the mouth, showing that the carbon is beginning to be 
 oxidized. This soon increases in size and brilliancy until a 
 white-hot flame, 30 ft. in height, pours from the vessel with a 
 loud roaring sound caused by the boiling of the bath and the 
 passage of the blast and gas through it. This boil lasts until the 
 end of the operation, and the bath is at all times violently agi- 
 tated and intimately mixed with the slag, which greatly facili- 
 tates the reaction between the two. The process at this period 
 
 * See page 158 of No- 2. Old edition, 1904. 
 
THE BESSEMER PROCESS 
 
 101 
 
 FIG. 79.— BESSEMER, FLAMES DURING A BLOW. 
 
ia2 THE METALLURGY OF IRON AND STEEL 
 
 presents a spectacle which is almost unmatched as a pyrotechnic 
 display. Soon the flame begins to flicker or ''feather'' at the 
 edges, as a warning that the carbon is becoming low, and finally 
 it shortens or drops, whereupon the converter is immediately 
 turned down and the blast stopped. The carbon at this time 
 will be about 0.03 to 0.10 per cent., depending on how ''young" 
 or how "full" was the blowing. 
 
 At all times there is a varying amount of slag and metal thrown 
 out of the mouth, so that the converter may be likened to a foun- 
 tain of sparks, the great bulk of which consists of slag. Through- 
 out the blow there is also a constant stream of fume issuing with 
 the flame. It is a brownish-red smoke, which rises to a good 
 height and consists principally of oxide of iron and manganese. 
 Dr. Charles F. Chandler, Professor of Chemistry at Columbia Uni- 
 versity, while observing this smoke at the Homestead steel-works, 
 suggested to me that it might be due to the formation in the bath 
 of iron carbonyl, a volatile compound of iron, carbon and oxygen 
 (Fe(C0)5), and perhaps of manganese carbonyl also. 
 
 Loss, — The difference in weight between the pig iron charged 
 into the converter and the steel ingots made will be 8 per cent, in 
 good practice, although running above that (say to 10 per cent.) 
 in some mills. This is distributed approximately as follows: 
 
 Carbon burned 3.5 per cent. 
 
 Silicon burned 1.0 per cent. 
 
 Manganese burned 0.5 per cent. 
 
 7 per cent, slag @ 15 per cent. Fe 1.0 per cent. 
 
 7 per cent, slag @ 7 per cent, iron pellets 0.5 per cent. 
 
 Volatilized and ejected 1.5 per cent. 
 
 8 . per cent. 
 
 Recarburizing, — It might be thought that a more economical 
 method could be found than that of burning up all of the carbon 
 in the pig iron and then adding the desired amount, and in fact 
 such a method was employed in Sweden, where the carbon was 
 burned down to the desired point, as estimated by the appearance 
 of the sparks issuing from the converter, the vessel being then 
 turned down and the charge held until a hammer-test confirms 
 this estimate. Such a complicated procedure, however, requires 
 a hotter bath and very slow working, and it is much cheaper to 
 burn the carbon until the drop of the flame and then add the 
 requisite amount. In making rail-steel to contain about 0.50 
 per cent, carbon, we will add to 15 tons of blown metal about 
 
THE BESSEMER PROCESS 103 
 
 3000 lb. of melted spiegeleisen, containing roughly 6 per cent, 
 of carbon, 12 per cent, of manganese, and 1,50 per cent, of silicon. 
 The mixture charged into the spiegel cupola for melting must be 
 higher in manganese than this, as there is a loss into the cupola 
 slag by oxidation. 
 
 In making dead-soft steel for wire, material to be welded, etc., 
 we add ferromanganese as high in manganese as possible. This 
 material will contain about 7 per cent, of carbon, 80 per cent, of 
 manganese, and 13 per cent, of iron. It is only necessary to add 
 about 500 lb. to a 15-ton bath, and therefore it will dissolve 
 without being melted, although it is customary to heat it to a red 
 heat in order to lessen the chilling of the metal. This dead-soft 
 material will then have the requisite amount of manganese, but 
 will be low in carbon and frequently less than 0.01 per cent, in 
 silicon. It is not improbable that pure manganese metal, if it 
 were readily obtainable, would be used in many cases instead of 
 ferromanganese, in order that the carbon might be still lower. 
 Ferromanganese and spiegeleisen are made in the blast furnaces 
 by smelting very high manganese ores in a manner somewhat 
 similar to the smelting of pig iron. 
 
 The chemistry of the recarburizing operation is important. 
 The recarburizer is added both to give certain elements to the 
 steel and to take certain impurities out of it. The importance 
 of this second function will be appreciated on recalling that the 
 makers of Bessemer steel were not able to produce a marketable 
 product until they learned to take oxygen out of the liquid bath 
 by means of manganese added with the recarburizer. Finally, 
 the chemistry of the recarburizing operation is complicated by 
 the circumstance that some reactions occur between the recar- 
 ,burizer and the slag of the process. 
 
 Deoxidizing the Bath. — The carbon, manganese and silicon 
 contained in the recarburizer will all deoxidize iron according 
 to the following type reactions: 
 
 (a) Mn+FeO = MiiO+Fe. 
 
 (b) 3C+Fe203 = 3CO+Fe. 
 
 (c) Si + 3FeO=Fe02.Si03 + 2Fe. 
 
 Of these, manganese is the most powerful deoxidizer and will 
 come nearest to ridding the bath of oxide. The CO has the 
 disadvantage of being somewhat soluble in the molten metal; 
 consequently, although large volumes of CO gas escape from the 
 steel during the "spiegel reaction," there is, under normal condi- 
 
104 THE METALLURGY OF IRON AND STEEL 
 
 tions, a constant evolution up to the time when the metal be- 
 comes solid, and there is probably even a small amount of CO 
 retained by the solid metal. This escaping gas is a cause of blow 
 holes, as described in Chapter VII. Silicon not only reduces iron 
 oxide, but also is a prime factor in preventing these blow holes, 
 as will also be described in Chapter VII. The oxides of man- 
 ganese and silicon tend to separate themselves from the liquid 
 bath and it is thus that they remove the oxygen; whereas, the 
 oxides of iron are much more readily retained by the steel. 
 Nevertheless, the removal of oxygen by manganese and silicon 
 is riot complete; because firstly, the deoxidation of the FeO and 
 FegOg is not complete and secondly, the oxides of manganese 
 and silicon do not completely separate from the steel, but, 
 on account of lack of fluidity, remain in tiny particles en- 
 tangled in the solidified mass. Hail steels will probably 
 contain 0.20 to 0.30 per cent, of entangled oxides.^ Un- 
 less there be an excess of manganese and silicon necessary to 
 fulfill reactions (a) and (c), their deoxidized effect is far from com- 
 plete, and it is partly on this account that the recarburizer is so 
 calculated as to leave about 0.70 to 1.10 per cent, of manganese 
 and 0.10 to 0.20 per cent, of silicon in rail steel. 
 
 Reactions Between Recarburizer and Slag. — The slag being an 
 oxidized product and the recarburizer being essentially a reducing 
 agent, there is always a certain interchange of elements between 
 the two, the slag giving up some of its oxygen to the carbon and 
 manganese of the recarburizer, and the latter contributing 
 manganese to the slag, both by virtue of reaction (a) and of the 
 folio Wing: 
 
 (d) FeO(Si02)2 + Mn = MnO(Si02)2+Fe. 
 
 It will be observed by the analysis in Table XIII and especially by 
 Fig. 78, that the slag is decreased in iron oxides and increased in 
 manganese oxide as a consequence of the recarburizing operation. 
 It will also be observed in the same places that there is more 
 iron oxide reduced than manganese oxide contributed and this 
 effect is produced by direct reduction of iron, by means of CO, 
 in accordance with the following reaction: 
 
 (e) 2FeO.Si02 + CO=FeO(Si02)2 + C02+Fe. 
 Calorific Equation of the Acid Bessemer Operation. — The pre- 
 
 ^ In this connection it is interesting to note the effect of titanium in supplementing 
 the deoxidizing effect of manganese and silicon and the consequent superiority of titanium 
 treated rails, which will be discussed on page 169. 
 
THE BESSEMER PROCESS 105 
 
 dominant part that silicon plays in furnishing heat for the acid 
 Bessemer process is shown by a calorific calculation. In making 
 this, let us assume that we have a charge of pig iron weighing 
 30,000 lb., and that we burn: silicon=l,00 per cent.; manganese 
 = 0.40 per cent.; carbon = 3.50 per cent.; iron = 2.00 per cent. 
 And let us assume further that the average temperature during 
 thfe operation will be 1500° C, and that the atmosphere is at 
 0° C. Then: 
 
 Surplus 
 Pound- pound- 
 
 calories calories 
 
 Si + 20 = Si02 produces 1,900,000 
 
 28.4 2X16 
 
 300 1b.^ + 3381b.-638 1b. 
 
 338 lb. of oxygen = 1470 lb. air. 
 
 Specific heat air = . 268 cals. per lb. 
 
 1470 lb.Xl500° C.X0.268 = 591,000 
 
 1,309,000 
 
 Mn + O = MnO produces 198,000 
 
 55 16 
 
 120 1b.2 + 35 1b. = 1551b. 
 
 35 lb. oxygen = 152 lb. air. 
 
 152 lb. X 1500° C.X 0.268 = 61,000 
 
 C + O. = C0 produces 2,552,000 
 
 12 16 
 1050 lb. + 1400 lb. =2450 lb. 
 1400 lb. oxygen = 6087 lb. air. 
 6087 lb. X 1500° C. X 0. 268 = 2,447,000 
 
 Fe + 0=FeO produces 704,000 
 
 56 16 
 
 600 1b. + 1711b. = 7711b. 
 
 171 lb. oxygen = 743 lb. air. 
 
 743 1b.Xl500° C..X0.268 = 299,000 
 
 137,000 
 
 105,000 
 
 405,000 
 
 Total net heat from chemical reactions = 1,956,000 
 
 ^ 30,000 lb. XI per cent. = 300 lb. 
 2 30,000 lb. X 0.40 per cent. = 120 lb. 
 
 Now let us suppose that the specific heat of the metal is 0.20 
 calories per pound, per degree Centigrade; then how many degrees 
 will it be raised by the heat produced in the chemical reactions 
 of the. blow? 
 
 30,000 lb. X 0.20 cals. = 6,000 cals. per 1° C. 
 1,956,000 cals. -^ 6,000 cals. =326° C. Answer. 
 
106 
 
 THE METALLURGY OF IRON AND STEEL 
 
 This simple calculation neglects the heat lost by radiation through 
 the vessel lining, and the heat necessary to raise the silicon, man- 
 ganese and carbon of the bath from their temperature at the , 
 beginning of the blow, to 1500° C, and also leaves out of account 
 the heat produced by the combination of FeO and MnO with 
 SiOj to form the slag. All these figures are relatively less im- 
 portant, but those who desire to calculate with greater delicacy 
 should consult J. W. Richards' very thorough little book, No. 
 6, Part II. 
 
 < 1 
 
 ) ] 
 
 [ 2 S 
 
 t i 
 
 5 6 !l 
 
 ' i 
 
 t ( 
 
 • Jfl U 12 13 llmin. 
 
 3.50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3.25 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3.00 
 
 ^.50 
 2.25 
 2.00 
 L75 
 1.50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 
 
 
 
 
 
 
 ^^1:25 
 
 
 
 
 — 
 
 
 
 — 
 
 k 
 
 
 
 A. 
 
 
 
 
 
 
 LOO 
 
 
 
 
 
 
 
 
 \ 
 
 V- 
 
 
 r\ 
 
 
 
 
 
 
 .75 
 Si 
 
 Mir.-.-50 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 V 
 
 
 
 
 
 s^ 
 
 
 
 
 
 
 
 \ 
 
 V 
 
 
 \ 
 
 
 
 
 
 S— 2S 
 
 \ 
 
 
 
 
 " ~ 
 
 — ■ 
 
 
 
 ^ 
 
 ._- 
 
 
 \ 
 
 /Z 
 
 
 Mn. 
 
 C 
 
 S 
 Ph. 
 
 CuTroft 
 
 
 \ 
 
 
 
 
 
 
 
 \ 
 
 
 
 ^ 
 
 v:.:: 
 
 mmw — 
 
 0123456789 10 11 1213U min. 
 FIG. 80.— REMOVAL OF IMPURITIES IN BASIC BESSEMER PROCESS. 
 
 Basic Bessemer Process. — The basic vessel is mechanically 
 the same as the concentric acid vessel, but the lining is made of 
 calcined dolomite held together by about 10 per cent, of tar and 
 rammed in around a pattern while still warm. In ramming up 
 the bottom wooden pins are rammed in with the lining, and 
 when withdrawn they leave 1 / 2-in. tuyere-holes, through which 
 the blast enters. The object of the basic lining is to resist the 
 chemical action of the basic slag, and it is not desired to have it 
 enter in any way into the chemical reaction. Therefore, before 
 
THE BESSEMER PROCESS 107 
 
 beginning the blowing, lime is added to the bath, equivalent 
 in weight to 14 to 20 per cent, of the iron, in order that it may 
 form a basic slag, take up all the silica formed by oxidation, and 
 prevent this silica from corroding the lining. The greater the 
 amount of acid-making elements in the pig iron — i.e., silica 
 and phosphorus — the more lime must be added to flux them. 
 
 Chemistry of the Basic Process, — There is a degree of similarity 
 between the removal of impurities in the acid blow and in the 
 first part of the basic blow, as will be observed in the analysis 
 in Table XIII, and in Fig. 80. Silicon is removed more rapidly 
 in the basic process, however, both because it is less in amount 
 and because its oxide is greedily absorbed by the basic slag. 
 Manganese, on the contrary, goes more slowly, because its amount 
 is larger and it is not attracted by the slag. Indeed it is not 
 uncommon to find "residual manganese" after a basic blow, and 
 this is considered an advantage because it reduces the amount 
 of manganese necessary to deoxidize the bath. Carbon passes 
 off after the silicon, and the phosphorus does not begin to be 
 substantially removed until after the carbon is practically 
 eliminated.^ Consequently the drop of the flame really marks 
 the beginning of phosphorus removal. 
 
 After-How, — This final, dephosphorization, period of the basic 
 blow is known as the "after-blow," because it follows what cor- 
 responds roughly to the ordinary blow of the acid process. 
 During this after-blow phosphorus is rapidly eliminated and 
 absorbed by the slag. We have no means of knowing by the 
 appearance of the flame how this chemical reaction proceeds, 
 and therefore the after-blow is continued on the basis of experi- 
 ence, by blowing for a certain number of minutes, or else a cer- 
 tain number of revolutions of the blowing engine. Phosphorus 
 goes into the slag in the form of a lean phosphate of lime per- 
 haps having the formula (CaO)4P205, which corresponds to 
 about 60 per cent, of lime and 17 per cent, of phosphorus. A 
 variable amount of sulphur is also eliminated during the after- 
 blow and goes into the slag in the form of CaS. Some also may 
 be volatilized as oxides of sulphur. A fluid slag and the presence 
 of manganese both assist in the removal of sulphur, which is 
 sometimes as much as 75 per cent, of that present. The addition 
 
 1 In short, we have here a reversal of the action, in the Bell-Krupp Process, and for a 
 reason already explained, namely, that carbon has a higher affinity for oxygen at high tem- 
 peratures, such as prevail in the Bessemer process, but has a lower affinity for oxygen at 
 low temperatures, such as prevail in the Bell-Krupp Process. 
 
108 THE METALLURGY OF IRON AND STEEL 
 
 of a certain amount of oxide of iron as a flux with the lime aids 
 in producing a fluid slag and also gives a slightly higher yield of 
 steel. 
 
 Recarhurization. — The bath must not be recarburized in the 
 presence of a basic slag because the latter gives up phosphorus 
 by reduction and the metal is "rephosphorized." What the 
 chemical action is that produces this- " rephosphorization " is 
 not clearly known, but it is probable that manganese is an im- 
 portant factor; hence an additional reason for adding as little 
 manganese after the blow as possible, although this cannot be 
 well controlled as the metal is more oxidised in the basic than in 
 the acid blow, and therefore requires a larger addition of man- 
 ganese to render it free from "red-shortness" and "rottenness." 
 The difficulty is avoided in great part by pouring off all the 
 basic slag possible, and then holding more back in the vessel 
 when the metal is poured into a ladle. The recarburization then 
 takes place in the ladle instead of in the vessel. The other chem- 
 ical reactions of the recarburizing operation are similar to those 
 given under the acid process. (See also page 136.) ' 
 
 Slag. — ^The proportion of basic slag made will naturally be 
 greater than that of acid slag, and will usually total about 25 per 
 cent, of the weight of the metal. It is important to so control 
 the blow that a slag shaU be produced containing 15 per cent, or 
 so of phosphoric acid, as this makes it salable as an agricultural 
 fertilizer. The amount of iron in the slag should not be above 
 10 per cent, in oxidized form, although this will be increased by 
 a few per cent, of pellets. 
 
 Loss. — The loss of metal will be much higher than in the acid 
 process, averaging perhaps 13 to 17 per cent., a part of which 
 is due to the fact that the pig iron used contains a larger amount 
 of impurities to be oxidized: 
 
 Carbon.- 3.7 per cent. 
 
 Silicon 0.5 per cent. 
 
 Manganese 1 . 5 per cent. 
 
 Phosphorus 2 . 5 per cent. 
 
 25 per cent, of slag @ 9 per cent. Fe 2 . 3 per cent. 
 
 Pellets 1 . per cent. 
 
 Fume and ejected 1.5 per cent. 
 
 13^.0 per cent. 
 Calorific Equation. — ^Phosphorus is the main source of heat 
 in the basic Bessemer process as shown by the following calcu- 
 lation: 
 
THE BESSEMER PROCESS 109 
 
 TABLE XVI.— CALORIFIC EQUATION OF THE BASIC BESSEMER 
 
 PROCESS 
 Charge: 30,000 1b. 
 
 Analysis: 0.50 per cent, silicon burned; Temperature of atmosphere = 0°C. ; 
 1.60 per cent, manganese Specific heat of air = 0.268 pound- 
 burned; . calories per 1° C; 
 3.50 per cent, carbon Specific heat of metal = 0.20 pound- 
 burned; calories per 1° C; 
 2.50 per cent, phosphorus Average temperature of blow = 
 
 burned; 1500° C. 
 
 3.00 per cent, iron burned; 
 Principle of calculation same as in Acid Bessemer Operation, p. 105. 
 
 Surplus 
 Reactions. Pound- pound- 
 
 calories, calories. 
 
 2 P + 5 O^PgOs produces 4,420,000 
 
 62 80 
 750 lb. 968 lb. 
 4209 lb. airXl500°^C.X0.268 = 1,692,000 
 
 2,728,000 
 
 Si + 2 = Si02 produces 950,000 
 
 28.4 32 
 150 lb. 169 lb. 
 
 735 lb. air X 1500° C.X 0.268 = 295,000 
 
 ^ 655,000 
 
 Mn + O = MnO produces 794,000 
 
 55 16 
 480 1b. 140 1b. 
 
 608 lb. air X 1500° C. X . 268 - ... : 244,000 
 
 550,000 
 
 C. + = C0 produces 2,552,000 
 
 12 16 
 1050 1b. 14Qpib. 
 6078 lb. airXl500° C.X0.268= 2,447,000 
 
 105,000 
 
 Fe + 0=FeO produces 1,056,000 
 
 56 16 
 900 lb. 257 lb. 
 
 1118 1b. air X 1500° Cx 0.268= 449,000 607,000 
 
 Total net heat from chemical reactions = 4,645,000 
 
 Thus, much more heat is generated in the basic process, but 
 more is required because a large amount is aborbed by melting 
 the additions of lime to form the basic slag, and also there is 
 greater radiation because the blows are longer and the operation 
 is slower. 
 
 Comparison of Acid and Basic Bessemer. — The basic Bessemer 
 
no ' THE METALLURGY OF IRON AND STEEL 
 
 process finds its chief field in localities where the ores produce a 
 pig iron very high in phosphorus, as, for instance, the Minette ores 
 smelted in Germany and elsewhere. Germany is the great do- 
 main of the basic Bessemer process, as shown in Table XIX on 
 page 153. The great skill of the Germans enables them to pro- 
 duce by this means a high grade of structural steel by a process 
 which is not intrinsically as well adapted to improve the 
 quality of the metal as any of the others, because of the greater 
 oxidation of the bath,- the danger of rephosphorization, and the 
 liabilities to error in the "after-blow." England too is a large 
 user of this process, but it is no longer in use in America be- 
 cause this country does not produce the desired grade of pig 
 iron in sufficiently large quantities. We must have as much as 
 1.75 per cent, of phosphorus in the pig iron or the metal will 
 give a cold blow, and this pig iron ihust be obtainable at a 
 cheap price or else the process is not commercial. The oper- 
 ating costs are high for several reasons: (1) the blows are 
 long; (2) the basic lining is more costly than silica and wears 
 out faster; (3) fluxes must be added to form the slag; (4) the 
 mechanical handling is more complicated, due to the extra pour- 
 ing off of slag, etc. The slag is a by-product as a fertilizer, 
 but the farmers have to be educated to use it, which has 
 proven slow and difficult in this country. 
 
VI 
 THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 
 
 Open-hearth Plant. 
 
 The different open-hearth cycles are as follows: 
 
 1. Getting the stock to, and in, the furnace. 
 
 2. Supplying the furnace with fuel and air and preheating both 
 of these. 
 
 3. Working the charge, repairing the furnace, etc. 
 
 4. Recarburizing. 
 
 5. Disposing of the steel and slag. 
 
 6. Repairing and preparing ladles, ingot molds, etc. 
 
 A plan and elevation of a typical open-hearth plant is shown 
 in Figs. 90 atid 91. It is not to be presumed that all plants are 
 laid out in the same manner, but that shown is a modern type 
 which is much favored in America. The furnaces are arranged in 
 a long single row, and with the level of the hearth several feet 
 above the general ground-level of the plant. 
 
 Melting Platform. — On the same level as the hearth, and in 
 front of the furnace, is the melting platform, or working platform, 
 upon which are placed one or more charging machines, depending 
 upon the number of furnaces to be served, running upon tracks ex- 
 tending the entire length of the working platform. The space 
 above this is usually spanned by one or more electric traveling 
 cranes which assist in repairs to the charging machines, in hand- 
 ling materials on thfe melting platform, and in pouring molten pig 
 iron into the furnace where such practice prevails. The melting 
 platform has a small extension around the back of the furnace to 
 afford access to the tap-hole and the ladle into which the steel is 
 poured, and lor putting the recarburizer into this ladle when 
 necessary. Upon the working platform are the valve handles for 
 regulating the admission of gas and air to the furnace and for 
 reversing the current of these at the proper time. 
 
 Gas Producers. — The gas producers are situated outside of the 
 furnace house and in a long line parallel to it. This arrangement 
 has the great disadvantage of placing the men on the working 
 
 111 
 
112 
 
 THE METALLURGY OF IRON AND STEEL 
 
 platform between the smoke of the gas producers and the heat of 
 the furnace, but there seems to be no good may of avoiding it. 
 As the regenerators are usually placed underneath the working 
 
 10 Furnaces 
 
 platform, this situation of the producers gives the least possible 
 distance which the gas has to be carried, and therefore the least 
 possible loss by deposition of tarry components. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 113 
 
 FIG. 92— OPEN-HEARTH MELTING PLATFORM AND CHARGING MACHINE. 
 
 Charging Boxes. 
 
 FIG. 93. 
 
114 THE METALLURGY OF IRON AND STEEL 
 
 Stock, — The stock yards are oftentimes placed between the 
 furnace house and the gas producers, but this nearness is not 
 necessary and stock is frequently stored by the end of the house, 
 or even at some distance. For its transfer to the furnace, the 
 stock is loaded into steel boxes similar to those in Fig. 93. Three 
 or four boxes are supported on a little car, which is transferred in 
 a train to the melting-floor, passing over a pair of scales on the 
 way, where the weight is taken. Between the track of the charg- 
 ing machine and the line of furnaces runs the track upon which 
 these cars are transferred, and if a constant supply of boxes is 
 brought to the machine, it can empty them upon the hearth of the 
 furnace at a rate of about 50 boxes (equivalent to about 125 tons) 
 per hour. 
 
 Casting-pit. — The casting-pit extends all the way behind the 
 furnaces; it is on the general ground-level of the plant and there- 
 fore several feet below the melting platform. This pit is spanned 
 by one or more electric traveling cranes of large capacity, which 
 are used to hold the ladles while the steel is running into them 
 from the tap-hole of the furnace and while it is being teemed into 
 the ingot molds. They are also used to serve the pit for several 
 purposes, such as carrying away the slag, transferring empty 
 ladles to and from the point where they are lined and dried, etc. 
 
 Open-hearth Furnace 
 
 The form of a modem 75-ton open-hearth furnace is shown 
 in Figs. 94 to 95. It consists of a long shallow hearth suitably 
 enclosed in fire-brick and bound together with steel. A rolling 
 furnace is shown in Figs. 90 and 91. There are many more 
 stationary than tilting furnaces, but the general principles of 
 the steel-making operation are the same in both. 
 
 Regenerators. — With the furnace are connected two pair of re- 
 generators which preheat the gas and air for combustion as de- 
 scribed on page 49. The internal volume of each of these cham- 
 bers is equal to 3/5 to 9/10 of that of the working-chamber itself. 
 The larger the regenerators the greater will be the amount of heat 
 intercepted in them, and therefore the lower the. temperature of 
 the gases that go to the stack. The amount of space actually 
 occupied by the bricks, or checkerwork, is the important con- 
 sideration, however, and this should be from 150 to 350 cu. ft., 
 total, for all four regenerators per ton of furnace capacity, the 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 115 
 
 volume of the two gas regenerators usually being about 0.3 to 0.4 
 less than that of the air regenerators, because the volume of gas 
 used is less than that of the air, and also because the gas does not 
 require to be preheated so much, since it is already somewhat 
 warm from the gas producer. During the operation of the 
 furnace more or less slag, dirt, and dust are carried over with the 
 
 ^5 
 
 y i:^ ::^^^C ^< Jr^^<y^^^ 
 
 FIGS. 94 AND 95. 
 "^ Longitudinal Section and Elevation. 
 
 outgoing gases. To intercept this the slag-pockets or dirt- 
 pockets A A (Fig. 95) are provided; but in spite of them the space 
 between the bricks of the checkerwork become partially choked, 
 and for this reason, as well as because the deposit of dust makes 
 the surface of the bricks rough, the total area between the bricks 
 must be much larger than the area of the ports, so that the 
 velocity of the gas will not be lessened. The furnace must be 
 
116 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Longitudinal Section and Elevation 
 
 MalD Gas Flue 
 
 ^4- 
 
 ""'Jr-Zr ' '_"'.r'~ -i-^ -' ' 
 
 
 
 r~r 
 
 ^ fe 
 ^"^ 
 
 
 
 ! 
 
 Casting Pit 
 
 
 P 
 
 
 
 
 
 
 FIGS. 96 AND 97. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 117 
 
 • 
 laid off for repairs when the passages between the bricks are 
 
 chpcked by dirt, but, on the other hand, the interstitial space 
 is limited, because the bricks must be laid in such a way that the 
 maximum amount of surface shall be exposed and the gases forced 
 to the greatest possible contact with them. Recent German im- 
 provements in regenerators include the use of bricks of special 
 shape that will not become clogged. The checkerwork with 
 such arrangements will last 2000 to 2500 heats. ^ The modern 
 construction makes the regenerators as tall as poss^le in order 
 that incoming gas and air may be forced into the furnace by the 
 draught, and also because this chimney effect causes the in- 
 coming gas and air to naturally seek the hottest places and the 
 out-going gas to naturally seek the coolest places, in this way 
 equalizing the temperature in the different parts of the regenera- 
 tors. They should be not less than 15 to 20 ft. high. 
 
 The space underneath the checkerwork should be so large that 
 the incoming gas and air will distribute itself nearly uniformly 
 through the different passages, and the temperature of the fire- 
 bricks at this lower part will be, say, 400° C. (752° F.), although 
 varying, of course, at different furnaces and at different times. 
 When the regenerator is receiving the waste gas from the furnace, 
 the temperature of these bricks will be that of the gases that go to 
 the chimney, say 400° tb 600° C, and when the air or gas is pass- 
 ing through the regenerator on its way to the furnace, these bricks 
 will be somewhat cooler, depending upon the length of time that 
 the regenerator has been in this phase of the operation. The 
 temperature of the bricks at the top of the regenerator will be 
 about 1000° C. (1832° F.), and therefore the air and gas entering 
 the furnace will be the same. 
 
 Valves, — The reversing valves of the open-hearth furnace are a 
 cause of large loss in producer gas, and sometimes the leak 
 amounts to 10 or 20 per cent. Ordinary leaks may be prevented 
 by having water-sealed valves, but sometimes the pressure of gas 
 reaches the point where it overcomes the water pressure and 
 escapes, causing a heavy loss. Moreover, water-sealed valves are 
 open to serious objections: (1) The water may freeze, causing 
 an endless amount of annoyance and trouble ; (2) some water on 
 the inside of the valve is vaporized and the vapor carried into the 
 regenerator or into the furnace, where it absorbs heat; (3) the 
 hoods are liable to warp with the heat; and (4) valves of this type 
 
 *See Iron Age, Aug. 25, 1910, p. 436. 
 
118 
 
 THE METALLURGY OF IRON AND STEEL 
 
 are so heavy that elaborate mechanism is necessary to reverse 
 them. The common butterfly valves burn out rapidly and warp 
 badly. If water-sealed, the hoods warp and leak. Lining the 
 hoods with brick makes additionarweight for shifting and adds 
 very largely to the repairs. The mushroom type requires eight 
 valves to a furnace and an elaborate arrangement for reversing 
 them properly. They burn out badly unless water-cooled, when 
 difficulty is met with from freezing (See Chap. XIX for cuts). 
 
 Ports. — The ports are so arranged that the flame shall be de- 
 flected away from the roof and yet not impinge upon the bath, 
 or impinge only very slightly, because the bath would thereby be 
 
 ^IG. 98.— REMOVABLE BUGGIES TO CATCH SLAG IN SLAG POCKETS. 
 
 oxidized excessively. The gas should be spread out all over the 
 width of the hearth beneath the air, and the two should be 
 brought together just before they enter the laboratory, or work- 
 chamber. The air especially must be kept awa'y from direct 
 contact with the bath, and for this reason the gas-ports are placed 
 below the air-ports, which arrangement has the further advant- 
 age of promoting a better mixing of the two, since the gas is 
 lighter and therefore rises. In America the favorite arrange- 
 ment is two gas-ports, above which is situated a long slit, ex- 
 tending almost the entire width of the furnace, which serves as 
 an air-port. This is not universal practice, however, for in some 
 cases there are two, or even three, air-ports. In one case the 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 119 
 
 two gas-ports are built wide and low, so that they will deliver 
 thin streams and get a better mixing of the two materials for 
 combustion. The area of the air-port in a 50-ton furnace should 
 be about 18 sq. ft., and the combined area of all the gas-ports on 
 one end should be from 8 to 10 sq. ft., depending upon the 
 quality of gas used. 
 
 The roof must be protected from the direct impact of the flame, 
 because even the most refractory silica bricks would be melted by 
 the intense heat. The chimney effect of the regenerators and up- 
 takes gives the gas and air a velocity which causes them to enter 
 the furnace with some force, and the construction of the ports 
 directs the steam in the desired manner. The mouths of the 
 ports are gradually melted away by the intense heat of combus- 
 tion of the outgoing gases, until they finally cease to serve this 
 purpose and it is impossible to get -the proper mixing and the 
 proper kind of a flame in the laboratory of the furnace. The 
 ports must then be repaired, else the temperature cannot be 
 maintained. 
 
 The ports have to suffer the full temperature of the out-going 
 furnace gases (flame), and it has recently become more common 
 to equip this part of the furnace with a water circulating system 
 to keep the brickwork cool, which materially increases its life. 
 The floor of the ports of basic furnaces are now commonly covered 
 with a layer of chrome ore — a neutral refractory material which 
 will not be slagged either by acid or basic fluxes. This protects 
 the silicious brickwork of the port from the basic slag which is 
 carried up by the outgoing gases. A recent novelty is to have 
 the ports capable of being removed and replaced bodily when 
 worn out, like the bottom of a Bessemer converter. 
 
 Draught and Chimney. — ^The draught must be sufficient to catch 
 the flame about in the middle of the laboratory and drag it out 
 through the ports on the opposite side from which it entered with- 
 out allowing it either to drop down and touch the bath (as this is 
 to be heated almost altogether by radiation) or to impinge upon 
 the roof. This draught also has to do the work of overcoming the 
 friction of the outgoing regenerators and flues. Its force will de- 
 pend upon the height of the chimney and the temperature of the 
 products of combustion after they have left the regenerators, 
 which should be about 400° C. (752° F.), though even better 
 economy (i.e., a lower temperature) than this is obtained in 
 many cases. All the heat carried away by these flue gases is of 
 
120 THE METALLURGY OF IRON AND STEEL 
 
 course wasted, but the great cost of the refractory bricks in the 
 regenerative chambers ma^es it unwise to reduce the temperature 
 of the flue gases too much by enlarging the checkerwork. The 
 calculation of the amount of draught produced by any given 
 height of chimney with any given temperature of flue gases, ac- 
 cording to the method of Professor J. W. Richards, is given in 
 Table XVII. 
 
 Roof, — The roof is made very thin and of the most refractory 
 bricks that can be obtained, i.e., almost pure silica, with only 
 enough lime to hold it together in a compact mass. The side 
 walls are also thin, and the radiation from the furnace chamber 
 is great, but this has to be endured, as thicker walls and roof 
 produce endless trouble by expansion and contraction. The 
 roof is arched and suspended from beams independent of the 
 side ^alls. 
 
 TABLE XVII.— CALCULATION OF DRAUGHT OF CHIMNEYS. 
 
 The draught in a chimney is produced by the gases within it being lighter 
 than the air outside, and this lightness is due to their being hotter. In this 
 calculation we assume the following data: 
 
 Air weighs 1.20 oz. per cu. ft. at 0° C. and 760 mm. pressure; 
 Chimney gases weigh 1.03 times air at 0° C. and 760 mm. 
 pressure; 
 
 Data Water weighs 772 times air; 
 
 assumed | Mercury weighs 13.6 times water; 
 
 The friction of a chimney absorbs the equivalent of 0.1 in. of 
 water on a water-guage showing draught, for every 100 ft. 
 height of chminey. 
 
 As the gases rise in the chimney they cool; therefore their average tempera- 
 ture will be about 50° C. cooler than their temperature at the foot of the 
 chimney. Call this average temperature T° C. Then: 
 
 273° C. . 
 
 1.29 oz. X 1 . 03 X -, — =the weight of a cu. ft. of the chimney 
 
 T C. + 273 C. 
 
 gases at the temperature T° C. 
 
 Subtract this weight from 1 . 29 oz. and we get the difference in weight of 
 
 air and the furnace gases at the temperature T° C. Then: 
 
 Height of chimney X difference in weight = ounces draught pressure per 
 
 square foot of chimney area. 
 
 Ounces draught X 12 in. 
 
 — — — = equivalent height of water-gauge. 
 
 X . ^y oz. X tt2i 
 
 From the height of a water-guage so found, we must subtract the amount 
 lost by friction, or 0.1 in. X height of chimney -^ 100. 
 
 For furhter modifications of this type of calculation see No. 53, Part I, pp. 
 164, 166, 194, 200, and 201. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 121 
 
 Life of the Furnace, — The "life" of an open-hearth furnace 
 means the number of heats that it can make cgntinuously without 
 stopping for any more extensive repairs than can be made in the 
 usual week's-end shut-down — i.e., without allowing it to cool 
 down. No figure can be given for this except in the most general 
 way. The life of the furnace will be ended usually in one of 
 three ways: (1) The falling in of the roof, (2) the eating way 
 of the ports, so that the flame can no longer be maintained prop- 
 erly, or (3) the giving out of the regenerators, which may occur 
 either throu^ the choking of the checkerwork, or through a 
 crevice formed by the contraction and expansion of the bricks, 
 so that there is a serious leak between the gas-chamber and the 
 air-chamber, and premature combustion takes place. If a basic 
 furnace makes 350 heats, it is considered good work, ajid we 
 may perhaps tentatively consider this figure as the 'Hhree-score 
 years and ten" of a furnace making steel for structural work and 
 similar purposes. Water-cooled furnaces, and furnaces having 
 greater length — 40 ft. and over — will last nearly twice as long. 
 Three hundred and fifty heats would mean about 18 to 24 weeks' 
 work in America. An acid furnace will last about 1,000 heats. 
 Foundry furnaces will only last about one-half as many heats as 
 given above in many instances. The reasons for this are the 
 higher temperatures usually attained when making steel for 
 castings, and the longer heats. Many steel foundries only take 
 two heats per day from their open-hearth furnaces. 
 
 Construction of Hearth and Bottom, — ^The hearth is made with 
 a thickness of 18 to 24 in. inside the furnace shell, in the form of a 
 shallow dish whose sides reach up to the level of the charging 
 doors, and so constructed that the depth of the metal will be from 
 12 to 24 in. — the former figure in the case of a very small furnace, 
 say 5 to 15 tons, and the latter in the case of one of 50-ton capacity. 
 If the bath is too shallow the oxidation will be excessive and the 
 wear of 'the lining by oxide of iron, with consequent production 
 of slag, will be great;^ if the bath is too deep, the melting and 
 oxidation will be slow. In the case of an acid bottom, that por- 
 tion of the lining next to the shell will be made up of refractory 
 clay brick, and the upper portion will be formed by shoveling in 
 silica sand, spreading it out in a thin layer about 1/2 in. thick over 
 the entire hearth, and then allowing it to sinter at the full heat of 
 
 * In an open-hearth furnace the lining suffers the greatest wear at the side of the bath 
 where the bath is thin, because the oxidation of the metal is the greatest there. 
 
122 
 
 THE METALLURGY OF IRON AND STEEL 
 
 the furnace for a^out 10 minutes, so as to set it firmly in place. 
 Upon this layer will be set another layer in like manner, until 
 the whole hearth is constructed, and then it will be "washed" with 
 a melted bath of old slag to fill up all crevices and give a glazed 
 surface. 
 
 In the case of the basic hearth, the bottom is made of calcined 
 magnesite, held together with 10 per cent, or less of anhydrous 
 tar. In this case the layer of brick next to the lining is very thin 
 and the magnesite and tar are set in in layers by the heat of the 
 
 It Is now oouldered best to bare 
 a 12 ' roof thxough-out 
 
 ThflBS ohMkor bdak 
 ntt on tUe 18 X 12 X 8 
 
 SIlIo* liriok tnta hen iq) 
 oiMj brick balow Un* 
 
 Loam, fstUlsg tto. 
 
 SCouiBMiDBsmoslUbriakonedgo^ 
 1 Coiuw obromo <> ■• >• 
 
 Qnnlte or oonoieto 
 
 These oheoken an not in the 
 uptake but la a obamber just 
 bejond uid openlDg Into in. 
 
 FIG 99.— MATERIALS OF CONSTRUCTION AND LINING FOR BASIC 
 OPEN-HEARTH FURNACE. 
 
 furnace as in the case of the acid lining. The tar burns to a 
 strong coke, which holds the mass together in a firm hard form. 
 In some cases no tar is used, and the calcined magnesite is fritted 
 slightly to hold it together; in other cases 15 per cent, of old slag 
 is used as a bond. Pure magnesite gives a more permanent lining 
 than dolomite, ^ and is now much used since its longer life more 
 than compensates for its greater expense. Even when mag- 
 nesite is used for the bottom, the topmost layer, or working 
 bottom, and the repairs, or "fettling," put in during the intervals 
 of the furnace life, are made of dolomite, because this sets more 
 quickly. As the sides and roof even of the basic furnace are 
 made of silica bricks, it is customary, although not absolutely 
 necessary, to put a layer of neutral material between these bricks 
 
 ^ Dolomite is a magnesia limestone, and after calcining consists of a mixture of lime 
 and magnesite (CaO, MgO), with a little silica and other impurities. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 123 
 
 and the basic hearth, and also to protect this joint frpm 
 excessive heat. The neutral material commonly used is 
 chromite bricks, which are made of ground chromite (FeO, 
 CraOg), held together with tar and then burned to form a firm, 
 hard mass. 
 
 Repairing Bottoms. — Between the heats, bottoms are repaired 
 by filling up holes with acid or basic material, as the case may 
 be, and by more extensive attention at the end of the week. In 
 this way the bottom may be made to last almost indefinitely, 
 unless a part of the charge works its way down a crevice and 
 forces up whole sections of the bottom lining, which not infre- 
 quently happens; while sometimes the charge even works its way 
 out through the bottom of the furnace. A sticky or viscous slag 
 is also liable to bring the bottom up by sticking to it. In the 
 
 FIG. 100. 
 
 tilting furnace, the bottom may be repaired along the point where 
 the worst corrosion usually takes place — i.e., at the edge of slag 
 line — during the operation, by tipping the furnace until this 
 place is uncovered. 
 
 Tap-hole. — In the stationary furnace, the tap-hole is made 
 according to the section in Fig. 100. This is closed by material 
 rammed into it from the door of the furnace. It sets quickly 
 into a solid mass, which is pierced with a pointed bar when it is 
 desired to allow the metal to run out. After tapping, the hole 
 must be entirely freed from metal, and then it is made up and 
 filled anew, ready for, the next operation. This always delays 
 work to some extent, and occasionally the tap-hole causes trouble 
 by the charge working through it prematurely, or, on the other 
 hand, by its becoming so hard that a hole is pierced in it only 
 
124 THE METALLURGY OF IRON AND STEEL 
 
 after a long delay and much difficulty during which the oxidation 
 continues in the charge beyond the desired p6int. In the tilting 
 furnace there is no tap-hole, strictly speaking, but the opening 
 into the metal-spout is closed by loose material, which is scraped 
 away before the furnace is tipped to pour the charge. 
 
 Tilting Compared with Stationary Furnaces. — Tilting or rolling 
 furnaces are more expensive to install than stationary ones, and 
 require repairs to the machinery and power to operate them. 
 They also require special arrangements of ports and uptakes, to be 
 described later, and means for cooling the junction between the 
 movable and stationary parts of the furnace. Their advantages 
 are that they do away with troubles and delays from the tap-hole, 
 the slag can be poured off at any time, and the charge may be 
 tapped at a moment's notice, which is especially advantageous 
 when making steels within very narrow limitg of composition. 
 One can also tap a part of the charge when using a rolling furnace. 
 Furthermore, the back-wall of the bottom may be repaired more 
 easily between heats when the furnace itself is still white hot. In 
 , the stationary furnace the angle of this back-wall will be about 
 60° or 70° from the horizontal, and loose material thrown in from 
 the front doors will not rest on this angle; but the tilting furnace 
 can be tipped until the back-wall is more nearly horizontal, and 
 then loose material will remain upon it until sintered into place. 
 The tilting type is of more advantage in the basic process than in 
 the acid, because it enables the slag to be poured off at will, which 
 is occasionally advantageous in basic practice. Except for 
 foundry work, it is now unusual to install tilting or rolling 
 furnaces. Foundry furnaces are generally small in size; 15 to 
 30 tons capacity is a common limit, and individual furnaces run 
 as low as 5 or even 3 tons. These furnaces are most often charged 
 by hand, and the rolling type facilitates this operation some- 
 what. The ease of tapping is an especial advantage where 
 castings are being made, because delays may cool the metal to 
 the point where castings will not run readily, or the carbon may 
 be burned out excessively and the metal lack fluidity on this 
 account. In that special type of practice known as the Talbot 
 Process, which will be discussed shortly, a rolling furnace is 
 essential because a part of the charge must be poured out of the 
 furnace without removing all. 
 
 Tilting Furnaces, — There are two types of tilting furnaces, 
 known respectively as the Campbell and the Wellman. In the 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 125 
 
 Campbell type the hearth of the furnace is arranged so that the 
 center of tilting is coincident with the center of the ports, and 
 therefore the furnace can be oscillated without shutting off the 
 supply of gas and air. In order to facilitate this, there is a little 
 clearance between the uptake and the furnace proper, and these 
 parts are surrounded by water-cooled castings. In the Wellman 
 type the' gas and air supply must be cut off when the furnace is 
 tilted. In tipping the Wellman furnace the ports move with the 
 hearth, 'and they are therefore seated in a water-tank, which 
 
 FIG. 101.— TILTING FURNACE. 
 
 makes an air-tight connection with the regeneratorsiwhen the 
 furnace is in a horizontal position, but breaks it wheniit is tipped 
 forward, :■, 
 
 The Wellman type is not as expensive to build as the Campbell, 
 and probably requires less repairs. The Campbell type has the 
 advantage that the bottom can be repaired along the slag line 
 without interrupting the operation, and that the lining can be 
 sintered into place by the heat of the flame when the hearth is in 
 any position. This is more important in the acid furnace than in 
 
126 
 
 THE METALLURGY OF IRON AND STEEL 
 
 the basic, where the mixture of dolomite and tar can be set by 
 the heat contained in the furnace walls themselves. In the Well- 
 man type, when the furnace is tipped for pouring, cold air can 
 
 csa 
 
 < 
 
 o 
 o 
 
 Ah 
 Q 
 < 
 o 
 
 enter it through the port-holes, and may oxidize the manganese in 
 the final product after recarburizing. Finally, a great advantage 
 of the Campbell type is the fact that a great deal of ore can be 
 used during the operation, and although the boiling of the charge 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 127 
 
 is violent on this account, metal does not flow out of the furnace 
 ' doors, because the hearth can be tipped in the opposite direction. 
 The slag which runs off during this period is allowed to pass 
 through a hole in the bottom of the port-opening, at the joint be- 
 tween the fixed and the rotating portion, where it is continually 
 exposed to the flame and therefore not liable to chill up. 
 
 Temperature. — ^There is almost no limit to the temperature 
 attainable in the open-hearth furnace by frequent reversing of 
 the valves, except the danger of melting the roof. The tempera- 
 ture is controlled by the intervals of reversing, and by throt- 
 tling the amount of gas and air admitted to the furnace. It is the 
 almost universal practice to reverse the valves every twenty min- 
 utes, in order to maintain a uniform temperature of the regenera- 
 tors. The melter endeavors to keep the charge always in a very 
 liquid state, and at the same time to have a slight excess of air, in 
 order that the atmosphere of the furnace may be slightly oxi- 
 dizing to burn the impurities in the metal. Every excess of 
 oxygen, however, causes a loss of heat, because each volume of 
 oxygen in the air is accompanied by four volumes of nitrogen, 
 which carries heat out of the furnace, but does not assist in any 
 way in the reactions. The actual temperature of the furnace 
 will depend upon the length to which the decarburization is 
 carried, because as the metal gets lower and lower in carbon, it 
 requires a higher heat to keep it fluid. It will average about 
 1600° to 1700° C. (2912° to 3092° F.). 
 
 FueL — The amount of fuel burned per ton of steel made will 
 depend not only on the quality of fuel, and skill of the operator 
 of the furnace, as well as that of the gas-producer man when. this 
 type of fuel is used, but it will also depend on the length of the 
 heats and the temperature desired for the product. The more 
 impure the metal being converted the longer will be the heats; 
 the greater the degree of purification, the longer the heats. For 
 example, if we purify a charge consisting chiefly of pig iron, and 
 remove all but traces of the phosphorus, and most of the carbon, 
 it will take longer than to melt a charge of 20 per cent, pig with 
 80 per cent, of steel scrap and leave 0.70 per cent, of carbon in the 
 steel. Furthermore, low-carbon steel must be delivered hotter 
 than high-carbon steel in order to have the necessary fluidity. 
 The more fluxes we have to melt, the greater the amount of heat to 
 to be supplied, etc. The average fuel consumption when using 
 producer gas will probably be between 400 and 800 lb. of coal 
 
128 THE METALLURGY OF IRON AND STEEL 
 
 per ton of steel made for ingot steel — and this includes the 
 small amount of gas used for preheating ladles, etc. — and 600 to 
 1000 lb. per ton for castings. Natural gas in the Pittsburgh 
 district for ingots will average between 5000 and 6000 cu. ft. per 
 ton of steel. The amount of fuel oil used will be within limits of 
 35 to 55 gallons per ton of steel. Good foundry practice making 
 two heats of steel per day will use not much jover 40 gallons per 
 ton. ' . 
 
 Size of Open-hearth Furnace. — The so-called "standard" open- 
 hearth furnace has a capacity of 50 tons. The bath in such a fur- 
 nace will have a length of about 30 to 35 ft., a width of about 12 to 
 15 ft., and a maximum depth of about 24 in. In America, how- 
 ever, there is now a tendency to increase the size to a capacity 
 of 80 or even 100 tons. The result is a much better opportunity 
 for complete combustion within the laboratory of the furnace, 
 and therefore less deferred combustion as the gases escape 
 through the ports and downtakes. This lengthens the life of 
 the ports and promotes fuel economy. 
 
 The smallest practicable size of an open-hearth furnace is about 
 15 tons, and this is very expensive to operate. There are fur- 
 naces, however, as small as 5 tons capacity, but this is not good 
 practice, even under the special circumstances which alone justify 
 the use of the 15-ton furnace. The maximum practical size will 
 not be far above 75 tons, and the real governing factor is the 
 ability of the mechanical apparatus to handle the raw material 
 and the product, besides which it is difficult to cast so much 
 metal out of one ladle without having the casting temperature 
 of the first metal too hot, or else that of the last metal too cold. 
 
 Basic Open-hearth Practice 
 
 Formerly, the open-hearth practice was divided into two types, 
 known respectively as the "pig-and-ore process" and the pig-and- 
 scrap process." In the pig-and-ore process the charge consisted 
 entirely of pig iron, and the oxidation was hastened by the addi- 
 tion of as much ore as the charge would stand without boiling 
 over; in the pig-and-scrap process the charge consisted of pig iron 
 with large amounts of steel scrap, the proportion of the latter 
 being so large in some cases that the operation became a mere 
 remelting process, there being only enough pig iron for its silicon, 
 manganese and carbon to protect the metal from excessive oxi- 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 129 
 
 datlon. At the present time normal charges consist of pig iron 
 and steel scrap, in proportions determined by financial considera- 
 tions, and the operation is hastened by judicious additions of 
 ore. The amount of steel scrap will vary all the way from to 
 90 per cent., and will average not far from 50 per cent, in America. 
 By decreasing the proportion of impurities in the raw material 
 in the charge the scrap enables the process to be completed in 
 less time. 
 
 It might seem as if the use of molten pig iron direct from the 
 blast furnace, or through the medium of a mixer, might be as 
 advantageous in open hearth practice as in Bessemer; but this is 
 not quite so, because the metal in the open-hearth is subjected 
 to oxidation at the same time that it is being melted, so that 
 charging molten material does not shorten the operation as much. 
 The time and labor cost for charging is less, but this advantage is 
 partly neutralized by the scorification and wear of the hearth 
 produced by the inpouring stream of melted metal. About one- 
 half the open-hearth steel of America is made from molten pig. 
 Where it is used the limestone is charged first, on top of that 
 the steel scrap and any other cold metal, and finally the molten 
 metal is poured in. In this way the hearth is protected as much 
 as possible from being cut, and splashing against sides and roof, 
 which would occur if solid pieces were dropped into liquid metal, 
 avoided. 
 
 Pig and Ore Process. — Where pig iron is cheaper than scrap, 
 the charge may consist entirely of pig and ore and it is then 
 customary to hasten the operations by some of the special pro- 
 cesses, such as the Talbot, the Monel, the Hoesch, etc. These 
 conditions prevail in Germany, and there is occasionally practised 
 the charging of molten pig iron from the blast furnace together 
 with a fairly large proportion of ore. The working down of such 
 a charge is shown in the Table XVIII. 
 
 Such a charge, however, must be under 1 per cent, phosphorus 
 at the start, or the carbon will be gone before the phosphorus and 
 the operation be much delayed by " pigging up." The ore should 
 be charged first and the molten metal poured upon it. After 
 approximately half the time of the heat is past, about one-half 
 the slag is removed, carrying with it the bulk of the phosphorus 
 and sulphur. The interposition of a mixer between the blast 
 furnace and the open-hearth furnace gives a greater uniformity 
 in the analysis of pig iron supplied to the open hearth, and also 
 
130 
 
 THE METALLURGY OF IRON AND STEEL 
 
 T3 
 
 a 
 
 03 
 
 I 
 
 h ^ 
 
 > 
 
 CO 
 
 X 
 
 
 
 oT 
 
 w 
 
 S 
 
 t-1 
 
 
 cq 
 
 ^ 
 
 < 
 
 in 
 
 b-t 
 
 tH 
 
 X 
 
 
 
 1 
 1 
 
 a, 
 e 
 
 <3 
 
 Warm and boils. 
 Begins to foam. 
 Warm and foaming. 
 Foams and is colder. 
 Foams and is colder. 
 Foams and is colder. 
 Cold. 
 Warmer and boils. 
 
 
 1 
 
 lb ore. 
 Warmer and boils. 
 Warmer and boil^. 880 
 
 lb ore. 
 Carbon low enough. 
 353 lb. ferro-manganese. 
 
 Steel. 
 
 c3 
 
 < 
 
 
 
 COOCDOSi-HIOOf-I 
 COCOOSiOOOOl-^tN 
 
 
 3§ 
 
 O i-H CO 
 
 CO cq t- 
 
 t^t>COl>t^t^t^l> 
 
 CO 
 
 lO CO 
 
 CO CO »o 
 
 o 
 
 1^ 
 
 
 CDOOIOOW-^OIO 
 000<NCDOSW.-H00 
 
 s 
 
 I> 00 
 
 CO (N 
 
 00 lO Tj< 
 CO OS CM 
 
 Oi-ti-fi-Hr-lNMW 
 i-Hi-ti-Hi-ri-fi-Hr-ii-l 
 
 o 
 
 Oi OS 
 
 00 t* C3S 
 
 o 
 
 
 
 
 CO O 
 tH OS 
 
 t^ Tt< lO 
 
 lO CO -^ 
 
 COCOO'-^OSlO^OS 
 CO CO CO (N 1-t »-H tH 
 
 I> 
 
 i-i OS 
 
 O O OS 
 
 o 
 
 
 O100lOiOCDC0iO<:D 
 OlO>i-lC0CD00iOI> 
 
 CO 
 
 CM CO 
 
 CO W3 
 
 !>. t* OS 
 
 00 rH CM 
 
 OOl-lfNCOCOTtllO 
 
 CM 
 
 CM CM 
 
 T-H T— 1 O 
 
 CM 01 CM 
 
 13 
 
 B 
 
 CO 
 
 m 
 
 c 
 
 OO00'£'CDO«D-^ 
 COCO(NIM(NCC(NC0 
 OOOOOOOO 
 
 oc 
 
 c 
 
 00 
 IN 
 
 o 
 
 O CO 
 
 O 00 CM 
 
 o§S 
 
 OOOOOOOO 
 
 c 
 
 o 
 
 o o 
 
 o o o 
 
 tG 
 
 c 
 
 OO^WiOCOCOiO 
 O>OC0rHCDC0i0C0 
 iCOOOOOOO 
 
 c 
 
 i-H 
 
 o 
 
 w lO 
 
 CO I-H 
 
 o o 
 
 o o -* 
 
 S§8 
 
 OOOOOOOO 
 
 c 
 
 o 
 
 o o 
 
 o o o 
 
 PLh 
 
 c 
 
 lOiCt^OO-^iO-* 
 OC0(N(NTPTjlCDa) 
 OOOOOOOO 
 
 c 
 -<* 
 
 1— 
 
 00 
 CO 
 1— I 
 
 , T-H OS 
 
 o o 
 
 OS W5 l> 
 lO CO CO 
 
 o o o 
 
 OOOOOOOO 
 
 c 
 
 o 
 
 o o 
 
 o o o 
 
 
 05 
 
 CC 
 
 COtH-^iOCDOOiOOS 
 OOOOOOt-h^ 
 
 r- 
 
 o 
 
 CC 
 
 CM CO 
 OS lO 
 iH C^ 
 
 ^ OS OS 
 CO CD OS 
 CM CM -^ 
 
 OOOOOOOO 
 
 c 
 
 ) o 
 
 o o 
 
 o o o 
 
 o 
 
 ^(NiO^tH«D00I>-«*< 
 (©t^-^COOSt^Oi-lTH 
 
 CO C^ (N (N'i-I 1-1 .-h' r-] tH 
 
 c 
 
 * C3 
 
 CO CO 
 
 d d 
 
 0.09 
 
 0.084 
 
 0.089 
 
 i 
 
 
 If 
 
 '^2^2^2^§ 
 
 1— 
 
 
 d d 
 
 O (N 
 
 d lo d 
 
 O t-( CO 
 
 CS 
 
 .-HC^INCOCOr^TjiiO 
 
 IT 
 
 ) "5 
 
 CO CO 
 
 t* t* t- 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 131 
 
 enables some sulphur to be removed in the mixer. In some cases 
 the mixer is used as a preliminary refiner, about 4 to 5 per cent, 
 of ore, and 5 to 6 per cent.^of lime being added to it. In the mixer, 
 manganese, silicon, phosphorus and sulphur are thus removed in 
 amounts worthy of consideration. Purifying in the mixer in- 
 creases the output of the open-hearth furnace; moreover, the 
 costs for repairs, labor and fluxes in the mixer are comparatively 
 small. 
 
 All-scrap Process. — In certain localities in America, steel scrap 
 is cheaper than pig iron, but its use is open to the objection that 
 it is severely oxidized and wasted during the melting period unless 
 there be some pig iron with it to give a reducing influence. In- 
 stead of using as much pig iron, however, some plants now partly 
 replace it with a carbonaceous material which is a by-product of 
 petroleum refining, and is known as "carbo," This "carbo" is 
 an impure form of carbon which dissolves readily in the steel and 
 serves the purpose of a reducing influence. 
 
 Normal Pig and Scrap Process. Order of Charging. — In the 
 pig and scrap process using about 50 per cent, of each, the order 
 of charging for a 75-ton furnace is as follows: about 6,000 lb. of 
 small steel scrap is spread evenly over the bottom, preference 
 being given to plate scrap because of its greater covering power. 
 On top of this is charged from 12,000 to 16,000 lb. of lime and 
 about 8,000 lb. of ore or scale. Then follows about 70,000 lb. 
 of scrap and 30,000 lb. of pig iron. After about 2 hours the 
 heat of the furnace has softened this charge so that it has sunk 
 in the furnace aiid there is room for the final charge consisting 
 of some 12,000 lb. of scrap and 60,000 lb. of pig. 
 
 Composition of Pig Iron Used. — Basic pig should be free from 
 adhering sand, and therefore only machine-cast metal will ordi- 
 narily be used, or else that cast in metal molds. Its silicon should 
 be below 1 per cent., and its manganese above 1 per cent., though 
 the price of manganese ore makes this ingredient an expensive 
 luxury. It is desired because it makes the slag more fluid and 
 aids in removing sulphur. The phosphorus will be almost any 
 figure up to 2 per cent., but a regular supply of pig with 
 more phosphorus than that would tempt the manager toward 
 the basic Bessemer process. Silicon and phosphorus increase 
 the amount of slag, lengthen the operation, and require lime to 
 flux them.^ 
 
 ^ Analysis of Basic Pig Iron in Table, p. 8. 
 
132 THE METALLURGY OF IRON AND STEEL 
 
 Fluxes, — The function of the basic lining is to remain inert 
 and serve simply as a container for the bath. In order that the 
 slag may be at all times rich in lime (35 to 45 per cent. CaO ordi- 
 narily), from 5 to 30 per cent, of lime is added with the pig and 
 scrap, the exact amount depending upon the impurities in the 
 metal charged. Especially if the charge is high in sulphur, the 
 slag must be kept as rich in lime as possible without making it too 
 infusible and therefore viscous, the limit being about 55 per cent. 
 CaO. The lime is usually added in the form of calcined lime- 
 stone, and, as already noted, the higher the silicon and phos- 
 phorus in the metal the greater must be the amount of lime used. 
 It is also customary to charge ore -with the pig and scrap, in order 
 to increase the amount of oxidation that takes place during the 
 melting period. This ore has very little effect on the basic 
 lining, although it would rapidly corrode a silicons one. Further 
 additions of ore are made during the operation if necessary. 
 The average total amount of ore added to a 50-ton charge 
 will be between 1/8 of a ton and 2 1/2 tons, equal to 0.25 to 5 
 per cent. 
 
 Chemistry. — It takes about 4 or 5 hours for the charge to 
 melt down level, and during that time the silicon is almost en- 
 tirely eliminated, while the proportion of carbon and manganese 
 is somewhat reduced, as shown by Figs. 103 and 104, which also 
 show the chemical changes that take place during the entire 
 operation. These reactions must be controlled by the melter, 
 because it is necessary that the carbon shall be eliminated last, 
 and therefore it is occasionally necessary to "pig up " the charge, 
 i.e., add pig iron in order to increase the amount of carbon. On 
 the other hand, in case the phosphorus is being eliminated very 
 fast, the oxidation of the carbon may be hastened by "oreing 
 down," as it is called, i.e., adding ore to produce the reaction. 
 
 Fe,03 + 3C = 2Fe + 3CO. 
 
 If the carbon is eliminated before the phosphorus, a great deal 
 of iron will be oxidized, because the phosphorus does not protect 
 it as much as the carbon. 
 
 When the charge is entirely melted,* and at intervals thereafter, 
 the melter takes a spoon-ladle full of metal and pours it into an 
 iron mold. As soon as this is set hard it is cooled in water, and 
 
 ^ The charge is not entirely melted until the lime has come up, which takes place about 
 2 hours after the charge has melted down level. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 133 
 
 ' Hours 
 
 FIGS. 103 AND 104.— CHEMICAL CHANGES IN BATH AND SLAG IN A BASIC 
 OPEN-HEARTH FURNACE. 
 
134 THE METALLURGY OF IRON AND STEEL 
 
 from the appearance of its fracture the melter can estimate very 
 closely the amount of carbon and phosphorus it contains. In 
 many plants it is customary to tap heats on these estimates alone 
 and astonishing accuracy can be obtained by these methods as to 
 the amount of carbon contained in the bath of steel. In all large 
 plants, however, the laboratory analysis is -made for phosphorus 
 and usually for carbon as well, and these determinations may be 
 made in 20 minutes or less. 
 
 Functions of the Slag. — The functions of the slag are: (1) To 
 absorb and retain the impurities in the metal, particularly silicon, 
 manganese and phosphorus, and as much sulphur as possible; (2) 
 to lie upon the top of the bath as a blanket and protect it from 
 excessive oxidation by the furnace gases; and (3) to oxidize the 
 impurities of the bath by means of its dissolved iron oxide, which 
 comes chiefly from ore, scale, etc., added by the "melter." 
 
 For the retention of phosphorus and sulphur, the slag must be 
 rich in bases. For the oxidizing it is necessary that the slag shall 
 be fluid, so that it will mix easily with the bath, and therefore the 
 content of lime must not be too high. During the boil, when the 
 carbon is passing off, there is an intimate mixture of metal and 
 slag, and there is even such a violent agitation of the bath that 
 the metal itself is frequently uncovered in places and exposed to 
 direct oxidation by the furnace gases. 
 
 The removal of sulphur is a variable quantity and cannot be 
 altogether depended upon, but there is usually a large loss of this 
 element during the operation, in spite of slight additions of sul- 
 phur from the coal through the gas. This sulphur reduction is 
 greatly assisted if the slag is thinned out (i.e., its melting-point re- 
 duced) by the addition of calcium fluoride (CaFg) just before the 
 end of the operation. The higher the slag is in lime, provided it 
 remains at the same time fluid, the more complete in general will 
 be the sulphur elimination, -but over 55 per cent, of lime usually 
 makes the slag viscous, unless the calcium be added in the form of 
 fluoride, or of chloride (CaClj). The latter agent was added in 
 many cases by the recommendations of E. H. Saniter, but the 
 practice was never general in the United States, and the use of 
 calcium fluoride (known as fluorspar) answers most purposes, 
 unless the sulphur in the charge is excessive. Manganese assists 
 in sulphur removal by forming MnS. The sulphur in this is 
 slowly oxidized and the manganese returns to the metal. 
 
 It is of importance that the final slag in the open-hearth 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 135 
 
 operation shall be a nonoxidizing one, lest the metal itself be 
 full of oxide and very "wild/' i.e., give off gas abundantly during 
 solidification. Therefore for the best quality no ore should be 
 added within one hour of the finish of the operation. 
 
 Melting Low Carbon Steel. — When steel for purposes requiring 
 dead soft metal is made, the charge is calculated so that it will 
 melt practically dephosphorized and of the desired carbon. In 
 other words, very little oxidization goes on after the lime is up. 
 Large amounts of scrap are used, and the necessary amount of 
 ore to take care of the phosphorus, is put in with the charge. 
 
 Slag, — ^Final basic slags will contain: 
 
 10 to 15, per cent. SiOg 
 45 to 55 per cent. CaO + MgO 
 10 to 18 per cent. FeO 
 5 to 15 per cent. P2O5. 
 
 The slag is the crux of the open-hearth operation, and it must 
 be kept slightly oxidizing at all times except just at the end. In 
 the basic process it must be kept always strongly basic in order 
 to retain phosphorus and sulphur and also spare the lining as 
 much as possible. Sufficient lime must be added when necessary 
 to maintain these conditions, but fluidity must not be destroyed 
 or the slag will lose its power to react well upon the impurities, 
 whose chief means of oxidation it is. 
 
 Weight of Slag. — The weight of slag will depend upon the silicon, 
 phosphorus and dirt in the metal, and will average from 10 to 
 30 per cent, of the weight of steel. Since the slag takes practi- 
 cally all the lime charged into the furnace, its weight can be calcu- 
 lated with sufficient accuracy by the same formula given for 
 calculating the blast-furnace slag: Divide the total lime (CaO) 
 used with the charge (f^lus 30 per cent, of itself to allow for wear 
 of the lining) by the percentage of lime in the slag. 
 
 Loss. — The weight of steel produced will be a variable propor- 
 tion of the weight of metal charged, depending upon the amount 
 of pig iron in the charge, the amount of ore used, the extent to 
 which the carbon was eliminated before tapping, etfc., but it will 
 average perhaps 93 to 96 per cent., the difference being made up 
 as follows, in a typical example ; 
 
136 
 
 THE METALLURGY OF IRON AND STEEL 
 
 ANALYSIS OF CHARGE 
 
 (50 per cent, pig; 50 per cent, scrap) 
 
 
 Pig iron 
 per cent. 
 
 Scrap 
 per cent. 
 
 Average 
 per cent. 
 
 Loss 
 per cent. 
 
 Carbon 
 
 3.75 
 
 0.70 
 1.50 
 1.30 
 0.37 
 
 0.25 
 0.06 
 0.40 
 0.04 
 0.03 
 
 2.00 
 0.38 
 0.95 
 0.67 
 0.20 
 
 2 00 
 
 Silicon.. 
 
 38 
 
 Manganese 
 
 95 
 
 Phosphorus 
 
 67 
 
 Sulphur 
 
 20 
 
 20 per cent, slag @ 15 per. 
 cent. Fe 
 
 3 00 
 
 • 
 
 
 
 
 
 Iron reduced from ore 
 
 7.20 
 1.50% gain. 
 
 5.70 
 
 Net loss 
 
 
 
 . 
 
 
 
 
 
 
 Recarhurizing. — Steel must not be recarburized in the pres- 
 ence of a basic slag, lest the carbon, silcon, and manganese of the 
 recarburizer reduce phosphorus from the slag and cause it to pass 
 back into the metal: 
 
 2(3CaO.P204) + 5C = eCaCPgOs + 5C0 +2P 
 4(3CaO.P206) + 5Si = 2(6CaO.P20^) +5Si02+4P. 
 
 Therefore in basic practice the recarburizer is added to the stream 
 of metal while it is pouring from the furnace into the ladle, and 
 special arrangements are made for allowing the slag which floats 
 on top of the metal to overflow at the top of the ladle and thus be 
 gotten rid of in a large part. In careful practice "rephosphoriza- 
 tion" need not exceed 0.01 to 0.02 per cent, of the steel, although 
 a much larger increase may take place through accident. 
 
 The recarburizer usually consists of ferromanganese, together 
 with anthracite coal, charcoal or coke which is broken into small 
 pieces and Ipaded into paper bags. About 45 per cent, of the 
 broken coke is burned and the other 55 per cent, will be dissolved 
 by the steel. It is practically impossible to melt spiegeleisen in 
 the cupola, because a cupola must be run continuously in order to 
 do satisfactory work, while the open-hearth process is too slow 
 and too irregular to use a continuous supply of moltAi recar- 
 burizer. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 137 
 
 In making soft steel the carbon is reduced to about 0.10 or 
 0.15 per cent, in the bath, and then the necessary amount of re- 
 carburizer is added. In making high carbon steel, on the other 
 hand, two methods are possible: We may reduce the bath to 
 
 0.10 or 0.15 per cent, and then add sufficient recarburizer to 
 bring the carbon up to the desired point, or we may bring the 
 bath down to only slightly below the desired point and recar- 
 burize it to the desired percentage. The former practice usually 
 
138 
 
 THE METALLURGY OF IRON AND STEEL 
 
 frees the steel more completely from gas and therefore makes it 
 less wild, besides reducing the danger of phosphorus being left 
 in the steel. It takes a longer time for the operation and more 
 fuel. 
 
 Teeming Basic Steel. — R. A. BuU^ gives the following figures 
 showing the oxidation of silicon and manganese, and the enrich- 
 ment in phosphorus and sulphur, by the slag during the teeming 
 of a series of basic heats. The practice here exemplified is in 
 
 AFTER 
 POUKING 
 
 P 
 
 S 
 
 MN 
 
 SI 
 
 MINIMA AND MAXIMA | 
 
 PHOSPHORUS 
 
 SULPHDR 
 
 MANGANESE 
 
 SILICON' 1 
 
 50^ 
 
 .018 
 
 .020 
 
 .79 
 
 .34 
 
 
 L i 
 
 
 i 4 
 
 
 
 
 2 ^ 
 
 i ^ 
 
 
 
 
 1 
 
 > 7 
 
 8 \ 
 
 ) 
 
 
 2 
 
 . 
 
 i 4 
 
 
 
 89" 
 
 .024 
 
 .020 
 
 .75 
 
 .30 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 93" 
 
 .027 
 
 .023 
 
 .74 
 
 .32 
 
 
 
 v 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 96" 
 
 .026 
 
 .023 
 
 .73 ■ 
 
 .31 
 
 
 
 \ 
 
 ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 100" 
 
 .031 
 
 .021 
 
 .73 
 
 .25 
 
 
 
 
 1^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 50" 
 
 .018 
 
 .021 
 
 .82 
 
 .32 
 
 
 
 
 t 4 
 
 E 
 
 
 
 \ 
 
 I ; 
 
 • 
 
 
 
 
 f 
 
 
 f 
 
 
 
 
 J 
 
 
 i 4 
 
 
 
 89" 
 
 .029 
 
 .024 
 
 .75 
 
 .26 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 f 
 
 
 
 
 ' 
 
 
 
 93" 
 
 .032 
 
 .024 
 
 .72 
 
 .26 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 ) 
 
 
 
 96" 
 
 .036 
 
 ,024 
 
 .71 
 
 .26 
 
 
 
 
 \ 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 / 
 
 
 
 100" 
 
 .038 
 
 .027 
 
 .68 
 
 .24 
 
 
 
 
 \ 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 / 
 
 
 
 50" 
 
 .018 
 
 .018 
 
 .81 
 
 .38 
 
 
 2 ; 
 
 
 E 
 
 
 
 i 
 
 
 i 
 
 
 
 E 
 
 f 
 
 7 
 
 ? 
 
 
 ■ 
 
 
 I 
 
 3 4 
 
 
 
 89" 
 
 .024 
 
 .019 
 
 .72 
 
 .34 
 
 
 1 
 
 
 
 
 
 
 y 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 1 
 
 
 
 98" 
 
 .024 
 
 .020 
 
 .68 
 
 .33 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 / 
 
 
 
 96" 
 
 .026 
 
 .022 
 
 .66 
 
 .32 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 / 
 
 
 
 100" 
 
 .031 
 
 .023 
 
 .58 
 
 .30 
 
 
 
 
 
 
 
 
 n 
 
 
 
 
 
 ^ 
 
 J— 
 
 
 
 
 
 
 ' 
 
 
 
 BO" 
 
 .028 
 
 .026 
 
 .82 
 
 .35 
 
 
 . 
 
 , 
 
 4 
 
 
 
 
 2 , 
 
 . t 
 
 I 
 
 
 
 ^ 
 
 
 
 } J 
 
 
 
 f 
 
 
 4 
 
 
 
 89" 
 
 .037 
 
 .028 
 
 .79 
 
 .33 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 93'. 
 
 .039 
 
 .028 
 
 .77" 
 
 .31 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 
 96" 
 
 .039 
 
 .028 
 
 .75 
 
 .31 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 100" 
 
 .056 
 
 .029 
 
 .71 
 
 .21 
 
 
 
 
 
 
 \ 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 60" 
 
 .024 
 
 .025 
 
 .70 
 
 .25 
 
 
 , 
 
 ; 
 
 i 
 
 
 
 
 
 
 I t 
 
 
 
 \ 
 
 ( 
 
 
 i 
 
 
 
 
 2 
 
 ) A 
 
 
 
 89" 
 
 .035 
 
 .029 
 
 .63 
 
 .22 
 
 
 
 \ 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 93" 
 
 .038 
 
 .029 
 
 .61 
 
 .21 
 
 
 
 
 \ 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 96- 
 
 .042 
 
 .029 
 
 .59 
 
 .22 
 
 
 
 
 \ 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 100" 
 
 .046 
 
 .032 
 
 .57 
 
 .21 
 
 
 
 
 
 s 
 
 
 
 
 ' 
 
 
 
 
 
 / 
 
 
 
 
 
 
 1 
 
 
 
 
 60" 
 
 .015 
 
 .025 
 
 .78 
 
 .29 
 
 
 . 
 
 ! 
 
 I 
 
 t 
 
 
 
 i 
 
 ; 
 
 1- 
 
 
 
 
 ( 
 
 1 
 
 \ 
 
 
 
 
 i 
 
 i . 
 
 ■ i 
 
 
 
 89" 
 
 .020 
 
 .026 
 
 .76 
 
 .24 
 
 
 \ 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 / 
 
 
 
 
 100" 
 
 .022 
 
 .027 
 
 .73 
 
 .22 
 
 
 > 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 / 
 
 
 
 
 50" 
 
 .023 
 
 .026 
 
 .73 
 
 .33 
 
 
 
 i 
 
 4 
 
 I 
 
 
 
 
 3 ^ 
 
 
 
 I 
 
 f 
 
 
 
 
 ? 
 
 
 
 I 3 4 
 
 
 
 89" 
 
 .034 
 
 .027 
 
 .71 
 
 .S 
 
 
 
 \ 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 ■ T ■ 
 
 
 
 100" 
 
 .044 
 
 .032 
 
 .70 
 
 .23 
 
 
 
 
 \ 
 
 ^ 
 
 
 
 
 > 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 1 
 
 
 
 FIG. 106.— PROGRESSIVE ANALYSIS OF A 50-TON BASIC OPEN-HEARTH HEAT 
 
 Note. — In "Minima and Maxima" figures for phosphorus and sulphur show hundredths 
 per cent; figures for manganese and silicon show tenths per cent; the carbons remaining 
 constant throughout are not shown. 
 
 making steel castings, so the time of teeming will be usually 
 somewhat longer than when ingots are cast, and the effects 
 studied will be correspondingly increased. 
 
 ^ The Foundry, September, 1909, page 34. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 139 
 
 Acid Open-hearth Practice 
 
 Acid open-hearth practice is in many respects similar to basic, 
 but the operations are shorter because : (1) a much larger pro- 
 portion of steel scrap is commonly used; (2) phosphorus is not 
 removed; and (3) no fluxes are added, except in rare instances, 
 when a little silica is charged at the beginning to prevent iron 
 oxide cutting the lining. 
 
 Chemistry. — The chemistry of the acid process is much sim- 
 pler, because neither phosphorus nor sulphur is removed; there- 
 fore it is necessary to start with pig iron and scrap low in both of 
 these elements. The progress of the operation is shown in Fig. 106. 
 This amount of silicon in the charge will determine the amount of 
 slag made, provided manganese is low; therefore low silicon pig 
 iron is preferred.^ The manganese is usually low also because: 
 (1) this element is costly; (2) it increases the amount of slag made; 
 (3) it forms a base which requires silica for fluxing it, thereby 
 cutting the lining; and (4) it increases the waste, since all the 
 manganese burned represents a loss in weight of metal purchased. 
 The manganese, as well as the silicon, in the bath is usually re- 
 duced to only a trace by the time the charge is melted. H, H. 
 Campbell has shown that the open-hearth slag at the end of the 
 acid operation automatically adjusts itself to contain about 46 
 per cent, of bases (FeO+MnO), the remainder being principally 
 silica, and that this ratio remains almost the same even when very 
 varying amounts of iron oxide are added. 
 
 Melting for Mild Steel. — In America approximately one-third 
 of the acid open-hearth steel made is used for castings containing 
 most frequently not very far from 0.20 per cent, carbon. Struc- 
 tural steel, with about the same analysis in carbon, forms an- 
 other large fraction of the annual tonnage. In this type of prac- 
 tice, the charge is so constituted of pig, scrap and ore that it wfll 
 contain when melted, only traces of manganese and silicon, and 
 not much more than 0.35 per cent, carbon. Where pig iron is 
 cheaper than scrap, the carbon after melting may be higher than 
 this, but this increases the length of the operation. The silicon 
 should be all oxidized in order that the slag may be fully formed 
 and cutting of the lining avoided. The analysis of the slag when 
 melted is practically the same as at the end of the operation, 
 provided it is always maintained in a fluid condition so it will 
 
 ^ See analysis of pig iron, page S, 
 
140 
 
 THE METALLURGY OF IRON AND STEEL 
 
 
 
 
 
 
 
 Diagram Showing Variations 
 
 in Composition 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 oia normal Aciq 
 Open Hearth Heat 
 
 
 
 
 
 
 
 
 
 
 ] 
 
 [eta^Blj 
 
 sltli 
 
 g 
 
 
 
 
 
 
 
 Meta 
 
 sM 
 
 elte 
 
 I 
 
 
 
 
 ' 
 
 
 
 Con) 
 
 men 
 
 •ed 
 
 Finjsbedl 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Heat 
 
 ' 
 
 
 
 01 
 
 iirgi 
 Ai 
 
 
 Cbs: 
 P 
 
 eing! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Tajp/ped 
 
 f 
 
 
 
 
 In 
 
 i 1 
 
 HalCarbd 
 
 12 
 a 
 
 09 
 
 4 
 
 02 
 
 4. 
 
 17 
 
 4. 
 
 12 
 
 4 
 
 17 
 
 5, 
 
 02 
 
 G 
 
 17 
 
 & 
 
 32 
 
 r 
 
 
 .tt) 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^0 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Mn. 
 
 .71 
 
 
 i'iO 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .1)0 
 .47 
 
 
 Inlt 
 
 al M nga; 
 
 lese 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 In^ 
 
 :Ial Sllico 
 
 a 
 
 
 T 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .40 
 
 
 
 
 
 
 , 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ll 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Bl. 
 
 81 
 
 .BO 
 
 
 
 
 
 
 11 
 
 
 
 1 
 
 
 ^' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 •he 
 
 >i^ 
 
 -— 
 
 -^ 
 
 s ff^ 
 
 i* - 
 
 
 
 
 1 r 
 
 lO. 
 
 23 
 
 .UO 
 
 
 
 
 
 
 \\ 
 
 
 
 
 
 
 
 
 
 
 .^ 
 
 >^'^ 
 
 Vi 
 
 ^-^ 
 
 n. 
 
 .^^ 
 
 T ' 
 
 
 
 
 
 
 
 t 
 
 
 11 
 
 
 
 
 
 
 
 
 
 
 
 
 -^tf^ 
 
 
 ;^ 
 
 1 
 
 
 
 ^0 
 
 
 
 
 p. 
 
 DO 
 
 
 Ij 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 
 
 .090 
 
 
 
 
 CiO 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 a 
 P. 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■a 
 
 i5 
 
 e^ 
 
 
 ,080 
 
 
 
 
 ? 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 cJ 
 
 Iff 
 
 a 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 .070 
 
 
 
 
 s 
 
 
 
 M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .000 
 
 
 
 
 1 
 
 
 
 ll 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .050 
 
 
 
 
 a 
 
 
 
 ll 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 ^ 
 
 
 
 ll 
 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 < 
 
 
 
 
 .010 
 
 
 
 
 
 
 
 1 
 
 
 "^ 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 8.. 
 
 I3K 
 
 ■ft-f? 
 
 — 
 
 Tnlt 
 
 (il-Pl 
 
 
 
 
 y 
 
 -'- 
 
 
 
 ^- 
 
 ::=:. 
 
 ^'' 
 
 "'' 
 
 - 
 
 -- 
 
 
 ^_^ 
 
 "-■*- 
 
 ■•-• 
 
 1 
 
 
 
 
 .V2a 
 
 .024 
 .020 
 
 
 
 
 
 -~^ 
 
 ^ 
 
 ' 
 
 / 
 
 >< 
 
 
 — 
 
 
 
 y 
 
 \ 
 
 — 
 
 
 
 1 
 
 ^ 
 
 V. 
 
 129 
 
 
 "hi 
 
 Klal 
 
 inlp 
 
 mr 
 
 • 
 
 
 
 
 
 X 
 
 
 
 / 
 
 * 
 
 
 \ 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 
 
 1 
 
 
 
 
 ^0 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 .--. 
 
 
 1 
 
 
 
 
 FIG. 106 A. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 141 
 
 react rapidly with the carbon. Ore is added from time to time, 
 but this is almost immediately reduced: 
 
 3C+Fe203 = 2Fe + 3CO 
 
 C+FeA^SFeO + CO. 
 
 The amount of slag increases all the time, showing that the 
 reduction of iron oxide is not complete, and that silica from the 
 lining is constantly entering the slag to maintain a proportion 
 of about 50 per cent. Si02. The slag also takes some oxygen 
 from the furnace gases and transmits it to the carbon of the 
 bath. At no time, however, when the charge is properly worked, 
 is the slag long in a strongly oxidizing condition, in contra-dis- 
 tinction to the situation in the basic process. Recar- 
 burization more usually is affected while the metal is still in 
 
 FIG. 107. 
 
 the furnace, because the objections to this practice obtaining 
 in the basic process do not apply here, and there are certain 
 advantages such as it being more readily melted; being stirred 
 into the metal better during the transfer into the ladle, etc. 
 On the other hand, a little manganese is oxidized and wasted. 
 Melting for High-carbon Steel. — When making steel for tires, 
 springs, and other objects requiring higher carbon, two general 
 types of practice are followed, depending to some extent upon 
 the relative cost of pig iron and steel scrap. If scrap is cheap, 
 
142 
 
 THE METALLURGY OF IRON AND STEEL 
 
 we may regulate the charge so as to melt a low-carbon bath, and 
 then bring this up to the desired point by melting pigs of iron in 
 it just before adding the ferro-manganese, etc. This has the 
 advantage of insuring more thoroughly the removal of dissolved 
 oxygen from the metal, but makes the final analysis a little 
 more uncertain because a variable proportion of carbon is oxi- 
 dized in the furnace during the melting period. The tempera- 
 ture of the furnace also affects this variable. The second type 
 of practice is to melt to carbon only a little above the desired 
 proportion, and then boil this down in the usual way. In either 
 event the ferro-manganese and ferro-silicon are usually added 
 in the furnace. Sometimes, as in the basic process, anthracite 
 coal in bags is used instead of pig iron to bring up the carbon, the 
 bags being thrown into the stream as it runs into the casting 
 ladle from the furnace. 
 
 Loss, — The loss in the acid process will not be as large as in 
 the basic, because the pig iron and scrap charged are not so 
 impure, and because the amount of slag made and the amount of 
 iron oxidized and retained by the slag are not so large. The loss 
 will vary on an average from 3 to 5 per cent., so that the final 
 metal will weigh 95 to 97 per cent, of the weight of the charge. 
 The analysis of this difference is as shown below. 
 
 ANALYSIS OF CHARGE 
 (30 per cent, pig; 70 per cent, scrap) 
 
 
 Pig iron 
 per cent. 
 
 Scrap 
 per cent. 
 
 Average 
 per cent. 
 
 Loss 
 per cent. 
 
 Carbon 
 
 3.75 
 
 0.90 
 
 0.60 
 
 0.035 
 
 0.03 
 
 0.25 
 
 0.10 
 
 0.40 
 
 0.045 
 
 0.03 
 
 1.30 
 0.34 
 0.46 
 0.04 
 0.03 
 
 1.30 
 
 Silicon. 
 
 34 
 
 Manganese 
 
 46 
 
 Phosphorus 
 
 00 
 
 Sulphur 
 
 00 
 
 8 per cent, slag @ 22 per 
 cent. Fe = 
 
 1 76 
 
 Net loss 
 
 
 
 
 3 86 
 
 
 
 
 
 
 Weight of Slag. — The weight of slag made in the acid process 
 may be determined, according to the method of H. H. Campbell,* 
 
 * See page 275 of No. 2. Old edition. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 143 
 
 from the total amount of manganese in the furnace, which will 
 include that put in with the charge (including any that may have 
 been in the ore) and that added with the recarburizing. First we 
 subtract from this total the amount of manganese in the metal 
 tapped; the weight of the remainder is then divided by the per- 
 centage of the manganese in the slag, which gives the weight of 
 the slag. The amount of acid slag will depend primarily upon 
 the amount of silicon in the charge and will vary on the average 
 from 6 to 18 per cent, of the weight of the metal charged, about 
 three-fourths of,, which is formed during the melting period.* 
 
 Coal Burned. — The amount of fuel used per ton of steel made 
 in the acid open-hearth furnace will be perhaps 50 to 100 lb. less 
 than in the basic furnace, but will depend again upon many vary- 
 ing conditions, so that figures should be used only with caution. 
 (See pages 127 and 128). 
 
 Special Open-hearth Processes 
 
 The original pig-and-ore process is abandoned very largely 
 in America, because of the length of time required to burn off the 
 impurities unless they are diluted by steel, and because steel scrap 
 is more abundant than it is in Europe. It is difficult for a novice 
 to understand why the reactions are not more rapid in the open- 
 hearth furnace, when the entire purification of pig iron in the 
 puddling-furnace is accomplished in an hour and a half, including 
 melting; but the difference is due to the very shallow bath in the 
 puddling operation and its extensive contact with the fettling. 
 If we should attempt to purify under such strong oxidizing con- 
 ditions in the open-hearth furnace, the molten metal would boil 
 violently, because of the high temperature, and for the same 
 reason would also become so charged with oxygen as to be worth- 
 less. Even at the low temperature of the puddling-furnace, the 
 boiling is so violent as to increase the height of the bath, and this 
 action would be proportionately increased at the temperature of 
 the open-hearth furnace, which, at the end of an operation pro- 
 ducing dead-soft steel, will be about 1650° to 1700° C. (3002° to 
 3092° F). The increase in volume of the metal when carbon 
 monoxide gas is escaping from it may be likened to that of cham- 
 pagne when the drawing of the cork allows a rapid escape of gas. 
 There is another reason why the boil causes more of an increase 
 
 ' In the case of making structural steel (say 0.18 per cent, carbon). 
 
144 THE METALLURGY OF IRON AND STEEL 
 
 in the volume of the bath in the open-hearth furnace than in the 
 puddling-f urnace : in the latter, the carbon monoxide has only 
 molten metal and slag to bubble through, but in the open-hearth 
 process, where cold ore is added to the charge, it produces a 
 certain amount of chilling of the metal and slag adjacent to it, 
 and the gas having to bubble through this somewhat pasty 
 material causes a greater increase in its bulk. By the time the 
 puddling charge becomes pasty, the carbon is largely gone and 
 therefore there is not a violent action. 
 
 Various attempts have been made by different metallurgists 
 to adapt the open-hearth process to the use of all pig iron rapidly 
 oxidized by iron ore or other agencies, and this has led to the 
 Talbot and the Monell processes, each of which is carried on in a 
 single furnace, molten pig iron being acted upon by a highly 
 oxidized liquid slag, formed prior to the addition of the pig iron 
 in the Talbot process and coincident with it in the Monell process. 
 It has also led to the Duplex process, whereby a large proportion 
 of the oxidation is effected in an acid Bessemer converter, the 
 operation being completed on a basic hearth. The ingenuity of 
 the German manufacturers has also enabled them to use the pro- 
 cess in a slightly revised form, as already described. 
 
 Duplex Process, — In several large American plants the com- 
 bined Bessemer and basic open-hearth process is in operation, an 
 acid converter being used to oxidize the silicon, manganese, and 
 most of the carbon, while the phosphorus and the remainder of 
 the carbon is eliminated in a basic open-hearth furnace. In the 
 different localities there are different ways of carrying out this 
 combination, but these divide themselves into two general 
 methods: In one method, the metal is blown in the converter 
 until it is purified to the point where it is practically equivalent 
 to so much high-phosphorus, molten steel scrap, which is then 
 mixed with either liquid or solid pig iron in the open-hearth 
 furnace and worked as any ordinary pig-and-sgrap heat after 
 melting. In another, and more common, method, the pig iron is 
 blown in the converter until it contains about 1 per cent, or so of 
 carbon, and this product, with little or no additional pig iron, is 
 then dephosphorized and decarbonized in the open hearth furnace- 
 The advantages of the combined process are: it shortens the 
 open-hearth purification by sometimes much more than half the 
 usual period; it relieves the basic hearth of the presence of silica 
 which tends to steorify it. Both of these improvements have a 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 145 
 
 secondary advantage in increasing the life of the hearth. An 
 advantage from the Bessemer standpoint is that a poorer grade 
 of iron can be converted and yet a higher grade of steel produced. 
 Open-hearth railroad rails, which are frequently made by the 
 Duplex process in America, bring $2 per ton more than Bessemer 
 rails. Finally, the Duplex process makes the steel-maker inde- 
 pendent of the variable price of so-called "melting steel scrap." 
 
 In carrying out the process, large mixers with capacities of 
 250 and even 750 tons each are used as a reservoir for the liquid 
 pig iron: sometimes a little purification is carried on in these 
 furnaces. They supply pig iron to the converters, and to basic 
 furnaces as well, if desired. The blown metal is also frequently 
 poured into a mixer of its own, which serves as a source of supply 
 to the open-hearth furnaces, and again a little purification may 
 be effected in this mixer. The use of these mixers not only 
 makes each part of the process somewhat independent of delays 
 in the other part, but also corrects slight irregularities in analysis 
 of pig iron or blown metal, thus providing more uniform con- . 
 ditions iir each furnace, which greatly facilitates the operation 
 therein. Under proper conditions, an open-hearth furnace with 
 auxiliary converter and mixers, may make as many as 40 to 50 
 heats per week, as compared with 18 heats by the ordinary 
 method. In one large American plant, -the Duplex process is 
 employed in connection with Talbot open-hearth furnaces, which 
 are described in the following section. 
 
 The name "Duplex process^' is also given to a combination of 
 Bessemer or open-hearth furnaces with electric " super-purifiers/' 
 but this practice will be discussed under the head of electric 
 processes. 
 
 Talbot Process. — The Talbot process has a basic lining and 
 contains a charge above 200 tonp in some cases, as, for example, 
 at the Jones & Laughlin plant, in Pittsburg, Pa. As the bath of 
 metal is over 3 ft. deep, however, which is about twice that of 
 an ordinary bath, the furnace is only as large in other dimensions 
 as a 100-ton furnace. The tilting furnace is commonly used 
 in order that any desired quantity of metal or slag may be poured 
 out at will. The operation is continuous and the furnace is 
 drained of metal only once a week. After the- first charge has 
 been worked down to the desired percentage of carbon, the great 
 part of the slag is poured off, and then about one-third of the 
 steel is poured into the ladle, recarburized, and teemed into the 
 
 10 
 
146 THE METALLURGY OF IRON AND STEEL 
 
 ingot molds in the usual way. To the charge of metal left ih 
 the bath is now added iron ore and limestone to produce a basic 
 and highly oxidized slag, and through this slag is then poured an 
 amount of pig iron equal to the steel removed. The reaction 
 between the impurities in the pig iron and the iron oxide in the 
 slag is very vigorous, but does not cause a frothing or foaming, 
 because all the materials are in the liquid form and the gas 
 bubbles through them without great difficulty; — however much 
 of the slag runs out of the furnace. 
 
 The oxidation is so rapid that the silicon and manganese are 
 said to be oxidized almost immediately, together with most of the 
 phosphorus; then the carbon is worked off in the usual way, using 
 more ore and limestone, if necessary. The temperature is low at 
 first, in order that the phosphorus may be more readily burned. 
 At the end of about three to six hours, the bath has again become 
 purified, and another cast removed, the whole operation being 
 then repeated. The yield of steel is 106 to 108 per cent, of the 
 weight of the pig iron charged, because of the large amount of 
 •iron reduced from the ore by the impurities in the pig iron. 
 
 3C+Fe203 = 2Fe + 3CO (absorbs 108,120 calories). 
 
 6P + 5Fe2O3 = 10Fe + 3P2O5 (evolves 117, 900 calories), et. al. 
 
 The advantages of the process are: We obtain three or four 
 heats of 75 tons each in 24 hours without the use of steel scrap; 
 the yeld is large (though this advantage is somewhat neutralized 
 by the cost of the iron ore used) ; and the temperature of the final 
 metal can be more easily controlled. The disadvantages of the 
 process are: The very large cost of furnaces and the slightly 
 higher cost for repairs. 
 
 Monell Process. — ^A. Monell, when metallurgist of the Carnegie 
 Steel Company, developed a pig-and-ore process in which highly 
 heated oxidizing and slag-making materials were made to react 
 with the impurities in molten pig iron without the necessity of 
 having a reservoir of metal into which to pour it. Upon a basic 
 hearth he heats limestone and a relatively large amount of ore, 
 or other form of iron oxide, until they become pasty, and then 
 pours molten pig iron, equivalent to the capacity of the furnace, 
 upon it. Oxide up to 25 per cent, of the weight of the pig iron 
 charge is used. The temperature of the bath is necessarily low, 
 since pig iron direct from a blast furnace or from a mixer will not 
 be more than 200° or 300° C. above its melting-point, and there- 
 fore the phosphorus will be oxidized very rapidly. The slag 
 
THE OPEN HEARTH OR SIEMENS-MARTIN PROCESS 147 
 
 foams up and pours out of a special slag-notch; in an hour the 
 bath is free from 90 per cent, of the phosphorus, most of the sili- 
 con and manganese, and all but about 2 per cent, of carbon. 
 The operation is then continued in the usual way to eliminate the 
 carbon, and the metal is tapped when this has been reduced to 
 the desired point. The total time is the same as an ordinary pig 
 and scrap process. The American rights to the process are 
 owned by the Carnegie Steel Company and they are operating it 
 at many of their furnaces. No details are known generally, but 
 it is to be presumed that the results are favorable. The apparent 
 disadvantages ^of the process are excessive cutting of thp hearth 
 and a heavy loss of iron in the rich slag which flows off at the 
 beginning of the operation. The Monell process has been used 
 successfully in England, with pig iron containing up to 2 per cent, 
 of phosphorus. 
 
 Processes in Two Open-hearth Furnaces, — At a low tempera- 
 ture phosphorus is very easily oxidized and absorbed by a basic 
 slag, even in the presence of carbon, but when the heat is high the 
 oxidation of phosphorus is hindered by the carbon, for the reason 
 already explained — that the aflSnity of carbon for oxygen in- 
 creases more rapidly with the temperature than the affinity of the 
 other elements in the bath. We could obtain the desired con- 
 ditions in -the open-hearth process, but the operation would be 
 extremely slow at this low heat, and the carbon would pass away 
 slowly. These difficulties have been met by the Campbell No. 2 
 and Bertrand-Thiel processes, the former of which was developed 
 at Steelton, Pa., and the latter at Kladno, Bohemia. 
 
 Campbell No. • 1 Process. — The pig-and-ore process using 
 molten metal has long been in operation in the Campbell tilting 
 furnace, and the frothing of the bath is taken care of by tipping 
 the furnace backward so. that no slag or metal will pour out of the 
 door, though a large amount of the former flows from the slag- 
 hole between the ends of the furnace and the ports. The opera- 
 tion is continued in this way for two or three hours, since, as 
 already noted, the furnace can be tipped without cutting off the 
 supply of gas and air, and the yield of steel is 104 to 106 per cent, 
 of the pig iron charged. 
 ^ Campbell No. 2 Process.- — At the same plant there is also a 
 combination process in which the charge, consisting of all pig 
 iron, or of pig iron^ and scrap, is placed in a basic open-hearth 
 
 ^ The pig iron charged may be either in the liquid or solid state. 
 
148 THE METALLURGY OF IRON AND STEEL 
 
 furnace and the purification carried on at a low temperature until 
 almost all the silicon and phosphorus and part of the sulphur and 
 carbon are eliminated. The bath is then tapped from the basic 
 furnace and poured into an acid-lined furnace, care being taken 
 that none of the basic slag goes with it. At this period the metal 
 is low in phosphorus and sulphur, and contains about the same 
 amount of carbon that a cold charge would have contained as 
 soon as melted. The conditions are therefore the same as if 
 low-phosphorus low-sulphur material had been charged into an 
 acid furnace and melted there, and the process is now continued 
 at a higher temperature, in the usual way to make acid open- 
 hearth steel. The disadvantage of this process is that the trans- 
 ferring of molten metal from one furnace to another is not an 
 easy matter, nor, in fact, is it possible with the arrangements in 
 many plants. 
 
 Bertrand-Thiel Process. Hoesch Modification, — ^The Hoesch 
 process is now in use in some of the German plants, where it is 
 said to give very satisfactory results. Instead of using two fur- 
 naces, as in the original Bertrand-Thiel process, and pouring from 
 one to the other, it -uses one furnace but pours the bath into a 
 ladle after. the first period of purification; the purpose of this is to 
 separate it from the highly phosphorized slag; when this is effected, 
 the metal is poured back into the same furnace and the purifica- 
 tion completed. In the first period, all the silicon and manganese, 
 most of the phosphorus and about one-half of the carbon, are re- 
 moved; in the second period, the remaining phosphorus, carbon 
 and some additional sulphur are eliminated. The slag separated 
 after the first period contains 20 to 25 per cent, of -PjOg, where- 
 fore it finds a ready sale as agricultural fertilizer. The chemical 
 history of a charge is given on p. 149. 
 
 Comparison of Purification Processes 
 
 Acid with Basic Open-hearth, — Acid open-hearth steel is 
 believed by engineers to be better than basic, and is usually 
 specified for all important parts of structures, although not so 
 rigidly to-day as a few years ago. This is in spite of the fact that 
 phosphorus and sulphur, two very harmful elements, are lower 
 in the basic steel. The basic process is less expensive than the 
 acid, because high phosphorus pig iron and scrap are cheap, and 
 the lower cost of materials used more than balances the greater 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 149 
 
 n 
 . a 
 
 o ^ 
 
 CO 
 
 td <=^ 
 
 u "^ 
 W I 
 
 w 
 
 pq ^ 
 
 H jT 
 O '^ 
 
 < ^ 
 
 u ^ 
 
 < a 
 
 O <£> 
 
 r-i 
 
 Pi; 
 o 
 
 03 
 
 
 « 
 
 
 
 
 
 
 
 
 O 
 
 1 
 
 
 • fH 
 
 1 
 
 
 1 
 
 
 - 
 
 (N 00 -* 
 
 <N r4 ,-1 
 
 f-l t-1 i-H 
 
 
 
 
 Oi ^ CO 
 
 
 OQ 
 
 
 
 CO <N CO 
 
 
 
 
 O i-( 1-t 
 
 -*3 . 
 
 
 
 
 • • • 
 
 
 
 
 
 o o o 
 
 4^ 
 
 
 
 
 
 ^ 
 
 ■o 
 
 
 
 if3 CO CO 
 
 h 
 
 o 
 
 
 
 GO (N 11 
 
 a> 
 
 N 
 
 
 
 
 Ah 
 
 PLh 
 
 
 
 (N IN N 
 
 1 
 to 
 
 
 
 
 (N (N (N 
 
 1^ 
 
 
 
 
 o3 
 
 
 
 (N O O 
 
 -S 
 
 
 
 CO o o 
 
 CQ 
 
 
 
 ^ Tt< Tt* 
 
 ^ 
 
 
 
 CO 00 CO 
 
 ■^ 
 
 
 
 lO '^ 00 
 
 ^ 
 
 d 
 
 
 
 tH Ift 00 
 
 ^ 
 
 
 
 
 ■* Tt< -* 
 
 o 
 
 
 
 CO X o 
 
 
 
 
 
 CO CO CO 
 
 
 <1 
 
 
 
 
 
 o 
 
 
 
 I— 1 1— I 1— I 
 
 
 
 
 
 1 
 
 
 
 CO CO CO 
 
 
 
 
 O OS OS 
 
 
 
 
 
 ifS -tjj CO 
 
 o 
 
 C3 
 
 
 
 lO CO t* 
 
 
 
 
 W (N CO 
 
 
 f=; 
 
 
 
 O l> ^ 
 
 
 
 ■ i-H 
 
 
 
 (N (N CO 00 N 
 
 -1^ 
 
 CQ 
 
 CC O OS OS M 
 
 ,-H »-< o O O 
 
 s 
 
 Ih 
 
 
 O O O O O 
 
 
 (M 
 
 
 02 
 
 d^^^^ 
 
 fl 
 
 o t^ cq IN Tj< 
 
 cd 
 
 OS iH (N (N CO 
 
 -fi 
 
 s 
 
 
 
 o o d d o 
 
 
 ?D OS t* r- CD 
 
 to 
 
 Pm 
 
 00 lO Tt< CO (N 
 
 •s 
 
 
 »-( o o o o 
 
 t>» 
 
 
 
 1 
 
 
 GO t^ O IC CO 
 
 o 
 
 (M '^ OS CD Tt< 
 
 
 
 00 (N tH .-1 i-t 
 
 
 
 'tS 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 u 
 
 •<r» 
 
 
 
 
 - 
 
 
 
 S 
 
 
 
 
 
 
 
 E 
 
 f4 
 
 H 
 
 !i- 
 
 > c 
 
 J c 
 
 c 
 
 w: 
 
 
 
 
 ^ 1-r ,-H CO -^ O 
 
 
 
 ^£ 
 
 
 ( r^ 
 
 1-H 
 
 M 
 
 
 CO 
 00 
 
 
 3 a g 
 
 fH CO 
 
 CO O 
 
 N 00 
 
 Q 
 
 , 
 
 ^ 
 
 m 
 
 a> 
 
 fl) 
 
 
 d 
 
 
 03 
 
 ,n 
 
 bo 
 
 
 a 
 
 o 
 
 03 
 
 a 
 
 "^ 
 
 
 N 
 
 ^ 
 
 (N 
 
 (N O 
 
 w -* 
 
 (N 
 
 1-i 
 
 1—1 
 
 CO 
 
 iH ,-1 
 
 CO (N 
 1-1 iH 
 
 O 
 
 iH 
 
 uo 
 CO 
 1-4 
 
 i 
 
 CO lO 
 
 §?5 
 
 CO rt< 
 1-H -«*< 
 CO CO. 
 
 o 
 
 o 
 
 O 
 
 o o 
 
 O O 
 
 (N 
 
 g 
 
 s 
 
 lO lO 
 
 28 
 
 CO 
 
 CO 
 
 lO 
 
 iO U3 
 
 iO iO 
 
 § 
 
 g 
 
 § 
 
 Sg 
 
 W (N 
 1-1 OS 
 
 CO 
 
 CO 
 
 CO 
 
 CO lO 
 
 CO lO 
 
 s 
 
 g 
 
 00 
 
 g§ 
 
 OS GO 
 1-1 <N 
 
 CO 
 CO 
 
 lO 
 
 CO 
 
 CO 
 
 CO lO 
 
 CD CD 
 
 1-1 
 
 o 
 o 
 
 (N 
 CO 
 
 <N (N 
 
 O O 
 
 O OS 
 
 CO 
 
 CO 
 
 N 
 
 cfl cq 
 
 (N iH 
 
 OS 
 
 GO 
 
 o 
 
 00 
 00 
 
 Oi (N 
 
 1-1 CO 
 
 l> lO 
 CO CI 
 
 CO 
 1-t 
 
 1~i 
 
 00 
 
 O CO 
 
 1—1 
 
 l> o 
 
 ft 
 
 CO 
 
 00 
 
 CO 
 00 
 
 CO CO 
 1-H W 
 
 §§ 
 
 OS 
 
 1-H 
 
 lO 
 
 1—1 
 
 •^ CO 
 
 1-1 1-1 
 
 r-t 1-H 
 
 o 
 o 
 
 r-t 
 
 OS 
 
 g 
 
 o 
 
 § 
 
 O 00 
 
 1^ CD 
 
 o o 
 
 o 
 
 o 
 
 o 
 
 o o 
 
 o o 
 
 ^^^ 
 
 ^^ 
 
 e^ 
 
 s 
 
 ?5 
 
 CO 
 (N 
 
 CO t- 
 
 
 o 
 
 o 
 
 o 
 
 o o 
 
 O O 
 
 OS 
 
 o 
 
 U5 
 
 o 
 
 o 
 
 CO CO 
 O C 
 
 §S 
 
 o 
 
 o 
 
 o 
 
 O O 
 
 o o 
 
 00 
 CO 
 
 § 
 
 OS 
 
 o 
 
 feg 
 
 ss 
 
 o 
 
 o 
 
 o 
 
 o o 
 
 o o 
 
 
 'Sod 
 
 5 "3 '-H 
 
 O lO IC O lO 
 
 CO n< o iH ^ 
 
 CQ 
 
 CO 'i* ^ -^ lo lO »o 
 
 o 
 
 I 
 
 
 t3 
 
 a 
 
 03 
 bj3 
 
 bXI 
 03 
 
150 THE METALLURGY OF IRON AND STEEL 
 
 cost of the basic lining and the lime additions and the circum- 
 stance that the acid furnace has a higher output because the heats 
 are shorter. Acid steel is preferred by many, however, for the 
 following reasons: 
 
 (a) A basic slag will dissolve silicon from the metal; we there- 
 fore recarburize in the basic process by adding the recarburizer 
 to the steel after it has left the furnace, instead of in the furnace, 
 as we do in the acid process. Should any basic slag be carried 
 over with the metal, however, which is liable to happen, there is 
 the danger that the ingots will be too low in silicon. They are 
 then impregnated with gas bubbles, or "blow hotes." Indeed 
 the last part of a heat of basic steel teemed from a ladle will be 
 lower in silicon and manganese and higher in phosphorus and 
 sulphur, as already indicated in discussing the basic process, 
 A goodly proportion of the acid open-hearth steel made in 
 America goes into steel castings, which receive no mechanical 
 treatment before going into service. In such products, of course, 
 the presence of blowholes is most harmful; hence the predomi- 
 nant use of the acid process, although this predominance in point 
 of tonnage is now being lost. Even the acid Bessemer process, 
 modified as explained under the chapter on Founding, gives a 
 product for castings, which, from the viewpoint of freedom from 
 blowholes, is preferred to the basic open-hearth. 
 
 (h) Moreover, the recarburizer does not mix with the steel as 
 well if it is not added in the furnace, because of the stirring in 
 pouring, and this sometimes produces irregularities. 
 
 (c) A basic slag is more highly oxidized than an acid slag; 
 therefore the metal at the end of the operation is more highly 
 charged with oxygen. For this reason we add a larger amount of 
 manganese in the recarburizer, but the remedy is never quite as 
 good as prevention. 
 
 (d) Since we cannot remove the phosphorus from the bath in 
 the acid process, it is necessary to use only picked iron and scrap, 
 whereas, in the basic process, good steel can be made from almost 
 any quality of material. Many engineers believe, however, that 
 a better grade of steel results from using the picked material. 
 
 (e) It occasionally happens in the basic process, after the 
 phosphorus has all been oxidized in the slag and the operation is 
 ended, a good deal may get back into the metal again. This is 
 especially liable to happen when basic slag is carried over into 
 the ladle before the recarburizer is all in. If this occurs, and if 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 151 
 
 the bath is very hot, a reaction may take place between the 
 basic slag and the acid lining of the ladle. In this way the slag 
 will be enriched in silica and phosphorus will be forced out of it 
 into the metal (see p. 136). 
 
 Basic Open-hearth with Bessemer. — Basic open-hearth steel 
 is better than Bessemer steel. The reasons for this are believed 
 to be : 
 
 (a) The open-hearth process being slower, more attention and 
 care can be given to each detail. This is particularly true of the 
 ending of the process, for if the Bessemer process is continued 
 only a second or so too long, the bath is highly charged with 
 oxygen, to its detriment, and even under normal circumstances 
 there is more oxygen in the metal at the end of the Bessemer 
 process than at the end of the basic open-hearth, because there 
 has been so intimate a mixture between metal and air. 
 
 (6) For the sarne reason the Bessemer metal is believed to con- 
 tain more nitrogen ^ and hydrogen, ^ which are thought to be 
 deleterious. 
 
 (c) The beat of the Bessemer process is dependent upon the 
 impurities in the pig iron, and especially upon the amount of the 
 silicon, and can be controlled only to a limited extent by methods 
 that are not perfect in their operation. Furthermore, the heat 
 is regulated according to the judgment of the operator and his 
 skill in estimating the temperature of the flame. Irregularities 
 therefore result at times, and these produce an effect on the steel, 
 because the temperature at which the ingots are cast should be 
 neither too high nor too low. It is true that the temperature of 
 the open-hearth steel is also regulated by the judgment of the 
 operator, but more time is afforded for exercising this judgment 
 and for controlling the heat. 
 
 (d) In the Bessemer process we get rid of all the carbon first, 
 and then recarburize to the desired point. In the open-hearth 
 process we may stop the operation at any desired amount of 
 carbon, and then recarburize only a small amount. Therefore 
 the open-hearth has the advantage of greater homogeneity when 
 maiking high-carbon steel, since a very large amount of recar- 
 burizer may not distribute itself uniformly. 
 
 In order to produce the best quality of steel, it must be cast 
 into ingot molds within a certain limited range of temperature, 
 which varies according* to the amount of carbon, etc., that it con- 
 
 ^ The hydrogen is said to come from the moisture in the blast. 
 
152 THE METALLURGY OF IRON AND STEEL 
 
 tains. Therefore, in casting the very large heats of the open- 
 hearth process, the ingots must be very large, else the first one 
 will be too hot and the last one too cold for the best results. On 
 the other hand, if the ingots are large, segregation is liable to be 
 excessive (see page 166). 
 
 For over fifteen years the Bessemer process has been fighting 
 a losing battle to maintain its supremacy against the inroads of 
 the basic open-hearth, which have been possible because of the 
 increasing cost of Bessemer pig iron, due to the exhaustion of the 
 low phosphorus ores. The Bessemer process will probably be 
 relegated in future to the place chiefly of an adjunct to the open- 
 hearth or electric furnaces in the "Duplex processes," unless large 
 bodies of Bessemer ores are discovered. 
 
 On account of its ability to make low carbon steel more readily 
 than the basic open-hearth, the Bessemer process has a hold on 
 the wire and welded steel-pipe industry, although even here the 
 open-hearth process has encroached. For rolling very thin 
 plates, for tinplate, etc., we want a metal relatively high in phos- 
 phorus, and therefore the Bessemer process is largely used here, 
 although in some cases ferrophosphorus is being added to basic 
 open-hearth metal to accomplish the same result. The reason 
 phosphorus is desired is because the plates are rolled very thin by 
 doubling them up and putting several thicknesses through the 
 rolls at the same time. Low phosphorus metal welds together 
 too much under these circumstances. 
 
 The chief requisites of railroad rails are lack of brittleness and 
 ability to withstand wear. The Bessemer process is able to pro- 
 vide such a material, and it works so well in conjunction with the 
 rapid, continuous operation of the rail-rolling- mill that it has a 
 decided advantage. It produces a small tonnage of ingots at 
 frequent intervals (say 15 tons every 7 minutes), while the open- 
 hearth process provides a large tonnage of ingots, which may 
 come at irregular intervals and thus alternately delay and over- 
 crowd the rail-mill operations. But notwithstanding these 
 advantages, an increasing tonnage of basic open-hearth rails 
 is made every year in the United States, and now the majority of 
 rails of large sizes are made by the open-hearth process, in which 
 we include the Bessemer-open-hearth Duplex process. The 
 lower phosphorus in the basic steel enables us to get a tough rail 
 with higher proportions of carbon, and this higher carbon gives 
 a steel more resistant to wearing down by abrasion. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 153 
 TABLE XIX.— STEEL PRODUCTION IN METRIC TONS,i 1912 
 
 X 
 
 United States 
 
 Germany 
 
 Acid converter 
 
 10,493,147 
 
 187,179 
 
 Basic converter 
 
 9,794,300 
 
 
 
 
 Total converter 
 
 10,493,147 
 
 9,981,479 
 
 Acid open-hearth 
 
 1,157,449 
 19,955,766 
 
 (a)295,256 
 
 Basic open-hearth 
 
 (a)6,871,896 
 
 
 
 Total open-heart 
 
 21,113,215 
 
 7,167,152 
 
 Crucible 
 
 123,461 
 21,501 
 
 79,190 
 
 Electric, etc 
 
 74,177 
 
 
 
 Total 
 
 31,751,324 
 
 17,301,998 
 
 
 
 (a) Includes all "Direct Castings." 
 
 TABLE XX.— MAKE OF ACID AND BASIC STEEL, 1912 
 
 
 Acid 
 
 Basic 
 
 
 Tons 
 
 Per cent. 
 
 Tons 
 
 Per cent. 
 
 United States 
 
 Germany 
 
 11,650,596 
 535,802 
 
 37 
 3 
 
 19,955,766 
 16,666,196 
 
 63 
 97 
 
 
 
 Total 
 
 12,186,398 
 
 
 36,621,962 
 
 ^ 
 
 
 
 
 Crucible Steel with Others, — Crucible steel is the most expensive 
 of all and costs at least three times as much as the next in price — ■ 
 acid open-hearth steel. It is also the best quality of steel manu- 
 factured,^ and for very severe service, such as the points and 
 edges of cutting tools, the highest grades of springs, armor- 
 piercing projectiles, etc., it should always be employed. The 
 reason for its superiority is believed to be that it is manufactured 
 in a vessel which excludes the air and furnace gases, and is there- 
 fore freer from oxygen, hydrogen and nitrogen. Perhaps the 
 fact that the process is in some ways under a little better control 
 than any of the others, and receives more care, on account of 
 
 ^ From "Mineral Industry." 
 
 ^ Except electrically purified steel, discussed in Chap. XVII. 
 
154 THE METALLURGY OF IRON AND STEEL 
 
 being manufactured in small units, assists in raising its grade. 
 Crucible steels are usually higher in carbon than Bessemer and 
 open-hearth steels, because the special service to which the cru- 
 cible steels are adapted is usually one requiring steel that can 
 be hardened and tempered — ^for example, cutting tools, springs, 
 etc., and only the high-carbon steels are capable of this hardening 
 and tempering. This is not an essential characteristic of the 
 product, but is most often the case. 
 
 Wrought Iron with Low-carhon Steel, — ^Wrought iron costs 
 from 10 to 20 per cent, more than the cheapest steel. Its claim 
 to superiority over dead-soft steel consists in its purity and 
 the presence in it of slag. Just how much advantage the slag 
 is has never been proven; it. gives the metal a fibrous structure 
 which, perhaps, increases its toughness and its resistance to 
 breaking under bending or under a sudden blow or shock. Some 
 think that the slag also assists in the welding of the material, but 
 this is doubtful, and it is probable that the easy weldability of 
 wrought iron is due alone to its being low in carbon. Some also 
 believe that the slag assists the metal in resisting corrosion; hence 
 one reason for the preference of some engineers for wrought-iron 
 pipe for boilers and other purposes. An advantage of wrought 
 iron in this connection is its rough surface to which paint or other 
 protective coatings will adhere more firmly than to the compara- 
 tively smooth surface of steel. Nevertheless the evidence goes 
 to show that properly made steel corrodes very little, if any, 
 more than wrought iron, especially in boilers, pipe, and other 
 articles which cannot be coated. (See page 418.) 
 
 The properties of wrought iron are the nearest to those of pure 
 iron of any commercial material, notwithstanding its slag. This 
 is because the slag is mechanically mingled with the metal and 
 does not alter some of its chemical and physical characteristics. 
 Therefore wrought ir,on is preferred for electrical conductivity 
 purposes and as a metal with high magnetic power, for armatures 
 of electromagnets, etc. When under strain greater than it can 
 withstand wrought iron stretches more "uniformly over its entire 
 length than steel, although the elongation closer to the point of 
 fracture may not be so great. Much wrought iron is made to- 
 day by the bundling and rerolling of wrought iron scrap, as 
 described on page 68. This makes a much inferior product, 
 the bundles sometimes even containing steel rolled into the pile. 
 The advantages I have mentioned, the conservatism of engineers, 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 155 
 
 and the capital previously invested in\puddling furnaces are the 
 chief factors in keeping alive the manufacture of wrought iron. 
 It was freely predicted that the invention of the Bessemer and 
 open-hearth processes would bring about the extinction of the 
 puddling process, but these prophecies have never been fulfilled, 
 although the importance of wrought iron has waned very greatly 
 in fifty years. 
 
 Summary, — In order of expense and of quality the different 
 steels are arranged as follows: (1) Crucible; (2) acid open-hearth, 
 (3) basic open-hearth, and (4) Bessemer. The amounts of the 
 different kinds made in America to-day are shown in Table XIX, 
 Though I have not made a direct comparison between certain of 
 the classes, e.g., acid open-hearth with Bessemer, their relations 
 may be easily learned by collating the other comparisons given. 
 
 Distinguishing Between the Different Prodttcts 
 
 Low-carbon steel pipe, merchant bars, horseshoe blanks, 
 etc., sometimes masquerade under the name of wrought-iron; 
 high-carbon open-hearth and Bessemer-steel merchant bars, 
 tool blanks, etc., sometimes masquerade as "crucible steel," or 
 perhaps "cast steel," which is the trade name for crucible steel; 
 other deceptions are not unknown; indeed, even malleable cast 
 iron is sold oftentimes as "steel castings." It is therefore 
 important for engineers to understand the essential differences 
 between these materials, although care in the wording of contracts 
 and specifications should be the important consideration and 
 should precede watchfulness over the products. The definitions 
 of iron and steel materials are in such a confused and unsettled 
 condition that it does not do to rely upon them at all, especially 
 where a lawsuit may be involved; and contracts in clear, simple 
 language, free from legal and metallurgical phraseology, are the 
 best safeguards. But even where it is entirely plain what 
 material is called for, there is always a temptation to substitute 
 steel for wrought iron, Bessemer for open-hearth, basic for acid, 
 and Bessemer or open-hearth for crucible steel. In case one 
 such substitution is suspected, there are means by which the 
 material may be tested, aside from its strength and ductility, 
 which may or may not be in the contract. The tests are some- 
 what delicate and usually require the judgment and experience 
 of an expert — one who has standard samples of the different 
 grades of material for comparison, because the details of manu- 
 
156 THE METALLURGY OF IRON AND STEEL 
 
 f acture vary from district .to district, and still more so with the 
 purposes for which the products are to be used. 
 
 Wrought iron may be distinguished from low-carbon steel 
 by the fact that it contains slag.^ Usually, there is more than 
 1 per cent, of slag in iron and less than 0.5 per cent, of slag (in- 
 cluding metallic oxides) in steel. The slag may be determined 
 either by chemical or microscopical analysis. Normal wrought 
 iron is practically free from manganese, while normal Bessemer 
 and open-hearth steel will contain 0.4 per cent, or more. 
 Wrought iron generally contains more than 0.1 per cent, phos- 
 phorus, while good steel should never do so. Charcoal iron has 
 less slag, less sulphur and less phosphorus than puddled iron. 
 
 Crucible 'steel normally has less than 0.4 per cent, manganese 
 and more than 0.2 per cent, silicon, while open-hearth and 
 Bessemer steels normally have more than 0.4 per cent, manga- 
 nese and less than 0.2 per cent, silicon, Iij. the case of steel 
 castings, however, this rule for silicon does not apply, as Besse- 
 mer and open-hearth steel castings are sometimes as high as 0.6 
 per cent, in silicon. It is possible — ^but not usual — to make 
 both Bessemer and open-hearth steels low in manganese. Be 
 it noted that when crucible steel is ordered low in carbon, there 
 is a much greater temptation to substitute another steel for it. 
 
 Acid open-hearth steel may be distinguished from basic open- 
 hearth steel by its being normally higher in silicon, and usually 
 in phosphorus also, but lower in manganese'. The same differ- 
 ences exist between acid and basic Bessemer steel. 
 
 Basic open-hearth steel may be distinguished from Bessemer 
 steel by its lower manganese, silicon, phosphorus and (generally) 
 sulphur, as well as by the fact that it dissolves much more slowly 
 in dilute hydrochloric and sulphuric acids. 
 
 It is possible to place such physical specifications in a contract 
 as to practically insure obtaining the grade of material ordered. 
 For example, such a high degree of ductility may be demanded,, 
 especially the percentage elongation in ten or twenty feet, that 
 nothing but wrought iron will give it; the strength and ductility 
 may be put so high as to make it too dangerous to try to supply 
 anything but crucible steel for the order; or they may be put a 
 little. lower so as practically to preclude Bessemer steel. The 
 average physical difference between acid and basic open-hearth 
 steels is not great enough to make this method of assurance so 
 
 ^ The threading test for wrought iron and steel pipe is unreliable. 
 
THE OPEN-HEARTH OR SIEMENS-MARTIN PROCESS 167 
 
 practicable, but it is possible in the case of basic and acid Besse- 
 mer steel in England, where alone both these kinds of steel are 
 made in important quantities. 
 
 Prices of Steel, Etc. 
 We may at this point insert, for, the sake of many persons in- 
 terested in the commercial side of the iron and steel industry, some 
 average prices of a few products, which will give an idea of the 
 representative value of all. 
 
 AVERAGE PRICES 
 
 1910 
 
 Ore and Pig Iron per Long Ton; Others per 100 Pounds 
 
 Iron ore, Mesabi, non- Bessemer* $4.00 
 
 Iron ore, Mesabi, Bessemer^ 4 . 75 
 
 Iron ore, Old Kange, non- Bessemer 4 . 20 
 
 Iron ore, Old Range, Bessemer 5 . 00 
 
 Pig iron, Bessemer 17 . 10 
 
 Pig iron. No. 2 Foundry 15.83 
 
 Pig iron, Southern Foundry, at Cincinnati 14 . 00 
 
 Pig iron, Basic 15.67 
 
 Fen^o-silicon, per long ton, 50 per cent, silicon' 52.00 
 
 Ferro-manganese, per long ton, 80 per cent, manganese 42.48 
 
 Ferro-chrome, high carbon, per long ton, 60 % Or 100 . 00 
 
 Ferro-chrome, low carbon, per long ton, 60 % Cr 450 .00 
 
 Ferro-titanium, per 100 lb., 15% Ti 15.00 
 
 Ferro- vanadium, per lb. of vanadium contained * 4.00 
 
 Rails, Bessemer, per 100 lb 1 . 25 
 
 Rails, Bessemer, nickel steel, 3 1/2 per cent, nickel 2.37 
 
 Rails, open hearth (usually Duplex process) 1 . 34 
 
 Structural beams and plates 1 , 45 
 
 Structural shapes, nickel steel, 3 1/2 per cent, nickel 3 . 45 
 
 Wrought iron bars 1 .43 
 
 Black sheets, No. 28 base 2.29 
 
 Wire nails 1 . 75 
 
 Cut nails ■. 1 . 72 
 
 Crucible tool steel bars* $4.50 to $17.00 
 
 Chrome steel, chrome-nickel steel 10 . 00 to 14 . 00 
 
 High-speed tool steels, etc * ^ 15 . 00 to 90 . 00 
 
 Manganese steel castings^ 11.00 to 25.00 
 
 Carbon steel castings 1 .25 to 15 . 00 
 
 ^ Base 51 . 5 per cent. iron. 
 
 ^ Base 55 per cent, iron, under 0.045 per cent, phosphorus. 
 
 ^ The price of all castings largely depends on molding costs. 
 
 * Price depends on maker, brand, etc. 
 
 ^ Price depends on content of Tungsten, etc., and on whether heat-treated by maker or 
 not. 
 
 « This is 1911 price; 1910, $5.00. 
 
 ^ This is manufactured in electric furnace; the blast furnace product containing 10 per 
 cent, silicon, about $22.00. 
 
VII 
 DEFECTS IN INGOTS AND OTHER CASTINGS 
 
 Besides the dangers already mentioned in connection with 
 improper methods of manufacture, excessive amounts of impuri- 
 ties, etc., iron or steel may suffer from damage caused or devel- 
 oped during casting. The commonest defects which may appear at 
 this time are: (1) Blow-holes, or gas bubbles enclosed in the 
 body of the metal; (2) a pipe, or shrinkage cavity; (3) ingotism, or 
 the formation of large-sized crystals; (4) segregation, or the con- 
 centration of impurities in localities; (5) checking or cracking of 
 the casting because of strains produced when the metal is hot and 
 tender. Avoiding these defects cannot make bad steel good, but 
 their presence may make good steel bad, and therefore to guard 
 against them is an important part of the processes. 
 
 Blow-holes. — Blow-holes are especially liable to occur in steel, 
 particularly in low-carbon steel. When the metal is in a molten 
 state, it readily dissolves certain gases, such as hydrogen, nitro- 
 gen, oxygen. Upon solidification these gases come out of their 
 state of solution, but may become entangled in the steel and cause 
 a gas bubble or cavity varying all the way in size from micro- 
 scopic proportions up to an inch or more in length. The for- 
 mation of these blow-holes is precisely similar to the formation of 
 air bubbles in ice; water dissolves a good deal of air while in the 
 liquid state and, as we all know, it is well-nigh impossible to 
 freeze the water without obtaining a great many air bubbles in 
 the ice, due to the separation of this air during freezing. The 
 danger in the case of steel is not so great as in the case of ice, and 
 it is by no means impossible to obtain steel absolutely free from 
 this defect. Apparently, the reason for this difference is that the 
 gas separates from steel a short time before solidificatipn is com- 
 plete, and thus the bubbles have some opportunity to escape 
 before they are enclosed in the solid. Also, steel passes through a 
 pasty stage during solidification, as we shall learn later, and 
 therefore gives a better opportunity for the gas bubbles to pass 
 away. 
 
 158 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 159 
 
 Another cause of blow-holes in steel is undoubtedly the pres- 
 ence of oxide of iron in the metal. This oxide of iron reacts with 
 the carbon added in the recarburizer and forms carbon monoxide 
 gas (CO), which may be produced during the entire solidification 
 period and thus cause many blow-holes. 
 
 Prevention of Blow-holes. — That oxide of iron is one of the chief 
 causes of blow-holes is shown by many things; for instance, (1) 
 steel known to be highly oxidized is very liable to blow-holes; (2) 
 cast iron, which from its chemical composition can never be 
 much oxidized,^ is never subject to blow-holes; and (3) the addi- 
 tion to steel of deoxidizers prevents the formation of blow-holes. 
 
 Chief among the deoxidizing elements which are added for this 
 purpose are manganese, silicon, and aluminum. These elements 
 seem to act partly by deoxidizing the iron and carbon, in both 
 ways preventing the formation of carbon monoxide, and partly 
 by increasing the solvent power of the solid metal for gases, so 
 that a less amount separates. ' -The amount of these deoxidizing 
 substances necessary to be added will depend largely upon the 
 extent to which we desire to prevent the formation of blow-holes. 
 In the case of steel castings it is often necessary that blow-holes be 
 entirely prevented, but in the case of ingots which are to be subse- 
 quently forged or rolled it is not necessary that blow-holes should 
 be absent altogether, because they will be closed up under the 
 pressure of the mechanical work and their sides welded together. 
 Indeed, their presence is sometimes desired, because when they 
 separate from the steel they occupy space, thereby counteracting 
 to a certain extent the shrinkage of the metal during solidification 
 and tending to reduce the volume of the shrinkage cavity or pipe. 
 For this reason a small number in some harmless locality is often 
 intentionally allowed to form in steel ingots. Mr. BrineU found 
 in his steelrworks- that if the percentage of manganese plus 5.2 
 times the percentage of silicon is equal to 2.05 or more, the steel 
 will be entirely free from blow-holes. In this case, however, the 
 pipe will be large. If this sum is equal to 1.66, the steel will con- 
 tain a harmless number of minute blow-holes, but the pipe will be 
 sniall. This figure is therefore about the correct amount for in- 
 gots under conditions similar to those of Mr. BrinelFs experi- 
 ments, and not far different in any event. Mr. BrineU also found 
 
 1 That cast iron is sometimes partially oxidized is claimed by several eminent authorities, 
 and the evidence presented makes us, hesitate to deny, that a certain variety of wild cast 
 iron owes its pecuUar behavior to the presence of some partially oxidized constituent 
 (perhaps the oxysulphide of iron, as suggested by J. E. Johnson). 
 
160 
 
 THE METALLURGY OF IRON AND STEEL 
 
 that the addition to the steel of 0.0184 per cent, of aluminum will 
 give approximately the same result as that given by the amount 
 of manganese and silicon last mentioned, 1.66. 
 
 Location of Blow-holes. — The number and size of blow-holes is 
 no more important, however, than the position they occupy in in- 
 gots in relation to the external surface. Even in castings, blow- 
 holes, if present, should be deep-seated, as they are then less liable 
 to be exposed by machine work performed on the surface. In the 
 case of ingots the deep-seating is of still greater importance, be- 
 cause then the blow-holes may be closed up before they have an 
 opportunity to break through to the surface and thus become 
 oxidized on their interior. The normal gases in blow-holes are 
 
 FIG. 115.-SKIN 
 BLOW-HOLES. 
 
 FIG. 116.— DEEP-SEATED 
 BLOW-HOLES. 
 
 FIG. 117. 
 
 , reducing in effect, and thus the interior surfaces of the holes are 
 bright and silvery in appearance and readily weld together; but 
 if they become oxidized they will never adhere firmly. For in- 
 stance, a blow-hole near the surface, as in Figs. 115 and 117, is 
 liable to break through to the exterior when the ingot is put under 
 pressure. This not only causes a crack in the steel, but allows the 
 air to oxidize the interior of the hole and thus prevent the crack 
 being welded up by the rolling. It is not at all uncommon to see 
 a number of these openings form during rolling when the blow- 
 holes are near the surface. 
 
 As the percentage of manganese plus 5.2 times the percentage 
 of silicon decrease^ from 1.66, the blow-holes become corre- 
 spondingly deeper-seated. Finally when this sum becomes as 
 low as 0.28, the blow-holes are harmlessly located in the interior. 
 It is usually impracticable, however, to get the manganese and 
 silicon as low as this in steel, because manganese is needed to 
 counteract the bad effect of sulphur and oxygen. 
 
 The location of the blow-holes is also very largely dependent 
 upon the fluidity of the metal when first cast into the molds. The 
 more fluid it is, other things being equal, the nearer will the blow- 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 
 
 161 
 
 holes be to the surface of the solid ingot. On the other hand, 
 if the casting temperature is too low there will be a dangerously 
 large number of blow-holes in the steel (see Fig. 117), because it 
 solidifies so quickly that very little opportunity is afforded for any 
 part of the dissolved gases to escape. The fluidity of the steel 
 depends partly upon its temperature and partly upon the amount 
 of impurities in it. For instance, pig iron is fluid at a temperature 
 at which steel is solid; high-carbon steel is fluid at a lower tempera- 
 ture than low-carbon steel. Therefore every different kind of 
 steel has a different correct casting temperature; but we have al- 
 ready learned (p. 89) how to determine this by means of the 
 skull left in the casting-ladle, and it is evident that this test ap- 
 plie^s equally well to all grades of steel. It is to be observed that 
 low-carbon steel suffers greatly from blow-holes, because the more 
 the carbon the less the liability to oxidation of the steel. 
 
 Pipes. — When steel is poured into a mold, it forms almost im- 
 mediately a thin skin of frozen metal against the cold surface of 
 
 Outside, 1, 
 Solidifies. 
 
 Outside 
 gets Cool. 
 
 Pipe 
 Begins. 
 
 All 
 Solid. 
 
 FIG. 118.-SOLIDIFICATION OF AN INGOT. 
 
 the sand or iron. The radiation of heat thereafter necessarily 
 takes place through these surfaces, and therefore a casting will 
 usually complete its solidification by the formation of thicker 
 and thicker layers of solid metal around all the sides. The top, 
 however, will usually remain molten longer than the rest because 
 the hottest metal is usually at this point, having been the last to 
 leave the ladle, and also because the heat is not conducted away 
 by the air as fast as by the walls of the mold. This is especially 
 true where the casting is poured into an iron mold — for example, 
 in the case of ingots (see Fig. 118). But it is evident that at some 
 period a stage will be reached when all the outside of the ingot, or 
 casting, will be covered by a skin of solid metal while the interior 
 11 
 
162 THE METALLURGY OF IRON AND STEEL 
 
 will still be liquid. The liquid interior will continue to freeze and 
 will, at the same time, contract. The result will be the shrinking 
 of the molten mass away from the solid walls and consequently 
 the formation of a cavity, known as a ''pipe," in the interior. 
 This pipe will be filled with the gases evolved by the steel during 
 solidification. Professor Howe has shown that the volume of the 
 pipe is too large to be accounted for altogether by the shrinkage 
 of the steel during solidification, and has shown that the rate of 
 contraction of the inner walls of the ingot being greater than the 
 rate of contraction of the outer walls, a virtual expansion of the 
 outer walls is caused and a consequent enlargement of the pipe.* 
 
 The portion of the steel containing the pipe is of course defect- 
 ive and should be discarded at some time subsequent to casting. 
 In the casting of ingots the upper part, which contains the pipe, is 
 cut off during the rolling or forging and goes back to the furnace 
 to be remelted as scrap. In the casting of steel castings there is a 
 large adjunct to the castings situated above it, and so regulated in 
 size and otherwise that it freezes after the casting itself, and thus 
 always contains a supply of molten metal which runs down and 
 fills any cavity that forms in the casting. This adjunct is cut off 
 when the casting has cooled. In other words, the ''riser" or 
 "feeder," as this extra part is called, serves the same purpose for a 
 steel casting as the upper part of an ingot does for the ingot. 
 
 Past-iron castings do not form a pipe under ordinary circum- 
 stances, because cast iron expands during solidification on ac- 
 count of the separation of graphite, as we shall learn later. 
 Under certain circumstances, however, there may be enough 
 difference in expansion between the inside and outside of cast- 
 iron castings to produce a porous spot which, while not exactly 
 a pipe, is due to similar causes. We shall discuss this matter 
 further in Chapter XII. 
 
 Lessening the Volume of the Pipe, — If the steel were poured into 
 the mold so extremely slowly that it would solidify in layers from 
 the bottom upward there would be no pipe. Therefore one 
 method of lessening the volume of the pipe is by slow casting. 
 We have already noted another method, namely, permitting a 
 small number of blow-holes to form, which causes a certain 
 amount of expansion to the steel during solidification and thus 
 diminishes its shrinkage. Another way is to use wide ingots 
 because this reduces the difference in contraction between the 
 
 1 No. 71. 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 163 
 
 inner and outer layers of the ingot, which, as I have already 
 stated, caused a virtual expansion of the outer walls and thus 
 enlarged the cavity. Casting in sand molds has the same effect 
 because radiation is not so rapid through sand as through metal. 
 Still another method is to prevent the steel forming a solid skin 
 over the top by constantly stirring and breaking it up with an 
 iron rod. This method is often resorted to with the risers of 
 steel castings, with very beneficial results. 
 
 Bottom Casting, — In the case of steel castings, and less fre- 
 quently in the case of ingots, the metal is poured from the ladle 
 into a runner which delivers it at the bottom of the casting (see 
 Fig. 193, p. 226). With steel castings this is often necessary in 
 order to prevent dirt getting into the casting. It also has a 
 similar effect on ingots, because it prevents slag getting into the 
 molds and also prevents metal from spattering up on the side of the 
 mold and forming what is known as a ''cold-shut/' that is, a part 
 of the metal which is not melted in with the rest. In both cases, 
 however, this bottom casting has the effect of increasing the 
 volume of the pipe and also of making the pipe extend deeper, 
 because at the end of casting the hottest metal is at the bottom 
 instead of at the top. 
 
 Casting with the Large End Up, — Risers on castings are al- 
 most always made with the top end larger than the bottom, in 
 order that the pipe may be less in volume and shorter in depth. 
 At steel-works, however, the ingot molds are always tapered 
 slightly with the large end at the bottom, in order that the mold 
 may be easily drawn off the top (see p. 90). This results in the 
 large end of the ingot being down, and consequently in the pipe 
 being larger in volume and very much greater in depth.* Be- 
 cause of this advantage Professor Howe has proposed certain 
 mechanical arrangements by which the ingot may be cast with 
 the large end upward.^ 
 
 Liquid Compression of Ingots, — If the pipe is caused by the 
 difference in expansion between the inside and the outside of an 
 ingot, it is evident that putting sufficient pressure upon the out- 
 side when the walls are solid but the interior is still liquid will 
 prevent the formation of a pipe. Numerous processes have been 
 devised for effecting this liquid "compression," some of which are 
 in operation at steel-works and produce ingots that are entirely 
 
 1 No. 72. 
 
 ^ See No. 71. 
 
164 
 
 THE METALLURGY OF IRON AND STEEL 
 
 free from pipes. In Whitworth's system the ingot is raised and 
 compressed lengthwise against a solid ram situated above it, 
 during and shortly after solidification.^ In Harmet's method 
 the ingot is forced upward during solidification into its tapered 
 mold.^ This causes a large radial pressure on its sides. In 
 Lilienberg's method the ingots are stripped and then run on their 
 cars between a solid and movable wall. The movable wall is 
 then pressed against one side of the ingots,^ 
 
 Ingotism. — When iron and steel freeze they crystallize, and 
 these crystals grow .with great rapidity, so that if the passage 
 through the solidification period is slow they will attain a very 
 
 FIG. 119,— INGOTISM. 
 
 large size. This formation of large crystals is known as "ingot- 
 ism." It is especially liable to occur if the metal is cast at too 
 high a temperature or is allowed to cool through the solidification 
 period at too slow a rate.^ In the case of steel, ingotism may be 
 detected by breaking the casting, when the large size of the crys- 
 tal faces or facets may be observed. 
 
 Damage Due to Ingotism, — Large crystals always produce 
 weakness and loss of ductility, for the large crystals do not adhere 
 to one another as firmly as when there is a more intimate asso- 
 
 1 See page 373 of No. 1. 
 
 ^ See No. 73. 
 In the case of cast iron, large crystals formed during solidification produce what is 
 -known aa an "open grain"; we shall consider this more particularly in Chapter XII. The 
 name "ingotism" is not usually applied to this open grain in cast iron. 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 
 
 165 
 
 ciation; consequently steel that shows ingotism will be tender. 
 In the case of steel castings, they will not give as good a result in 
 the testing machine; in the case of ingots they will be liable to 
 tear during rolling or under the hammer. 
 Remedy for Ingotism, — Ingots in which 
 large . crystals have formed during solidi- 
 fication may be brought to a high degree 
 of strength and ductility by forging or 
 rolling, because the mechanical work 
 crushes the crystals and reduces them to 
 a smaller srze. The work must be done 
 very carefully at first, however, or cracks 
 
 — B 
 
 CD, 
 
 fig; 120.— lines of 
 equal carbon- 
 percentage in 
 a steel ingot. 
 
 Top in ingot 
 
 O • O 
 
 o 
 
 .61 m 
 
 .27 .28 .29 .28 .27 .29 
 
 .25 .30 .21 .26 .2G .SO 
 
 o o o o o 
 
 .24 .21 .83 
 
 o o o o o 
 
 .28 .23 .22 
 
 o o o 
 
 .22 .38 .24 
 
 .24 .26 .20 
 
 ■BO ,29 .80 .29 
 
 FIG. 121.— CARBON-PERCENTAGE AT 
 DIFFERENT PARTS OF A STEEL 
 INGOT. 
 
 will be formed that are not*afterward welded up. Ingotism in 
 steel castings is not so easy to cure; indeed, it is maintained by 
 some authorities that its bad effects are never entirely obliterated. 
 I am inclined to agree with this opinion, although annealing the 
 
166 
 
 THE METALLURGY OF IRON AND STEEL 
 
 steel at a proper temperature <see Chapter XIV) will produce a 
 very beneficial effect. 
 
 Segregation. — When either iron or steel is molten, the various 
 impurities are dissolved in it, and some of them, especially carbon, 
 phosphorus, and sulphur, make the metal more fusible, that is, 
 they lower its melting-point. But the impurities are not as 
 soluble in the solid metal, and therefore tend to separate on solid- 
 ification; so it can readily be conceived how each layer that 
 freezes, beginning at the outside, rejects some of its impurities to 
 be dissolved by the still liquid mass in the interior. When the 
 next layer freezes that too will tend to reject a part of its im- 
 purities into the contiguous molten layer, and thus the concentra- 
 tion will proceed so that as a general thing the portion of the 
 metal richest in impurities, especially in carbon, phosphorus, and 
 sulphur, will be that which freezes last. With ingots, this will 
 evidently be at a point just below the bottom of the pipe, and it 
 is found to be so in the great majority of cases; but the location 
 of the richest segregate is very liable to vary, and rules can only 
 be used for general guidance. For example, in Fig. 120,^ the 
 most impure metal is found at a point higher than the bottom of 
 the pipe, and other unexpected exceptions occur. In the case 
 of iron and steel castings, the most impure point will generally 
 be near the top of the thickest section of metal. The riser is 
 calculated to be the last portion to freeze and the richest segregate 
 should be located in it. 
 
 In iron castings which contained on an average less than 1 per 
 cent, phosphorus and 0.1 per cent, sulphur, I found on one occa- 
 sion a segregate portion containing 1.856 per cent, phosphorus and 
 0.144 per cent, sulphur; and on another occasion I found one 
 containing 2.43 per cent, phosphorus and 0.236 per cent, sulphur. 
 An extreme case of segregation in steel is shown in the following 
 analysis:^ 
 
 Carbon 
 Per cent. 
 
 Silicon 
 Per cent. 
 
 Manga- 
 nese 
 Per cent. 
 
 Phosphorus 
 Per cent. 
 
 Sulphur 
 Per cent. 
 
 Average. . . 
 Segregate. 
 
 0.24 
 1.27 
 
 0.336 
 0.41 * 
 
 0.97 
 1.08 
 
 0.089 
 0.753 
 
 0,07 
 0.418 
 
 1 From page 205 of No. 71. 
 
 2 From page 373 of No 1. 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 167 
 
 Treatment of Segregated Steel. — Segregation cannot be pre- 
 vented, although, of course, it seldom takes place to the degree 
 shown in the extreme cases that I have cited above. Neverthe- 
 less, there are always certain portions of the ingot or casting 
 which are richer in impurities than others. An attempt is made 
 to get this richer portion into the upper part of an ingot, or into 
 the riser of a casting, and then it is cut off when the ingot is 
 rolled, or when the riser of the casting is removed. It is there- 
 fore advantageous to cause the segregate to go as high up in the 
 ingot or casting as possible. Whatever tends to raise the whole 
 pipe higher up in the casting would, in general, tend to raise the 
 segregate also; but wide ingots, or ingots cast in walls with low 
 conducting power, though they tend to decrease the volume of 
 the pipe, would not necessarily raise the segregate to a higher 
 point. Furthermore, wide ingots will probably have a much 
 greater degree of segregation than narrow ingots, other things 
 being equal, because the wider the ingot the greater will be the 
 number of ^layers of solidification, and consequently the greater 
 concentration of impurities in the center. 
 
 Lessening Segregation. — Benjamin Talbot^ has shown that 
 quieting the steel by adding aluminum to it will lessen the segre- 
 gation. J. E. Stead^ argues that this result is due to the branches 
 of crystals (commonly called ^' fir-tree crystals'); which grow per- 
 pendicularly to the cooling surfaces when steel solidifies and me- 
 chanically entangle some of the impure metal, thus preventing it 
 from traveling inward. Professor Howe calls this type of freezing 
 the "land-locking type." When the steel is violently agitated by 
 the escape of ^as its rapid movement washes off the fir-tree crys- 
 tals and prevents them from growing out into the liquid mass and 
 entangling the impure metal. The quietness produced by alu- 
 minum, however, makes this growth possible. 
 
 Another important means of lessening the segregation is by 
 making ingots narrow,^ that is, by reducing the area of the hori- 
 zontal cross-section; but this is often difficult of accomplishment. 
 For example, if we cast fifty tons of open-hearth steel out of one 
 ladle, it will take a very long time to cast all of this into small in- 
 gots, and therefore the first ingots cast will be too hot or else the 
 
 1 Pages 204 to 223 of No. 74. 
 
 2 Pages 224 to 228 of No. 74. 
 
 ^ Some metallurgists disagree with this and believe that the large ingots do not segregate 
 so much. Nevertheless, I am inclined to think that the greater weight of evidence is 
 against them, in this contention. 
 
168 THE METALLURGY OF IRON AND STEEL 
 
 last ingots will be too cold. There is a difference of opinion as to 
 whether or not rapid cooling decreases or increases the degree of 
 segregation, and it seems probable that it acts in both directions, 
 sometimes prevailing one way and sometimes in the opposite. 
 On first thought it would seem that slow cooling must necessarily 
 increase segregation, because it would allow-more time for the im- 
 purities to separate from one layer of metal and dissolve in the 
 next. On the other hand, slow cooling also favors the growth 
 of the fir-tree crystals, and therefore opposes segregation. It 
 does not seem possible, at the present time, to tell under what 
 conditions we should have the one influence prevailing or the 
 other. 
 
 It seems to be pretty well established that the greater the per- 
 centage of impurities present the greater will be the extent of the 
 segregation. Therefore high-carbon steel should be cast with due 
 care and narrow iAgots used wherever possible. Generally when 
 the phosphorus and sulphur are low (say, not more than 0.05 per 
 cent, each), much segregation is not liable to occur, especially in 
 low-carbon steels. 
 
 'Occluded Oxidized Substances.- — Steel and iron castings often 
 contain after solidification particles of oxidized foreign subsljances. 
 When large in size, these enclosures are commonly the result of 
 accident, negligence or ignorance, and therefore could usually 
 have been avoided by the exercise of due care, but, in micro- 
 scopic particles, aggregating as much as 0.25 to 0.30 per cent, of 
 the weight of the metal, they are often found in Bessemer and 
 open-hearth steels, not as a result of chance, but as a regular 
 occurrence. Oxides and sulphides of iron and manganese, oxide 
 of silicon, and tiny particles of slag, are the commoner examples 
 of these enclosures. Entangled particles of slag seem to be due 
 to their having become caught in the solidifying metal before 
 they had an opportunity to rise to the surface. No doubt the 
 bulk of the oxides of manganese and silicon originate through 
 the recarburizing reactions in the bath, and the tiny particles 
 thereof are retained because they cannot separate immediately. 
 Like the cream in milk, time must be allowed for them to rise 
 to the surface. The same necessity is felt to an even greater 
 degree in the case of sulphide of manganese, which separates 
 from a molten bath of iron or steel only upon long standing. 
 The sulphide of iron appears to be actually soluble in the molten 
 metal and not to separate until after solidification. 
 
DEFECTS IN INGOTS AND OTHER CASTINGS 169 
 
 All these occluded oxidized particles are harmful to the metal, 
 to the same extent as microscopic blow-holes, for example. It 
 is now commonly believed that many previously unexplained 
 failures of steel are due in part at least, to such defects, and the 
 "shelling" off of the surface of railroad rails and locomotive 
 tires is often directly traceable to this cause. 
 
 Titanium. — ^Titanium is now often used in steel and cast iron 
 baths for the purpose of a final deoxidation of the metal and its 
 impurities, following the usual treatment for this object by means 
 of manganese and silicon. Titanium has a strong chemical 
 affinity for oxygen and will remove from the metal traces of 
 oxygen which do not yield to the ordinary recarburizing reactions. 
 It also has the further and apparently unique property of causing 
 occluded oxidized substances to separate more readily, thus avoid- 
 ing some of the harmful effects noted in the previous section. 
 The most notable instance of titanium treatment is in the manu- 
 facture of railroad rails and during the year 1910 more than 5 
 per cent, of the total rail production of the United States was 
 treated with this element. The beneficial effects of the titanium 
 treatment, from the standpoint of the railroad rail user, are: 
 (1) Increased life of rails — say from 200'to 400 per cent.; (2) less 
 brittle (i.e., tougher) rails under the drop test; (3) a stronger 
 metal, and (4) less flaking off of the metal in the head. Some 
 railroads find that the slightly higher cost of the titanium-treated 
 metal justifies its use in rails in all parts of the line, and others 
 employ it at present only on cuj'ves or at other points where the 
 wear is unusually heavy. 
 
 The treatment consists of a physicing of the molten steel (either 
 in the bessemer or open-hearth furnace, or in the ladle into which 
 the product of the furnace is being poured after the end of the 
 operation) by means of about 0.3 to 0.6 per cent, of an alloy con- 
 taining 10 to 15 per cent, titanium, about 80 per cent, iron, 5 to 
 8 per cent, carbon, 0.35 to 1 per cent, silicon and all other im- 
 purities guaranteed to total less than 1 /2 of 1 per cent. The titan- 
 ium alloy must be added last, after the requisite amount of carbon, 
 silicon and manganese. Aluminum must not be used for puri- 
 fying the same steel in which the titanium alloy is employed 
 and if the titanium alloy is not used subsequently to the carbon, 
 manganese and silicon, its essential ingredient is liable to be 
 excessively oxidized and wasted. 
 
 In the bessemer process the titanium should be shoveled into 
 
170 THE METALLURGY OF IRON AND STEEL 
 
 the stream of metal which is being poured into the ladle, the steel 
 having been previously recarburized in the converter. As the 
 titanium alloy reacts energetically with the slag and may there- 
 fore be wasted, it is especially recommended that it have the 
 maximum possible amount of churning and stirring with the 
 metal before much slag enters the ladle. For this purpose it is 
 well to let the stream enter a little off center of the ladle in order 
 to give the molten bath a sort of gyratory stirring. After this 
 treatment it is well to "hold the metal about 3 min. in order that 
 the titanium may complete its action and carry off as much as 
 possible of the dissolved nitrogen, oxygen and entangled slag. 
 
 In the acid open-hearth process the same general procedure is 
 followed, except that the metal should be held perhaps 5 or 10 
 min. to give the impurities released by the titanium time to 
 pass off with the slag. Where, however, the titanium is added 
 in the acid open-hearth furnace the greatest care must be exer- 
 cised to keep it from contact with the.slag and it must be rabbled 
 into the metal as quickly and thoroughly as possible. The bath 
 must then be gotten out of the furnace at the earliest possible 
 moment as every delay tends to undo the purification work per- 
 formed by the alloy. In basic open-hearth steel it is well to add 
 at least three-quarters of the manganese in the bath; then add 
 the remainder of the manganese and silicon at the earliest possi- 
 ble moment in the ladle and follow with the titanium alloy, using 
 the precautions already mentioned for mixing and dissolving it 
 thoroughly in the metal and keeping it out of contact with the 
 slag. 
 
 Under no circumstances should the titanium alloy be pre- 
 heated before it is added, or be used in conjunction with alumi- 
 num.* These warnings are more than precautions that should 
 be observed; they are quite necessary to obtain success in the use 
 of titanium, and should be regarded with at least as much respect 
 as a good steel maker pays to any essential of correct practice. 
 
 Besides its use as a "super-purifier" for rail steel, titanium is 
 being used for treating steel and iron rolls, whoee resistance to 
 abrasion and shock it is said to increase; for steel castings; for 
 chain of high and low carbon, gears and pinions, tires, die plates 
 and heads, propeller shafts, driving rods and other forgings, 
 
 1 However, W. Venator testifies to the contrary, and states that he finds aluminum 
 beneficial. — Stahl and Eia&a. Vol. XXX, pp. 650-4 — but he used titanium metal instead 
 of the 10 per cent, ferro-alloy. 
 
DEFECTS OF INGOTS AND OTHER CASTINGS 171 
 
 automobile steels, tool steels, cast iron ingot molds, acid pots 
 and chilled car wheels. It is said to improve the qualities of 
 other alloy steels such as nickel steels, chrome steels, etc. 
 
 Titanium is also said to remove nitrogen from iron and steel, 
 titanium being one of the very few elements which attacks this 
 comparatively inert gas. It is also said to lessen the segregation 
 of steel ingots, which may be believed because of the well-known 
 influence in this respect of agents which deoxidize and quiet the 
 metal. ^ 
 
 Instead of introducing titanium by means of the so-called 
 ferro-titanium alloy, it may be reduced by the Goldschmidt 
 thermit process, although this is not as common practice. In 
 this process we use powders consisting of oxide of titanium mixed 
 with pulverized aluminum. When a reaction is started by heat- 
 ing the mixture, either with the heat in the molten bath, or by 
 means of certain fuses, a reaction is produced whereby titanium 
 is reduced, and at the same time, heat is generated. 
 3Ti02 + 4A1^3Ti + 2Al303(evolves 65,200 cals. estimated). 
 
 Vanadium in Steel. — While titanium is used for the purpose 
 of treating and superrefining steel and cast iron, it is not ordi- 
 narily employed to manufacture one of the socalled alloy steels 
 in accordance with the definition on page 389. Vanadium, on 
 the other hand, is used,^although generally, in conjunction 
 with some other element, such as chromium or nickel, — to pro- 
 duce an alloy steel, but it is also used, like titanium, for the pur- 
 pose of carrying the deoxidation beyond the point obtainable by 
 means of manganese, silicon and aluminum. Vanadium has a 
 strong chemical afiinity for oxygen, and when added to the liquid 
 iron or steel, to the extent of 0. 15 to 0.30 per cent., after the usual 
 recarburizing, it seems to cause further deoxidation of the metal 
 and its impurities. Vanadium is usually added by introducing 
 into the liquid bath small pieces of vanadium alloy, containing 
 iron with varying amounts of vanadium, of which the 35 per cent, 
 alloy is a common example. It is melted without difficulty by 
 the heat of the molten bath and the vanadium dissolves in the 
 
 metal, attacking the oxides present. 
 
 > 
 
 1 George B. Waterhouse, Proceedings, American Society for Testing Materials, Vol. X, 
 1910, pp. 201-211. 
 
VIII 
 THE MECHANICAL TREATMENT OF STEEL 
 
 Metals may be shaped either by pouring them while molten 
 into^ mold, as described in the following chapter, or by mechan- 
 ical pressure. The choice of the casting or the mechanical 
 method of shaping will depend on the cost of production, the size 
 and form of the finished product and the purpose for which it is 
 intended. Some shapes must be produced by casting, because 
 they are either too intricate or too large to be shaped by pres- 
 sure; others must be produced by pressure, because the service 
 in which they are to be used demands the higher strength and 
 ductility which mechanical work produces. Between these two 
 classes, however, is a large number of forms, each of which is a 
 study by itself. Financial considerations will govern in some 
 cases, and the importance of quality in others. The advantage 
 of quality is usually with the pressed material. Cost is in favor of 
 castings except where many pieces are to be produced alike in 
 shape and size. 
 
 Effect of Work. — Mechanical work will multiply the strength 
 of steel from two to five times. In order to accomplish as much 
 as this latter, however, it is necessary (1) to reduce the material 
 to very small sizes, in order that the beneficial effect of the knead- 
 ing action may extend throughout the mass, and (2) to finish the 
 work cold, in order that the metal may have no opportunity to 
 recrystallize. The ductility also will be increased at first by 
 working, but again decreases if the metal is worked cold. The 
 increase in strength and ductility is due (1) to decreasing the size 
 of the crystals and closing the grain of the steel; (2) to the closing 
 up of blow-holes, both large and small, which are almost all 
 welded together under pressure at high temperatures, unless they 
 are near enough to the surface to become oxidized inside, and 
 (3) to increasing the cohesion and adhesion of the crystals. 
 All these effects increase the specific gravity and hardness of the 
 metal, and are more effective in these respects, as well as in 
 increasing strength, if hot work is followed by cold work. 
 
 172 
 
THE MECHANICAL TREATMENT OF STEEL 173 
 
 Crystallization of Steel, — Metals are crystalline substanceS; 
 • the individual components arranging themselves in regular 
 forms unless opposed by the rigidity of the mass in which they 
 form. Indeed, the metallic crystals grow with astonishing 
 rapidity when the metal crystallizes from the molten state (i.e., 
 solidifies), or even when it is in a mobile condition (i.e., at tem- 
 peratures near or above a red heat). Once crystals have formed 
 they cannot be reduced in size except by annealing (see Chapter 
 XIV), or by breaking them up by mechanical crushing. These 
 facts are important, because large crystals do not adhere to 
 each other firmly, and thus they cause a weak and brittle mass. 
 Iron and steel follow the same laws as other metals in these 
 respects. 
 
 FIG. 125.— STRAIGHT SLIP BANDS FIG. 126.— CURVED SLIP BANDS 
 
 IN WROUGHT IRON MAGNI- IN WROUGHT IRON fitAGNI- 
 
 FIED 60 DIAMETERS. FIED 60 DIAMETERS. 
 
 Unetched. (William Campbell.) Unetched. (William Campbell.) 
 
 Effect of Strain. — When a metal is strained, the crystals first 
 stretch, and the amount of this stretching is directly proportional 
 to the strain; when the strain is relieved, the crystals and the mass 
 as a whole, return to the original dimensions. If the strain is 
 greater than the "elastic limit," however, the crystals yield, and 
 the particles composing them slip along the cleavage planes, so 
 that a permanent deformation or extension occurs in the direc- 
 tion of the strain. This "elongation" is accompanied by a 
 "reduction in cross-sectional area," and gives "warning that the 
 material is suffering from excessive strain. The extent to which 
 these two forms of distortion precede rupture is usually taken 
 as the measure of the " ductility" of the metal. 
 
174 THE METALLURGY OF IRON AND STEEL 
 
 Rationale of the Effect of Work. — Mechanical pressure upon a 
 metal crushes the crystals, mixes them intimately together, and ' 
 breaks up the cleavage planes along which they would yield. If 
 the work is finished above a red heat where the mass is still mobile, 
 the crystals reform to a certain extent, decreasing the strength. 
 The elastic limit of a structural steel rolled in this way will be a 
 little more than one-half its ultimate strength. If the work con- 
 tinues while the metal is cold, there is no opportunity for the 
 reformation of the crystals, and the strength, hardness, and 
 brittleness are much increased. 
 
 Methods of Applying Pressure. — Aside from the differences of 
 hot and cold working, the mechanical pressure may be exerted in 
 one of three ways: (1) Instantaneously, by a blow, in which 
 method the pressure is relieved before the metal has fully yielded 
 to it; (2) more slowly, by rolling or wire-drawing, in which the 
 pressure is relieved almost as soon as the metal has yielded to it; 
 and (3) slowest, by presses, in which the pressure remains for a 
 second or so after the metal has yielded. See page 214 to 
 compare. 
 
 The Forging of Metals 
 
 The instantaneous application of pressure is man's first method 
 of shaping metals and is accomplished by a blow from a falling 
 weight, frequently aided by some other force. Examples of the 
 first practice are found at the present time in the helve-hammer 
 used at many small forges and steel-works. After the introduc- 
 tion of steam, however, this was used to raise the weight, and 
 very soon steam-power was employed not only to raise the 
 weight but also to force it downward for the blow, whose momen- 
 tum was thus greatly increased. Hammers of this type, which 
 is now the most prevalent, have been built capable of delivering 
 a blow estimated as equivalent to 150 tons weight. Such large 
 sizes are not now approved of, however, because of the inordinate 
 expense for foundations, which must be deep and powerful in 
 order to take up the force of the blow, while the constant jarring 
 disturbs the foundations and alinement of machinery, even at 
 distant parts of the plant. For very heavy forging work, such 
 as armor-plate, etc., the hydraulic press is therefore preferred, 
 and hammers are not often built in sizes above 30 or 50 tons. 
 
 Effect of Hammering. — A blow creates in a metal practically 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 175 
 
 nothing but compressive strains, which act chiefly in the vertical 
 direction, and, by transmission, in the two horizontal directions. 
 Because the pressure is relieved almost as soon as felt, the elas- 
 ticity of the metal causes it to recover somewhat from the effect. 
 This makes the effect of hammering superficial. Also the 
 amount of yield to it is not great in proportion to its force, and 
 therefore it takes more pressure to accomplish a result than 
 
 FIG. 127.— STEAM HAMMER. 
 
 would be the case if the application was slower. This makes 
 hammering a slow process of reduction, but results in a better 
 and more uniform working of the crystals on the surface at least, 
 which is one of the chief reasons for the superiority of hammered 
 over rolled material. Another and,' perhaps, even more potent 
 reason is the exact control of the operation which can be exercised 
 by the expert forger, and more especially his control over the 
 
176 THE METALLURGY OF IRON AND. STEEL 
 
 temperature at which the work is finished, and over the varying 
 force of pressure applied at different stages and temperatures* 
 On the other hand, the effect of forging extends only skin-deep 
 from the upper and lower surfaces. 
 
 Finishing Temperatures for Forging. — ^Forging seldom con- 
 tinues after the red heat is lost, but the exact temperature will 
 depend upon the article and the properties which it is desired 
 to have. The colder it is finished the closer to the exact size 
 required it can be made, because it has less shrinkage to undergo; 
 but it will also be harder, less ductile and stronger. The relation 
 between the finishing temperature of mechanical work and the 
 critical points of steel will be discussed in Chapter XIV. 
 
 Drop-forging. — There is a large variety of articles, such as 
 parts of machinery, hammer-heads and similar tools, which are 
 formed by the process known as "drop-forging." In this opera- 
 tion a piece of metal of the desired size is forged by repeated blows 
 between a lower die, upon which it rests, and an upper die at- 
 tached to the head of the hammer. These two dies are made in 
 the desired form of the finished article, and the metal is squeezed 
 into them until it has assumed the proper shape. Sometimes 
 several pairs of dies are necessary to complete the finished shape, 
 (see Fig. 128). Drop-forgings are directly comparable with steel 
 castings, to which they are superior in quality on account of the 
 beneficial effect of the working. To be economically made, they 
 must be ordered in large quantities, so that it will pay to make 
 the costly dies of hardened steel — often an alloy. Even then, 
 castings are sometimes cheaper, though f orgings are stiU preferred 
 on account of their quality. There are cases, however, in which 
 drop-forgings may be made more cheaply, either because the 
 shape is one that lends itself to rapid production in this way, or 
 because it is one' liable to checking, or requiring a large riser, if 
 cast. 
 
 Forging Bars. — Crucible-steel ingots are often forged instead 
 of rolled, because the material will bring a price high enough to 
 pay for the superior method of working it. The ingot, after the 
 top third has been broken off to remove the pipe and segregate, 
 is heated to a bright-red heat, out of contact with the flame and 
 fuel, and then tilted down under a hammer of about 10 to 15 
 tons size until it is about one-half as large on the sides and four 
 times as long. One end is then reheated and drawn down to a 
 bar of the desired size, under the same hammer, or under one of 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 177 
 
 less weight, and the long bar is then used as a handle while the other 
 end undergoes heating and reduction. The finished size is pro- 
 duced by light taps of the hammer just before the blue heat 
 appears, and often a piece of cold steel is laid beside it on the anvil 
 
 FIG. 128.— SOME AUTOMOBILE DROP-FORGINGS. 
 
 to more correctly arrest the downward blow. The finished bar 
 will be so straight and true as to lead one to believe that it was 
 produced by drawing through a die, or rolling in grooved rolls. 
 
 Forging Razors. — ^Flat bars for razor stock, made of cemented 
 
 ^ 12 
 
178 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 129.— SOME STAGES IN THE MAKING OF A DROP-FORGING. 
 
THE MECHANICAL TREATMENT OF STEEL 179 
 
 steel melted in crucibles without additional carbon, are produced 
 in a manner similar to that outlined above, and then forged down 
 by hand to the rough size of a razor. They are then stamped with 
 t-he appropriate name and mark, drilled with a hole, heated to the 
 correct temperature and hardened in water, after which the 
 temper is drawn to the light or medium straw color (see page 
 383). The exact shape is then produced by grinding, care being 
 taken not to heat the razor during this operation, lest it be tem- 
 pered thereby, and the blade polished and fitted with a handle. 
 
 Forging Cannon, — Large cannon tubes are made from open- 
 hearth steel ingots weighing perhaps 65 tons or -so, more than 
 one-half of which is discarded or "scrapped'' during the process. 
 In France and Germany, cannon-tube ingots 
 have been made of crucible steel by pouring 
 many crucibles into one mold, but the expense 
 and the liability to heterogeneity because of 
 the many small units is belived to outweigh 
 the advantages due to the quality of crucible 
 steel. The heating of the ingots must be done 
 with great care, lest a crack or hollow be fig. i30. 
 
 formed by too rapid expansion or by the ex- 
 pansion of the outside away from the interior, and for the same 
 reason ingots of the form shown in Fig. 130 are usually em- 
 ployed, and those cast in sand molds are preferred .because they 
 are not liable to contain surface cracks produced by tearing the 
 steel when the iron mold is withdrawn. Moreover, as reducinga 
 flame as possible must be maintained, lest the carbon be oxidized 
 in the outer layers of steel during the many hours required to 
 attain the bright-red heat necessary. 
 
 During the long forging process reheatings of about an hour or 
 so are necessary at intervals. Not more than the lower two- 
 thirds of the original ingot is allowed to be present in the tube at 
 the finish of the forging operation in order that the segregated 
 and piped portion may be avoided. The blows are delivered on 
 all sides of the ingot in order that the center of the tube shall be 
 the same as the center of the original ingot, for this center portion 
 is to be drilled out in the subsequent operations, and, as we 
 have already seen, the center of the ingot is of looser texture and 
 contains more of the segregate. 
 
 After the inner tube is forged, an outer tube is produced in a 
 similar manner, but of larger size, so that it may be bored out to 
 
180 THE METALLURGY OF IRON AND STEEL 
 
 fit over the carefully turned inner tube. After the boring the 
 outer tube is too small to pass over the inner one, and it is there- 
 fore heated to a temperature of about 280° C. (550° F.) in a tall 
 vertical furnace, which expands it so that it may be passed over 
 the inner tube and "shrunk" upon it, greatly increasing its 
 compactness and reinforcing it against the tremendous strains 
 it is subjected to in service. Cannons are now frequently forged_ 
 under presses instead of under hammers. 
 
 The Reduction of Metals in Rolls 
 
 If two rolls rotating as shown in section in Fig. 131 be made to 
 grip a piece of metal, A, they will drag it between them and force 
 it out on the other side reduced in thickness. The metal between 
 the points and NN is being compressed vertically, while its 
 
 FIG. 131. 
 
 outer layers are suffering tension. In the case of a deep section, 
 the unequal strain produced by unequal speed of travel is liable 
 to tear the steel (see Fig. 165). The mechanical pressure is there- 
 fore not as uniform as in hammering, and acts for a longer period 
 of time. Reduction can only take place vertically, as in forging, 
 there being always a certain amount of expansion sidewise, and a 
 large amount of extension in length. The metal at the points 
 N N being forced backward, and that at the points being 
 forced forward, the ends of the rolled section assume a shape 
 somewhat like that shown in Figs. 132 and 133. The reduction 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 181 
 
 in area at each "pass" will vary between 5 and 50 per cent, of 
 the original, and the work is very rapid. For example, a rail- 
 road rail may be produced, from an ingot having a section 18 in. 
 square, in 22 passes, varying in amount of vertical squeeze from 
 8 to 50 per cent., only about five minutes being required for the 
 whole operation, the piece traveling through some of the passes 
 at a rate of ten miles per hour, and in some mills not being re- 
 heated after the ingot comes to the first pair of rolls. Some 
 American rolling-mills produce over a mile of single rail per hour 
 for 24 hours a day and 26 days per month. The temperature 
 
 FIG. 132.— SHAPE OF ENDS OF 
 ROLLED METAL WHEN THE 
 INSIDE IS THE HOTTER. 
 
 FIG. 133.— SHAPE OF ENDS OF 
 ROLLED METAL WHEN THE 
 OUTSIDE IS THE HOTTER. 
 
 at which the rolled material is finished is gaged with much less 
 accuracy than in forging operations, and is always too high for 
 the best quality of the steel, because economy of power urges 
 the manufacturer to work. the metal hot. 
 
 Pull-over Mill. — In a single pair or rolls, such as shown in 
 Fig. 131, the metal, after passing between them once, must be 
 handed or ptiUed over the top of the mill, to be fed in for a second 
 pass. This type of train is known as a "pull-over" or "pass- 
 over " mill. It can be used only for shapes small in size and that 
 can be handled readily, and the action is slower than in a con- 
 tinuous operation, such as in a "three-high mill.*' The pull-over 
 mill is simple and cheap to construct and operate, and is used 
 especially for the rolling of plates and shapes from crucible steel, 
 whose high price renders it less important to seek rapid output. 
 It is also used very largely for the rolling of steel to be used for 
 tin-plate. The upper roll is adjustable, so that any thickness 
 may be produced. 
 
 Multiple-ply Plate, etc. — Three-ply plates for plow-shares, 
 and five-ply plates or bars for burglar-proof safes and jail 
 bars, are often made in this type of mill. We first roll in- 
 dependently thin plates of high-carbon, crucible, chrome steel 
 
182 
 
 THE METALLURGY OF IRON AND STEEL 
 
 (or an equivalent alloy steel capable of becoming very hard upon 
 quenching in water from a red heat), and thicker plates of wrought 
 iron. For a plowshare, a plate of wrought iron will then be sand- 
 wiched between two plates of chrome steel, tied into a bundle 
 with wire, and raised to a welding heat. The wire burns off in 
 the furnace, but the bundle is grasped with a pair of tongs and fed 
 into a pair of plain rolls, where it is welded into a plate of three- 
 ply steel which is reduced to a thickness of a little over a quarter 
 of an inch. This plate is trimmed and then hardened and used 
 
 FIG. 134. 
 
 for a plow, the hard outer layers resisting the attrition and 
 wear of service, and the ductUe core resisting the shocks. For 
 safes and jail bars we have an inner layer of wrought iron, then 
 two layers of chrome steel, and then two layers of wrought 
 iron, welded together and then hardened. A prisoner can neither 
 file this, on account of the hardened chrome steel, nor break 
 it with a sledge, on account of the soft iron, which is not har- 
 dened by quenching. 
 
THE MECHANICAL TREATMENT OF STEEL 183 
 
 Three-high Mills. — When a piece is passed over a two-'high 
 mill, it is often rested upon the top of the upper roll, whose travel 
 assist somewhat in the transfer. While watching this operation 
 at the Cambria Iron Company's mill in 1857, John Fritz con- 
 ceived the idea of the three-high mill, which is shown in section 
 in Fig. 134. It will be seen that the piece receives work in both 
 directions. At the present time the great bulk of the tonnage 
 of steel and wrought iron produced, consisting of structural 
 shapes, railroad rails, plates, wire rods, billets and bars, is 
 finished in this type of mill. The output is large, because the 
 rolls can be run very fast indeed (rod mills running 600 to 1200 
 revolutions per minute and sometimes passing the rod through 
 at the rate of half a mile a minute in American practice^), and 
 
 
 
 fW 
 
 I 
 
 m. i^iMMitiilHil 1 
 
 ^^^^H 
 
 m. -,S*^" ""'■"" ....'- 
 
 2 
 
 FIG. 135.— THREE-HIGH MILL. 
 
 two or more pieces may be passing through at the same time. 
 The disadvantage of the three-high mill is the power neces- 
 sary to raise large weights up to pass over the middle roll. Most 
 Bessemer ingots are cast two tons or more in weight, and most 
 open-hearth ingots from three tons to ten or more tons. 
 
 Reversing Mills, — Therefore ingots are often "cogged" in two- 
 high reversing mills to avoid this consumption of power. More- 
 over, the two-high mills, which have an adjustable upper roll, 
 have the advantage of being able to work an ingot gently at first, 
 in case it shows a tendency to be "tender,'* that is, to crack in 
 
 1 The reason for this rapid rolling is not only large product, but that the thin rods may not 
 radiate their heat during the operation and thus be finished too cold. This rapid work 
 actually raises the heat of the metal faster than it can be radiated, an^ rods are hotter at 
 the end than at the beginning of the rolling. 
 
184 
 
 THE METALLURGY OF IRON AND STEEL 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 185 
 
186 THE METALLURGY OF IRON AND STEEL 
 
 spots when the pressure is applied. The disadvantages of the 
 two-high reversing mill are its slowness and the severe strain on 
 the engines, which are often reversed while running full speed. 
 
 Universal Mill. — During the rolling of metal there is a cer- 
 tain amount of expansion sidewise, which gives the piece a cross- 
 section somewhat bulging on the sides, and makes the edges 
 uneven, unless the rolls have collars which form a groove through 
 which the metal passes. In 1855 R. M. Daelen, at Hoerde, 
 Germany, devised a mill in which even edges could be produced 
 at any width by having an auxiliary pair of vertical rolls, between 
 which the piece passes immediately after it emerges from the 
 horizontal rolls. These vertical rolls are adjustable to any width 
 up to the capacity of the mill, and give only enough pressure to 
 keep the edges even without producing any reduction. They 
 are usually made to rotate with a surface velocity greater than 
 that of the horizontal rolls, so as to prevent the serious buckling 
 that would take place if the conditions were reversed. As this 
 tension is not good for the edges of the metal and wears out the 
 vertical rolls, some mills have independent control of drive for 
 each pair of rolls, and others have friction-clutches connected 
 with the vertical rolls, which allow them to run faster if pushed 
 by the metal, but ordinarily run them at a slower speed. Uni- 
 versal mills are made two-high or three-high, and with vertical 
 rolls on one or on both sides of the horizontal rolls. In England, 
 the Universal mill is not as much in favor as grooved rolls because 
 rolling-mill managers believe that the faster work of the latter 
 more than makes up for the necessity for changing rolls when- 
 ever a new width is to be produced. 
 
 Parts of Rolling Mills 
 
 Rolls. — Rolls may be plain cylinders, by which plates and 
 rectagonal shapes are produced, or they may be cylinders with 
 "collars" at intervals, as shown in Fig. 138, in which large 
 rectagonals with even edges may be produced; and the collars 
 may be on both rolls, giving an ''open pass, " or may be on only 
 one roll and extend into grooves on the other roll, as shown in 
 Fig. 139, giving a "closed pass." With open passes, the collars 
 cannot be made to quite touch, hence the name; and the pressure 
 may squeeze some metal between them, forming a "fin" along 
 the side of the piece. This results from "overfilling the pass." 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 187 
 
 The closed pass makes the upper roll weaker, and there is also a 
 liability of the metal becoming wedged tightly between the 
 collars and thus drawn all the way around the roll, with the 
 result that something will be broken. Wedge-shaped grooves 
 may be cut in the rolls, producing the ''diamond" pass, in which 
 small squares are made (see Fig. 139); or oval grooves make 
 nearly round bars which are finished round in the last pass with 
 almost no draft. Other forms of passes are shown in Figs. 135 
 and 141. In case rolls are weakened by deep cutting, as shown 
 
 FIG. 138.— CG=COLLARS. W= WOBBLERS. 
 
 in Fig. 139, they may be strengthened by stiffeners, D, while 
 long rolls for producing wide plates are sometimes stiffened by 
 an idle roll running on top, lest the springing of the roll make 
 the plate thicker in the middle than at the edges. 
 
 Cast-iron versus Steel Rolls,— Ca,si-iTon rolls are chilled upon 
 ihe outside so as to produce a surface layer of white iron (see Fig. 
 282), which, after turning in a lathe, makes a very smooth sur- 
 face for rolling and is especially advantageous for finishing-mills. 
 They are not so good for the mills which do preparatory work, 
 however, because they are not so strong, and because in pre- 
 
188 
 
 THE METALLURGY OF IRON AND STEEL 
 
 paratory work we want a rough surface to assist in gripping the 
 metal and drawing it through. Furthermore, they cost more to 
 turn to the desired shape, and they cannot be turned down many 
 times (see page 336), lest we get below the "chill." The greater 
 cheapness of cast iron over steel, however, counteracts these 
 factors of higher cost. Where the rolls must be very strong and 
 yet not too large in diameter, and for sharp corners, which would 
 crumble if made of cast iron, steel rolls are often used. The 
 steel employed should be high in carbon — say 0.50 to 0.75 per 
 cent.; but any case-hardening of steel is useless here, because the 
 
 A AND C = OPEN PASSES. 
 B- CLOSED PASS. 
 
 B — 
 
 C = DIAMOND PASS. 
 
 FIG. 139. 
 
 heating of the roll by the material passing through will soon 
 draw their temper. This heating cannot be prevented alto- 
 gether, though it is customary to have a stream of water 
 flowing over the rolls. Sometimes nickel-steel rolls are used 
 for strength. An analysis of roll metal of a very large American 
 company is: 0.40 to 0.50 per cent, carbon, 0.65 per cent, man- 
 ganese, 3.25 per cent, nickel, and 0.15 to 0.20 per cent, silicon. 
 
 Diameter of Rolls. — With smaller rolls, the amount of power 
 consumed is less because the area of metal under vertical pressure 
 is less. There is a limit below which the diameter cannot go, 
 however, either because the rolls will not be strong enough to 
 give the desired pressure, or they will not grip the bar. In 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 189 
 
 order to be gripped the upper and lower edge of a piece must 
 touch the rolls at a point not more than 30° from the center 
 line of the two rolls (see Fig. 140). Every effort is made to use 
 smaller rolls, because the size of all the mills is regulated by them. 
 The surfaces of all but the finishing-mill are usually "ragged" 
 (i.e., made, rough, Fig. 141), to make the rolls give a better grip. 
 Those to receive the ingots are 
 ragged the most, with deep in- 
 dentations somewhat like the 
 cogs of cog-wheels, whence the 
 name of " cogging rolls " for this 
 mill.^" The next trains, known 
 as the_" roughing rolls," are also 
 deeply marked, but even then 
 the piece must come within the 
 30° line, or time is lost in trying 
 to make them bite the piece. 
 
 Speed of Rolls. — ^The more 
 work "the rolls do, the slower 
 must they revolve, because the 
 piece entering the train gives 
 a shock to the mechanism that 
 is dependent upon the power 
 exerted and the momentum of 
 the moving parts. Thus (1) 
 the larger the pieces treated (2) 
 
 the colder they are, and (3) the larger the rolls, the slower must 
 be the speed. In America, speeds are at the high limit. Revers- 
 ing slab-mills may do the work at 20 or 30 r.p.m.; three-high 
 blooming rolls may run over 50 r.p.m.; mills for finishing rails, 
 100 r.p.m,; and rod mills from 550 to 1200 r.p.m. )^ 
 
 Making of Rolls, — Cast-iron or steel rolls are cast in ajpproxi- 
 mately the desired shape and then turned accurately inA lathe, 
 being fitted exactly to a templet when completed. After rolling 
 some thousand tons of material, they become worn and.|produce 
 too large a size of finished shapes. They may then be %sed for 
 a larger size of the same kind of article by putting them back in 
 
 1 In America, the train that produces blooms (i.e.» pieces of steel usually about 6 to 8 1/2 
 in. square) from, ingots, is sometimes, butnot always, known as "bloom rolls," or "blooming 
 rolls," instead of cogging rolls; and the train that produces "slabs" (i.e., thick, wide, rec- 
 tangular pieces that are to be rolled into plates) from ingots is known as the "slabbing- 
 miU." 
 
 FIG. 140. 
 
190 THE METALLURGY OF IRON AND STEEL 
 
 Screw dowa mechanism. 
 
 FIG. 141.— INDENTATIONS ON A COGGING MILL. 
 
 FIG. 142. 
 B, Coupling boxes, C, collars; EEE, rollar table engine; F, fingers, or boms on manipulator; 
 H, housings; HC, housing cap; M, msmipulator or '.'Go-devil"; R, roll; S, spindle; T, 
 roll table; TB, table rollers. 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 191 
 
 the lathe and turning to another templet. For example, a roll 
 for a 20-in. I-beam, with a certain thickness and width of flange, 
 may be converted to one for a 20-in, I-beam with thicker web 
 and longer flange. 
 
 The MilL—The different parts of a rolling-mill may be seen in 
 Fig. 142. The wobblers_ (Fig. 143) are made of the same cross- 
 section as the spindle, some examples being shown in Fig. 138. 
 The coupling boxes (Fig. 142) fit over the spindle and wobblers, 
 
 FIG. 143. 
 
 B, coupling boxes; C, collars; E, roll engine; G, guides; H, housings; S, spindles; TR, table 
 
 roller; HC, housing cap; W, wobblers; TM, table motor. 
 
 so that neither can turn without the other. In some mills the 
 coupling-boxes are made of cast iron in order that, if any shock 
 comes upon the driving mechanism, the boxes shall give way and 
 relieve the strain. In other mills the boxes are made of cast steel, 
 as it is thought that the constant delays due to broken couplings 
 are more costly than breakages in other parts of the mill. The 
 spindles (Fig. 143) are at least twice as long as the coupling- 
 boxes, in order that they may carry both of them at once when 
 the train is uncoupled. Both boxes slip back upon the spindle. 
 
192 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Pinions (Fig. 144) are now usually made of steel for the sake 
 of strength. Housings (Fig. 143) are made of either steel or of 
 cast iron, depending on the strains to which they are subjected 
 and the opinion of the manager. In America, they are usually 
 
 FIG. 144.— A PINION. 
 W= wobbler. 
 
 made so that the top can be removed and the whole train of rolls 
 removed at once, together with, the chocks, and several mills 
 have spare sets of rolls all made up ready and carried in a sling, 
 so that a new set may be dropped into place with a crane with 
 
 the least possible delay to the 
 
 mill. Delays in rolling-mills are 
 
 very costly, because of the idle 
 
 Coiiar__L^^^^^^^^ \ labor and capital, and because 
 
 other parts of the plant may be 
 delayed thereby. 
 
 The screw-down mechanism 
 (Figs. 141, 146) which adjusts 
 the distance between the rolls 
 Co llar f. ^^^^^^;>M(:\ \ I is operated by hydraulic pres- 
 sure or by electric motors. It 
 is connected with a telltale gage 
 which advises the roller exactly 
 FIG. 145. as to the distance separating the 
 
 rolls. 
 Guards (Fig. 145) are of steel and serve to peel the piece off the 
 roll and prevent it encircling the roll (called "collaring") in 
 case it becomes wedged between the collars. They may be upon 
 the lower roll, as shown in Fig. 146, or upon the upper roll, and 
 counterbalanced to hold them in position, when they are called 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 193 
 
 "hanging guards." Guides (Fig. 147) are on the opposite side 
 of the train, and'assist in conducting the piece straight into the 
 groove. 
 
 FIG. 146. 
 
 A, hanging guards; B, coupling boxes; H, housings; P, pinions; PH, pinion housing; R, 
 
 rolls; S, spindles. 
 
 Roll Tables. — Heavy pieces are handled at the rolls by sup- 
 porting them upon a series of rollers, situated in front of and 
 
 FIG. 147. 
 G, guides; HC, housing cap; PH, pinion housing; RR, rolls. 
 
 behind the roll train, and known as the "tables." (Fig. 148.) 
 At two-high mills the tables are stationary; at three-high mills 
 
 13 
 
194 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 148. 
 D, Screw-down mechanism; EEE, table engine; TM, Table motor; TR, table roller. 
 
 . c- ■ ■ 
 
 \ 
 
 1L 
 .-T^ 
 
 i 
 
 _^ 
 
 t^ 
 
 
 " ui 
 
 J 8 "^ 
 
 
 '"" " 5" 
 
 1^. 
 
 ^..-■^; '1 
 
 ^' ■St; A. 
 
 
 
 T- 
 
 
 
 
 FIG. 149.— TWO-HIGH, REVERSING UNIVERSAL MILL. 
 VVV, Vertical rolls; RR, Horizontal rolls. 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 195 
 
196 THE METALLURGY OF IRON AND STEEL 
 
 the front and back tables are sometimes raised and lowered 
 together by hydraulic or electric mechanism, and sometimes 
 they are pivoted near the middle, so that the end next the rolls 
 can be tilted upward in order to bring the piece between the 
 guides which direct it into the groove. The rollers to handle 
 large pieces are "live," that is, they are made to revolve by 
 electric motors and thus move the piece back and forth. "Dead" 
 rollers are used where- pieces are to be moved by hand. 
 
 Transfer Tables. — Roller tables are sometimes made so that 
 they may be moved bodily from one roll train to another, carrying 
 the^ piece of metal with them, and so connected electrically that 
 the roller can be caused to revolve when the table is in any loca- 
 tion. 
 
 Manipulators (Fig. 142).— If two or more posts, supported on 
 a carriage which can be moved laterally, project between the 
 rollers of a table, their sidewise motion will transfer the piece 
 from one pass to another. If the table is of the lifting type, 
 the posts, or "horns," or "fingers" can be brought to such a 
 
 FIG. 151 
 
 position that the lowering of the table will bring the edge of the 
 piece upon the horns and thus tip it on to the other side. This 
 form of manipulator is much used at three-high blooming rolls, 
 and is very efficient and rapid in its work. The same type is 
 used at reversing blooming rolls, but the piece is more usually 
 tipped over by the roller with the tool shown in Fig. 151. 
 
 Roll Engines. — The service on rolling-mill engines is very 
 severe, because the full load comes upon it when the piece enters 
 the rolls, and then leaves it as suddenly again. To equalize these 
 sudden variations of power, all but the reversing engines are 
 built with very large and heavy fly-wheels and run at a high rate 
 of speed (from 30 to 250 r.p.m.), with governors of a quick-acting 
 
THE MECHANICAL TREATMENT OF STEEL 197 
 
198 
 
 THE METALLURGY OF IRON AND STEEL 
 
 type. The Allen engine with the Porter governor serves these 
 purposes, and the Porter-Allen type is much used. The ordinary 
 slide-valve is used on the smaller engines. Corliss valves are com- 
 moner in America for engines doing heavy work (1000 to 3500 H. P., 
 
 FIG. 153.— UNIVERSAL MILL. 
 
 E, Engine for horizontal rolls; EE, engine for vertical rolls; EEE, roller table engine; TT, 
 
 roll tables; TR, TR, table rollers. 
 
 while piston-valves are favored in England. The fly-wheel is 
 placed upon the crank-shaft, to which the roll train is directly, 
 coupled. The fly-wheels very exceptionally weigh as much as 
 75 and 100 tons or more. 
 
 FIG. 154.— MOTOR-DRIVEN ROLLING MILL. 
 
 Piston-valves are used almost always for reversing engines 
 which are compounded, so they may never come to rest at a dead 
 point. There is, of course, no fly-wheel, and the engine is directly 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 199 
 
 coupled from the crank-shaft to the roll train in the large American 
 mills, but is geared down so that the engine can develop a higher 
 speed than is desired for the rolls, thus requiring less power, 
 Eeversing slabbing-mill engines have capacities up to 25,000 H.P. 
 each. 
 
 Electric-motor Drive, — In many mills in America and Europe 
 electricity has replaced steam for all power purposes. The ad- 
 vantages of electricity over steam are a lower operative cost, 
 greater security of operation, fewer breakdowns, and a more 
 
200 
 
 THE METALLURGY OF IRON AND STEEL 
 
 flexible relation between the prime mover and the load, the 
 result of .the electric motors receiving a sudden shock more elas- 
 tically. On the other hand, the advantage of steam, is that, 
 although it receives the load less elastically, it adjusts itself 
 
 quicker and better to the extreme variations in load that always 
 occur in rolling-mills. This is especially true of reversing mills. 
 As already noted, the smaller the mill the less will be the load, 
 and therefore the variation in load. Consequently, in England, 
 Sweden and Germany there are many motor-driven roll trains of 
 
THE MECHANICAL TREATMENT OF STEEL 201 
 
 the smaller size, and a few up to several hundred horse-power, in- 
 cluding one reversing motor of 1200 H.P. Even large mills 
 are now most commonly operated by electric power, if of recent 
 installation, aiid this applies even to reversing mills. One reason 
 for the extent of this change is of course the cheapening of the 
 cost of electricity due to the better utilization of the waste 
 power of blast furnaces at large plants by the installation of gas 
 engines. 
 
 RoLLiNG-MiLL Practice 
 
 Troubles in Rolling.— Hhere are more difficulties met with in 
 rolling-mill practice than we can discuss here, but it may be said 
 that the seriousness of the difficulty is estimated almost altogether 
 in proportion to the delay it causes in the operation of the mill, 
 rather than in the loss of a small amount of material or of a part 
 of the mill itself. For example, the breaking of a table engine, 
 roller, or even a roll, is regretted more because of the time neces- 
 sary to put in a new one than because of the loss of the part. 
 This is one reason why electric motors to operate the tables 
 have, in many cases, replaced small steam engines. The same 
 conditions have also resulted in different parts of the mill being 
 made interchangeable. In many mills it is customary to have 
 spare table engines, or^ motors, etc., always ready, and the least 
 accident to one of these machines would result in its being im- 
 mediately replaced by a whole new one. 
 
 The most important common troubles in roUing-mill opera- 
 tions, probably, are: (1) Bending and breaking of the rolls, due 
 to their being placed under too severe a strain, 
 either because the draft is too heavy of* because 
 the piece is cooled too much, (2) fins caused by 
 metal being squeezed out between the collars of 
 the rolls, as shown in Fig. 157; these fins, besides 
 spoiling the material, are liable to break the rolls; (3) collaring. 
 
 Rolling Plates, — In the rolling of plates an ingot, usually of 
 open-hearth steel and weighing 2 to 10 tons, is first cogged down 
 in the slabbing-mill, producing a long, flat piece of metal. The 
 slabbing-mills are frequently of the two-high, reversmg, uni- 
 versal type. The front end of the piece is cut off in a huge hy- 
 draulic or electric shear to remove the pipe and then it is cut up 
 into slabs of the desired size or into slabs of a size such that each 
 one will make one plate. The slabs are then transferred to the 
 
 FIG. 167.— FIN. 
 
202 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 158.— TWO-HIGH PLATE MILL. 
 
 FIG. 159.— THREE-HIGH PLATE MILL. 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 203 
 
 heating furnace, heated to about 1300° C, and rolled in a three- 
 high or, more rarely, a two-high reversing plate-mill, in some 
 cases there being a pair of vertical rolls to keep the edges straight. 
 During the rolling a shovelful of salt is occasionally thrown upon 
 the surface of the plate, which carries in between the rolls some 
 
 FIG. 160.— THREE-HIGH PLATE MILL, TWO STANDS OF ROLLS. 
 
 B, Coupling boxes; D, screw-down mechanism; E, roll engine; H, housings; P, pinions; 
 
 PH, pinion housing; RR, rolls; S, spindles. 
 
 of the water which is always trickling over them to keep them 
 cool. Sand may be used for the same purpose, and in England 
 heather is sometimes used. As soon as this water is pressed 
 against the hot plate it is converted into steam, causing a rapid 
 series of explosions which blow the scale off the upper surface of 
 
 FIG. 161.— OTHER VIEW OF FIG. 160, SHOWING METHOD OF RAISING THE 
 ENDS OF THE TABLES NEXT TO THE ROLLS. 
 
 the plate and give it a smoother finish. As the process continues, 
 the operator tests the thickness of the plate with a gage, and 
 when it is of the desired thickness, it is passed up to the straight- 
 ening rolls and then to a cooling table, being marked with a 
 distinguishing mark on the way to indicate the heat of steel from 
 
204 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 162.— PLATE STRAIGHTENING ROLLS. 
 
 
 H'.— J 7. -Ml', , ^ 1,1 ... ....l'^ J 1 
 
 1 i'.^^p^VjMi« 
 
 
 / ' ■.■' ' ': 
 
 ^^ 
 
 FIG. 163.— PLATE SHEARS. 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 205 
 
 M^ = 
 
 U-LWlM 
 
 
 Mi 
 
 FIG. 164. 
 
 FIG. 165. 
 
206 
 
 THE METALLURGY OF IRON AND STEEL 
 
 which it was manufactured. When cooled it is sheared to the 
 desired size and shape. The weight of finished plate will prob- 
 ably be not more than 80 per cent, of the weight of the steel sent 
 to the rolling-mill in the form of ingots. 
 
 Rolling Rails. — An ingot of about 3 tons in weight is sent 
 to the rolling-mill, where it is kept in the heating furnace for 
 50 minutes or more until the interior is entirely solid and it is of 
 a uniform temperature throughout. It is then rolled into 
 blooms, either in a three-high mill, such as shown in Fig. 142, 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 207 
 
 or in a two-high reversing mill. In the three-high mill, an ingot 
 18 1/2 in. square at the middle (tapering about 1/8 to 1/4 in. toa 
 foot in order that the mold may be more easily removed) will be 
 reduced to a bloom of about 8 in. square in nine passes, the 
 amount of reduction in each pass being about 12 to 18 per cent. 
 
 , of the original area. The top 
 end is then cut off to remove 
 the pipe, the bottom end to 
 remove the irregularity due to 
 
 FIG. 167. 
 
 FIG. 168.— MANDRIL. 
 
 the rolling, and the piece cut in two to make two blooms. The 
 blooms are then generally reheated in a heating furnace and passed 
 through the series of changes shown in Fig. 165, until they have 
 assumed the proper size and form, the greatest amount of draft 
 being usually not more than 22 per cent., except upon the middle 
 
 FIG. 169.— PIPE-WELDING ROLLS. 
 
 portion of the web. In some cases the blooms are not reheated, 
 but go directly from the bloom rolls to the first roughing train. 
 This makes the metal crack more in rolling, however, and these 
 cracks will ultimately show as a mark on the finished product, 
 which causes the rails to be classified by the inspector in the 
 second or third class. Railroads will accept only 5 or 10 per cent. 
 
208 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 170.— PIPE-WELDING ROLLS WITH MANDRIL IN POSITION. 
 
 I 
 
 FIG. 171. 
 
THE MECHANICAL TREATMENT OF STEEL 209 
 
 of their order in second-class rails, while third-class rails are not 
 acceptable and must go into the tracks of the steel company 
 itself or else be reheated and rolled into smaller sizes, whereby 
 the marks will often be eliminated. If the blooms are reheated 
 before going to the roughing train, many of the cracks formed 
 during blooming will be seemingly closed up, or in any event 
 will not show. Furthermore, if this reheating is to take place, 
 the ingots need not be heated so hot in the first instance, and 
 therefore will not be so tender and so liable to crack. 
 
 Making Lap-welded Tubing. — The wrought iron or steel low 
 in carbon is first rolled out into skelp about 20 to 25 ft. long, and 
 of a width a little more than three times the intended diameter of 
 the tube. The skelp if large is then rolled up into a rough form 
 of a pipe, as shown in Fig. 167, by passing it sidewise through 
 rolls, which bend it roughly to the shape of a pipe, with edges 
 overlapping. The same is done in the case of small 2- to 8-in. 
 tubes, by drawing them through a die. It is then passed at a 
 welding heat through a pair of rolls, with the seam that is to be 
 welded upward. Between the rolls is a mandril on the end of a 
 long rod and of the size of the inner diameter of the tube (Fig. 168). 
 The rolls press the two parts of the weld together 
 over the mandril, and the pipe, after another 
 rolling to give true size and after straightening 
 and testing, is ready for service. 
 
 Making Seamless Tubes. — Seamless or weldless 
 tubes are made either by distorting a steel plate 
 between dies, as shown in Fig. 171, or else by 
 piercing a hole through the celiter of a hot steel 
 billet and then rolling it successively between j,jq i72_sec- 
 rolls over a mandril. The hole is sometimes first tion of bell. 
 ^ of small size and then expanded by pressing 
 larger and larger expanders through it. The pierced billet is 
 then rolled over mandrils constantly decreasing in size until the 
 inner and outer diameters are broiight to the desired size. 
 
 Butt-welded Tubes. — Butt-welded tubes are made by heating 
 the skelp to a welding temperature and then drawing it out of the 
 furnace through a bell, as shown in Fig. 172, which curls it up and 
 welds the edges together, without lapping. Butt-welded tubes 
 are not so strong as lap-welded tubes, and are not usually used for 
 boilers or high pressures, or where they will be expanded much by 
 heat during service. They are mostly made in the small sizes. 
 
 14 
 
210 THE METALLURGY OF IRON AND STEEL 
 
 Wire Drawing 
 
 Wire is a product formed by being drawn cold through a die. 
 The commonest shapes are "rounds/' and the next, hollow tubes, 
 but a great variety of forms may be produced at will. 
 
 FIG. 173.— WIRE ROD ROLLING TRAIN. 
 These mills will roll steel down to from 7/8 to 1/2 inch rods, which are then drawn into wire. 
 
 FIG. 174.— WIRE ROD FRAME. 
 
 Effect of Drawing. — The effect of the drawing is to produce a 
 very exact size of material and to increase the strength, hardness, 
 and brittleness of the metal. In the drawing of steel, the crystals 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 211 
 
 of the metal are actually pulled apart and flow by each other, the 
 outer layers of the metal being dragged back over the central core, 
 there being at the same time a pressure exerted in all directions 
 toward the center, which results in a certain amount of backward 
 flowing even there. Because the crystals are so broken up dur- 
 ing the operation, and because the metal is never heated above 
 its critical temperature during annealing, the grain of the steel 
 is very fine and the crystals are intimately mixed, which is prob- 
 ably the cause of the great strength of wire. 
 
 Annealing. — With each draft the wire becomes harder and 
 more difficult to draw. As it is pulled through the die by a force 
 equal to 40 to 80 per cent, of its 
 tensile strength, it is necessary to 
 soften it at intervals by annealing, 
 lest it break. The annealing is ac- 
 complished by enclosing the wire in 
 some receptacle that protects it • 
 from oxidation' and th^n heating 
 to a low-red heat. In the case of 
 steel, it is required after every eight 
 to three passes, depending upon the 
 amount of carbon in the metal and 
 the amount of "draft"; i.e., pro- 
 portionate reduction in size. 
 
 Dies, — Wire dies are usually fig. 175.— section of wire die. 
 made of high-carbon steel (say 
 
 about 2 per cent.), through which a tapered hole is made, as 
 shown in Fig. 175. The object of using this material is that, as it 
 becomes worn in service, it can be reformed and used for larger 
 sizes, which could not be done with white cast iron. 
 
 Bench. — ^A ''bench" on which wire is drawn consists of a 
 reel which holds the coil of undrawn wire, a die support, and a 
 second reel which draws the wire through the die and coils it up, 
 and which is driven by bevel gears. The die rests against the 
 support, and the wire, having a tapered point, is thrust through 
 the hole and grasped by a pair of tongs, which pulls it out until 
 it can be attached to the reel. This is then set in motion and 
 draws the wire through. The die-holder is heaped up with 
 lubricant of some kind, in order that the metal may pass more 
 easily through the hole. The speed at which wire is drawn will 
 vary from 75 to 750 ft. per minute, depending upon the size and 
 
212 THE METALLURGY OF IRON AND STEEL 
 
 hardness of the material drawn and the amount of reduction 
 during each draft. In many cases there is more than one die, 
 and the wire passes successively through two, three or more, 
 being constantly reduced in each one. Between each pair of 
 dies is a reel, around which the wire passes two or three times, 
 since the strength of the wire emerging from the last die would 
 not be sufficient to draw it through all of the holes. 
 
 Draft. — The heavier the draft the greater is the hardness 
 produced in the wire and the greater the wear of the dies. 
 The average amount of draft will probably be from 20 to 25 
 per cent. 
 
 Drawing Tubes. — Hollow wire or small tubes are drawn some- 
 times over a mandril. This mandril may be a wire of about 
 the size of the inner diameter of the finished tube. After several 
 drafts, the tube is wedged so tightly on the mandril that it cannot 
 be separated. It is then given an unbalanced squeeze between a 
 pair of rolls, so that the tube is reduced in thickness, whereby 
 its diameter is increased .and the mandril may easily be with- 
 drawn. 
 
 Pressing 
 
 Steel may be pressed either hot or cold, the latter method 
 being used chiefly for thin and soft steel, and the former for 
 very large work, such as armor-plate, cannon, etc., for which 
 hydraulic presses have now largely replaced the heaviest steam- 
 hammers. 
 
 Effect of Pressing. — The effect of pressing upon the metal 
 differs from hammering, in that its action extends deeper into 
 the material, thus giving a somewhat superior texture to the 
 interior. Tests cut from the center of large pieces forged under 
 the press are very much superior to those cut from the same 
 place in pieces forged under the hammer. 
 
 Hot-pressing. — ^Presses vary in size Usually from 600 to 14,000 
 tons. They may be either of the continuous or of the inter- 
 mittent type. In the latter, the amount of pressure exerted 
 increases step by step as the work progresses. The amount of 
 work that can be done by the press in large-sized pieces is greater 
 than that done by hammers for the same amount of power used. 
 This results in a double saving of fuel, since more work can be 
 accomplished with one heating. By means of the 10,000-ton 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 213 
 
 press at the Homestead Steel Works, a 50-ton armor-plate has 
 been reduced 2 in. in thickness and moved forward 6 in. for each 
 squeeze, while a 3000-ton press at the Firths Works in England 
 has reduced a 30-ton ingot from 49 to 28 in. diameter in 30 
 piinutes, and from 51 to 26 in. diameter in 65 minutes. Small 
 
 FIG. 176.— FOURTEEN-THOUSAN:p-TON ARMOR-PLATE PRESS. 
 
 pieces can, however, be turned out a little faster under the 
 hammer. 
 
 Cold-pressing. — Thin plate for steel railroad-car construction, 
 and many other purposes, is often formed by pressing it cold 
 between dies under hydraulic presses of from 30 to SOO-tons 
 capacity. In this way bolsters, braces, and many other parts 
 
214 
 
 THE METALLURGY OF IRON AND STEEL 
 
 are formed with great economy. Sometimes two or three presses 
 are required with different dies to complete the shaping, and 
 occasionally it is necessary to press some of the work hot, be- 
 cause the distortion is so great that the steel would otherwise 
 be torn. Cold-pressing is also known as ''flanging." It has 
 one great difference from hot-pressing, in that there is no reduc- 
 tion in the sizes of pieces treated. 
 
 ■-'^-■■^:-... 
 
 X--" .^ '" , dfi*.f ' ^A \ .. ^-'- 
 
 IS— ^ 
 
 
 :-,■,---"-■; J, ' 
 
 ^tes:...,. '^ -'■ 
 
 
 
 ■ ^ 1 
 
 
 FIG. 177.— DROP-FGEGING PRESS FOR PLATES. 
 
 Comparison of Mechanical Methods 
 
 Hot-working with Cold-working.— Cold-wovkmg gives greater 
 strength, a harder material and more accurate finish as to size 
 than any form of hot-working. Furthermore, it produces a finer 
 grain on the surface of the metal. If the cold-working is fol- 
 lowed by annealing at a temperature below the critical range of 
 the steel (see pages 290 to 293), the material retains its fine 
 grain, and is Stronger and more ductile in proportion than metal 
 that has been worked hot. Before cold-working the metal is 
 pickeled in dilute sulphuric acid to remove the scale, and is there- 
 fore produced with a bright surface which is suitable, without 
 machining, for use as shafting, for nickel-plating, etc. The an- 
 nealing is usually effected inside closed vessels, in a reducing 
 atmosphere of illuminating gas or some similar medium, which 
 prevents the formation of scale. Cold-rolled steel is used for 
 shafting and for articles that are to be drawn or stamped to 
 shape — watch and clock springs, hack-saw blades, corset steels, 
 
THE MECHANICAL TREATMENT OF STEEL 215 
 
 etc. Cold rolling, wire-drawing and "flanging" are the com- 
 monest forms of cold-working. 
 
 Hammering^ Rolling and Pressing Compared. — Of all the me- 
 chanical methods, rolling gives by far the largest output per day, 
 per unit of power, and usually, per unit of fuel for heating. It 
 is therefore the cheapest method, especially for labor. It does 
 not work the metal as well as either hammering or pressing, both 
 of which produce a much better crystalline structure, beside 
 affording a better control of the temperature at which the opera- 
 tion is ended. Pressing works the metal at greater depths than 
 hammering, and is therefore especially advantageous for pro- 
 ducing large pieces. In such large pieces the size of the hammer 
 which would be sufficiently heavy to forge the piece all the way 
 into the center of the section, becomes so large that it is un- 
 wieldy. The press turns out a much larger production of large 
 pieces than would a hammer suitable for the same Work, and is 
 less destructive on the dies and other equipment. The press 
 is also used on special forging work, such as drawing, special 
 punching and closed die forging, where it is almost impossible 
 to use a hammer. Where a shape is intricate, rolling is more 
 liable to tear the metal than hammering or pressing because, 
 at the point where the roll is deeply cut, its surface velocity 
 is much less than where the diameter is greater, and thus it 
 tends to drag the metal through at different speeds. 
 
 Heating Furnaces 
 
 Heating furnaces are usually of the reverberatory type, burn- 
 ing soft coal or gas. The flame produced must be as reducing as 
 possible in order to produce a small amount of scale. Much better 
 control is obtained if the ash-pit is enclosed and forced draft is 
 used to burn thefuel. In this way about half a ton of fuel will be 
 required to heat a ton of steel from the atmospheric temperature 
 to that necessary for mechanical work, with a loss of about 4 to 5 
 per cent, of the metal as scale. The gases must necessarily leave 
 the furnace at a high temperature, and therefore it is not uncom- 
 mon to have boilers situated over the heating furnace, in which 
 steam is raised by means of the waste heat. With this economy 
 the amount of fuel chargeable against heating the steel will be 
 from 350 to 450 lb. per short ton of steel heated. 
 
 Regenerative Furnace. — If the heating furnaces are fired with 
 
216 
 
 THE METALLURGY OF IRON AND STEEL 
 
THE MECHANICAL TREATMENT OF STEEL 217 
 
 producer gas and the regenerative method is employed, we get a 
 far better controLof the temperature and of the reducing influence 
 of the furnace gases. By this means a short ton of steel may be 
 heated with from 150 to 200 lb. of fuel and with a loss of metal of 
 from 1 to 5 per cent, by oxidation. 
 
 Continuous Furnaces, — Billets and other small pieces may be 
 heated in furnaces whose action is continuous. Such a one as 
 this is shown in section iix Fig. 179. Along the hearth stretch 
 two lines of pipe, which are kept cool by a stream of water inside. 
 Upon the pipes is laid a long series of billets, which are gradually 
 moved forward toward the end at which the gas and air enter. In 
 this way the flame is always met by colder material and finally 
 leaves the furnace at a relatively low temperature. As the gases 
 pass out, they go through a series of pipes, Bj B^ around which 
 circulates the air that is afterward led to the fire and used for 
 combustion. As soon as tjie billet nearest the fire end is heated 
 to the desired temperature, a new one is pushed in at the bottom, 
 causing the hot billet to be shoved onto the inclined plane, whence 
 it rolls out of the furnace to the point A, whence it is transferred 
 to the rolling-mill. In this type of furnace a short ton of steel 
 may be heated with from 120 to 145 lb. of fuely with a loss in 
 weight of less than 1 per cent, by oxidation. 
 
 Soaking-pits. — Ingots with molten interiors must be put in 
 some form of the furnace in which they will stand upright until 
 they have solidified throughout and are ready to roll, in order 
 that the pipe may form in the upper portion. The type of 
 furnace used for this is known as a soaking-pit and is shown in 
 Fig. 180. The original intention of soaking-pits was to have 
 the heat in the ingot itself bring the interior of the furnace and 
 the mass of metal to the desired temperature; but this is not 
 found practicable in the United States, and soaking-pits are usu- 
 ally heated by regenerated gas and air. The ingots must be 
 kept in these soaking-pits long enough to be entirely solid in the 
 interior, and for this purpose at least 55 minutes are required for 
 3-ton ingots when stripped and charged as soon as possible after 
 teeming, and more for ones of larger size. 
 
 Furnace Bottoms. — Heating-furnace bottoms must be of some 
 material not readily corroded by oxide of iron scale, and basic 
 bottoms are very commonly employed with success. Heating- 
 furnaces are also frequently supplied with a tap-hole, from which 
 the slag, composed chiefly of oxide of iron, can be tapped at 
 
218 
 
 THE METALLURGY OF IRON AND STEEL 
 
THE MECHANICAL TREATMENT OF STEEL 
 
 219 
 
 intervals. Soaking-pit bottoms are frequently covered with a 
 layer of coke breeze to absorb the slag and prevent corrosion of 
 the furnace bottom. This is shoveled out when an opportunity 
 is afforded, and new breeze substituted, or is knocked out through 
 a hole in the bottom, for which see Fig. 180. 
 
 Heating Practice. — In heating steel for rolling, the lower the 
 temperature the better will be the quality of the product. On 
 the other hand, if the metal is to undergo many passes before it 
 receives another heat, it must be correspondingly hot, in order 
 
 FIG. ISO.— BRICKWORK OF REGENERATIVE GAS SOAKING-PIT. 
 
 that the finishing temperature may be high enough to avoid ex- 
 cessive power for reduction. There is no doubt that rolling 
 temperatures at the present time are higher than they should be, 
 for the metal when finished should be only just above the critical 
 temperature of the steel. Until within recent years no suitable 
 pyrometers for measuring the temperature have been available, 
 and the temperature for drawing the material is judged by eye, 
 so that no figures can be given. There can be no doubt that more 
 careful attention to this point will result in less waste in rolling 
 (on account of the production of cracked or second-class material 
 because of too hot steel at the starting) and in the production of 
 a higher quality of steel. 
 
220 
 
 THE METALLURGY OF IRON AND STEEL 
 
 It must be remembered, however, that the steel must be hot 
 enough to cause it to weld together wherever it has become 
 cracked. This is especially to be observed in low-carbon steel 
 whose welding, as well as its melting-point, is higher than that 
 of high-carbon steel. The casting temperature and the absence 
 of ingotism which the author has discussed elsewhere! is probably 
 
 FIG. 181.— CHARGING A SLAB INTO A HEATING FURNACE. 
 
 more important than any other factor in preventing cracking 
 during rolling, as properly made steel can stand without injury 
 a high temperature which would be very harmful otherwise. 
 High-carbon steel is very delicate to roll especially when the 
 silicon also is high. 
 
IX 
 
 IRON AND STEEL FOUNDING 
 
 Founding is a mechanical art, and consists in pouring melted 
 metal into a mold of any desired size and form, which the metal 
 assumes and retains when cold. The mold is made of some 
 kind of sand, with rare exceptions to be mentioned hereafter. 
 The art is a very complex one, added to which it is now passing 
 through an important transition period in which science is very 
 
 mm^ ^ 
 
 
 1 -^i*l 
 
 
 Z^-1 '-^'.^^ -l^''-^ 
 
 
 
 FIG. 185.- 
 
 Sweeping up a mold. 
 -VIEW IN AN IRON FOUNDRY. 
 
 rapidly takiag the place of rule of thumb. It is impossible to 
 treat the subject adequately in a single chapter, but several 
 books are now available, to which foundrymen, metallurgists 
 and chemists are referred, and which are also recommended to 
 all engineers, to whom a knowledge of the art is of prime im- 
 portance. 
 
 221 
 
222 THE METALLURGY OF IRON AND STEEL 
 
 The Matcing of Molds 
 
 There are various kinds of sand molds made for foundry work, 
 but the three principal kinds are loam molds, dry-sand molds, 
 and green-sand molds. 
 
 Loam Molding. — In molding "^ith loam, sand is usually Built 
 up into the required shape by hand, aided by machines. Fig. 185 
 
 FIG. 186.— MACHINE FOR FORMING THE TEETH OF A BEVEL-GEAR. 
 
 shows the molding of a gear in which the parts are built up of 
 brick and sand and then " swept '^ into the proper shape by means 
 of the wooden sweeps. Large wheels and gears ^re often swept 
 up in this way, the teeth being formed subsequently by means of a 
 small pattern that is moved around as the molder progresses, or 
 by means, of a machine, as shown in Fig. 186. In the case of a 
 gear, the arms are usually formed by placing within the swept-up 
 
IRON AND STEEL FOUNDING 
 
 223 
 
 mold forms of sand known as ''cores/' as shown in Fig. 187. 
 Loam moulding is common in iron foundries, but less often 
 used for steel castings. 
 
 Pattern-molding. — To only a limited class of work is loam- 
 molding applicable, and the commonest manner of making a mold 
 is to press or ram sand around a pattern, which is subsequently 
 removed, leaving the desired cavity. Usually the pattern is en- 
 
 FIG. 187.— PLACING COKES IN A MOLD. 
 
 closed by a " flask ^' much larger than itself, between which and 
 the pattern the damp sand is rammed. The pattern (sometimes) 
 is split into halves, one half being in the lower part, or " drag," of 
 the flask, and the other half being in the upper part, or ''cope."^ 
 The cope is now taken off and turned upside down, after which a, 
 lifting-screw is inserted into each half of the pattern in turn, by 
 means of which it is drawn from the sand; and when a " grate " is 
 cut through the cope, the flask is again fastened together, and a 
 receptacle is formed of the shape of the pattern into which the 
 metal may be poured. 
 
 1 The old English word " cope," meaning a covering for the head, which has now largely 
 been replaced by the name " cap." 
 
224 
 
 THE METALLURGY OF IRON AND STEEL 
 
 M 
 
 
 
 
 cfl 
 
 
 o 2 
 
 
 hJ O 
 
 
 ii 
 
 
 
 w « 
 
 
 ffl s 
 
 
 H N 
 
 
 0) 
 
 
 o ^ 
 
 
 IS^ 
 
 
 S^s 
 
 
 
 t-i d 
 
 [V, fc< 
 
 Ph 
 
IRON AND STEEL FOUNDING 
 
 225 
 
 The art does not consist of these simple operations alone, 
 however, for in drawing the pattern from the sand, even though 
 the lifting-screw be lightly tapped with a hammer in four hori- 
 zontal directions to loosen the pattern, the slightest tremble of the 
 molder's hand, or of the crane used for lifting, may cause the sand 
 to be broken in places, and the chief skill of the molder as well as 
 a large share of his time is employed in repairing the damage thus 
 produced. Furthermore, the mold may be "washed," that is, 
 
 yii^yyy^^^^^^A^A^^MM^^^^^ 
 
 
 
 FIG. 192,— PATTERN IN SAND. 
 
 painted inside; the proper cores must be put in place; parts of the 
 sand liable to drop off must be nailed in place with thin, large- 
 headed wire nails thrust in with the thumb; before the pattern is 
 taken from the sand the cope must be "vented," that is, made 
 porous, by jamming a wire into it many times and pulling it out 
 again, so that the air and gases will escape when the metal is 
 poured in; and so on. 
 
 Furthermore, it may readily be imagined that parts of the 
 pattern as shown in Fig. 192 might be of such a shape, with flanges 
 on the bottom, or something of that kind, that they could not be 
 drawn without breaking the sand. In the case of such a design 
 
 15 
 
226 
 
 THE METALLURGY OF IRON AND STEEL 
 
 the pattern and flask must be split into three or more parts,* or 
 else a core must be put in to make an offset. It will be evident to 
 every engineer that he will have to pay more for making a casting 
 so designed. (See Fig. 193.) 
 
 Ramming, — In pattern molding, it is essential that the pres- 
 sure of the sand around the pattern shall be nearly uniform in 
 all places; because (1) when the metal is, poured into the mold, it 
 
 Cope 
 
 Che^ 
 
 >5v! Cheet 
 
 FIG. 193.— SECTION OF FLASK AND PATTERN. 
 
 drives out the air already there by forcing it through the inter- 
 stices between the particles of sand, and if the sand is too hard in 
 any place, the pressure of air collected there is liable to form a 
 depression, or "scab," in the casting; and because (2), if the sand 
 is too loose in any place, the pressure of the metal upon it is 
 liable to ''swell" it outward and thus cause an enlargement of 
 the casting at that point. To obtain uniformity it is necessary 
 
 1 The bottom and top parts being still known as the ' 
 while the intermediate parts are known as " cheeks.' ' 
 
 drag " and " cope," respectively, 
 
IRON AND STEEL FOUNDING 
 
 227 
 
 FinUhiog Trowel 
 
 Finishing Xrowel 
 
 Lifter 
 
 Bench Lifter 
 
 Yankee 
 
 Finishing Txowel 
 
 Sqjiare. Trowel 
 
 Slick and Spoon 
 
 FIG. 194. 
 
228 THE METALLURGY OF IRON AND STEEL 
 
 that the sand be packed around the pattern, and not the pattern 
 pushed into the sand. This packing is accomplished by the hands 
 for the sand immediately adjacent to the pattern, and by rammers 
 for the layers further distant. In the case of bench molding 
 hand rammers are used, and for making larger molds on the floor 
 long iron rammers are employed. The molder's skill is shown in 
 applying the proper amount of power in ramming each different 
 kind or part of pattern. 
 
 Dry-sand Molds. — After ramming up the mold, drawing the 
 pattern and applying the "wash," the mold may be used green 
 (when it is called a "green-sand mold"), or it may be put in the 
 ovens and dried (when it is called a "dry-sand mold"). The 
 drying has the effect of driving off the moisture and leaving a 
 firm, hard mass, semi-baked into a sort of brick. Sand for these 
 molds should have slightly more clay than for green-sand 
 molds; otherwise, instead of baking into a hard mass, they would 
 be liable to crumble with the heat. The temperature of drying- 
 ovens should be about 350° to 400° F. (170° to 200° C), and 
 they are heated by coke, coal, gas or oil. If the temperature 
 is too high, the mold will be burned, that is, it will crumole under 
 the fingers after drying; if not hot enough, the mold will not be 
 baked hard. The length of time in the oven will depend upon 
 the size of the mold, and will vary from an hour or so to a day or 
 so. During the drying the molds are liable to shrink somewhat, 
 due to the action of the clay in binding together. 
 
 Green-sand Molds. — Green sand requires less clay than dry 
 sand, because it has a certain coherence due to its dampness. 
 Many natural sands are found suitable for both the green-sand 
 mold and the dry-sand mold, or they can be made up as desired 
 by mixing a good clay with a sand rich in silica. Green-sand 
 molds xaust not be rammed as hard as dry sand, so the moisture 
 may more readily evaporate. 
 
 Washes. — ^For iron castings the common wash is graphite 
 dust, which is made up into a paint with water and applied with a 
 brush or dauber to the inside of a dry-sand mold before it goes to 
 the oven. In the case of a green-sand mold, pulverized coal is 
 dusted onto the surface through a piece of cloth, and then spread 
 uniformly with the "slicker." In the case of a dry-sand steel 
 casting, the wash is composed of pulverized silica rock, running 
 from 98 to 99 per cent, silica, which is made up into a thick paint 
 with water, thickened with molasses, and applied to the inside of 
 
IRON AND STEEL FOUNDING 
 
 229 
 
230 THE METALLURGY OF IRON AND STEEL 
 
 the mold with a brush or dauber before the mold is dried. -Green- 
 sand molds for steel castings cannot ordinarily be washed. 
 Graphite washes cannot be used for steel molds, because the hot 
 metal attacks the graphite and becomes rough upon its surface. 
 
 The functions of washes are: (1) To make a very smooth face 
 on the sand, so that the surface of the casting shall be smooth 
 (this they accomplish by the very fine size of their particles) ; (2) 
 to give a surface that shall resist the melting and chemical action 
 of hot metal, so the castings may be easily cleaned of sand. 
 
 Skin-dried Molds. — The inside surface of green-sand molds is 
 occasionally dried by painting or spraying it with some inflam- 
 mable liquid, such as gasolene, and then applying a match. This 
 is more common in steel-foujidry than in iron-foundry practice, 
 and produces a little better surface to the casting. Herbert B, 
 Atha patented a mold wash for green-sand steel castings, which 
 would enable them to be skin dried. ^ The formiila for this wash 
 was devised by me with the aid of Parker C. Mcllhiney. It con- 
 sists of ordinary gasolene in which is dissolved as much rosin as it 
 will take up without warming. The rosin increases the specific 
 gravity of the gasolene so that it forms a paint with the silica 
 wash, which would otherwise sink to the bottom and not adhere 
 to the brush. After the wash is applied, it is touched with a 
 match, the burning resulting in giving a dry skin to the mold 
 and leaving it coated with the silica. The rosin does not com- 
 pletely burn off, but binds the sand together and gives a tough 
 skin, so that the sand is not so liable to drop when the cope is 
 turned over to place it upon the drag. 
 
 Dry- versus Green-sand Molds: For Iron Castings. — Dry-sand 
 molds are often cheaper to make and require less molding skill, 
 because (1) the sand does not have to be tempered so carefully, 
 that is, brought to the proper condition of dampness, since the 
 moisture is eventually to be driven off by the drying. In green- 
 sand molds, if the sand is too wet, it is liable to "cut" (be eroded 
 by the stream of metal) and get dirt into the casting, and also 
 to be impervious to the gases. (2) The sand requires less care 
 in ramming, because, whether too hard or too soft, the expansion 
 and contraction in drying will adjust its firmness and porosity, 
 (3) Furthermore, the dry sand is stronger, which is an advantage, 
 when the sand is liable to be under pressure from the metal, or to , 
 have the metal fall upon it from a height. (4) Dry-sand castings 
 
 » U. S. Letters Patent No. 686,189. 
 
IRON AND STEEL FOUNDING 
 
 231 
 
 are also more liable to be sound, because there is less gas in the 
 pores of the sand. 
 
 The disadvantages of dry-sand castings are: (1) The mold is 
 Uable to shrink during drying and therefore be less true to the 
 pattern; (2) the castings are more liable to '' check '* (that is, 
 crack in cooling), because the mold is firmer and so does not 
 give way so easily to the crushing action when the casting con- 
 tracts; (3) molds exposed to the direct action of the flame during 
 
 FIG, 196.— CORES FOR FORMING THE INSIDE OF A GAS-ENGINE CYLINDER 
 
 CASTING. 
 
 drying, or heated too hot, are liable to be burnt and therefore 
 rendered useless, causing a loss; (4) in handling, the molds are 
 liable to be damaged; and, furthermore, it is more costly to 
 repair a dry mold than a damp one, because the adjacent sand 
 must first be damped, the damage repaired, and then a flame 
 applied to dry the wound; (5) it takes longer to complete an 
 order. The actual cost of heat is not very great, and usually 
 is less of an item than the extra labor of handling for drying. 
 For Steel Castings. — Dry-sand steel castings have a surface 
 
232 
 
 THE METALLURGY OF IRON AND STEEL 
 
 much superior to those made in green sand.^ They are also 
 stronger and more liable to be sound. Soundness is much more 
 difficult to obtain in steel castings than in iron castings. Green- 
 sand moldS; however, have the great advantage of allowing their 
 
 FIG. 197.— CORE. 
 
 sand to crush more easily when between two parts of the casting 
 that are being drawn together by the shrinking of the metal. 
 This is doubly advantageous in steel work, because steel shrinks 
 twice as much as cast iron and is therefore more liable to- check- 
 ing. Special means may be employed for making the sand 
 easily collapsible after the metal is poured, such as mixing with 
 
 FIG. 198. 
 
 it an inflammable substance like flour, chopped hay, hay-rope, 
 sawdust, coke, etc., which burns away after the hot metal is 
 poured in. 
 
 Cores. — The function of cores has already been referred to: 
 they are set inside the mold proper to assist in forming the metal. 
 
 ^ One reason that dry-sand steel castings have a better surface is because they can be 
 ■washed. In practice, however, the drying and washing is often improperly performed; 
 
 therefore green-sand castings often have the better surface. 
 
IRON AND STEEL FOUNDING 
 
 233 
 
 The commonest use for them is to extend through a casting in 
 some place to make a hole, as, for instance, the inside of a cylinder, 
 the bore of a pulley, etc. In this position they are subjected to 
 great crushing strain when the metal shrinks, and therefore the 
 
 FIG. 199. 
 
 bond which keeps the sand together, consisting of linseed-oil, or 
 flour paste, a patented core compound, etc., must be of such a 
 nature that when subjected to the heat of the liquid metal it will 
 burn away and allow the sand to disintegrate, which both pre- 
 
 FIG. 200. 
 
 vents it bursting the shrinking casting and permits of its being 
 more easily cleaned out. Cores are often built up around an 
 iron pipe riddled with holes, so that the gases formed may readily 
 escape through this '' vent. " In the case of large cores, the pipe 
 
234 
 
 THE METALLURGY OF IRON AND STEEL 
 
 is frequently wound with hay-rope, or some similar material that 
 will burn away and make the sand more collapsible. Some cores 
 have coke breeze or cinder in the center to make them light as 
 well as porous. 
 
 Cores are supported sometimes by being set in the drag, some- 
 times by being hung in the cope (see Fig. 202) ; but it is more com- 
 mon to have a hollow adjunct to the mold, known as a "core-print" 
 
 FIG. 201— OVEN FOR DRYING SMALL GORES. 
 
 into which an extension of the core fits (see Fig. 190). Some- 
 times both ends of the core are so supported, and sometimes only 
 one end is thus supported, while the other end rests upon a metal 
 chaplet that is absorbed in the casting when the metal is poured. 
 Cores are often dried, lest their gases make the casting unsound 
 or cause it to blow, that is, boil with the rapid escape of gas 
 through the metal. 
 
 Chill Molds.^-lt is often desired to chill certain parts of a 
 mold, or cool them more rapidly than the remainder, in order 
 
IRON AND STEEL FOUNDING 
 
 235 
 
 either to make a thick part of a casting solidify as soon, or 
 sooner, than the thinner portions, or else to produce white cast 
 
 FIG, 202.— CORES HUNG FROM THE COPE. 
 
 FIG. 203.— CHAPLETS. 
 
 iron at that point. The former may be desirable in the case of 
 either an iron or steel casting, becaused the shrinkage cavity 
 
236 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 204.— POSITION OF MACHINE WHEN PRESSING THE DRAG. 
 
 FIG. 205.— POSITION OF MACHINE WHEN PRESSING THE COPE. 
 
IRON AND STEEL FOUNDING 
 
 237 
 
 FIG. 206.— CUTTING SPRUE WITH TUBULAR SPRUE CUTTER. 
 
 I^K 
 
 
 - 
 
 
 
 
 
 . -:/^ 
 
 
 
 r 
 
 f 
 
 ^ 
 
 ^^^^s 
 
 y 
 
 't^^^^l 
 
 w 
 
 i 
 
 f. 
 
 >-^- 
 
 r 
 
 jp^^ 
 
 J n 
 
 
 ^8 
 
 
 -^ 
 
 . ■ p^mskToR--.-. .'^ ^."^^ 
 
 m.- 
 
 
 |ra 
 
 ■; ;%,■, ,^ 
 
 V!tr^'^ 
 
 '■'^'--' q 
 
 1 "^ 
 
 ijM 
 
 
 sfldi 
 
 '^^ 
 
 T- - 
 
 i ^^'ngrm ^•.^ ^ ^— * 
 
 r 1 
 
 ^^ 
 
 I J 
 
 w 
 
 ^;'- """"V 
 
 
 ^- .::^T^Wi 
 
 1 
 
 M 
 
 1^^* J* 
 
 
 
 -.-.^ 
 
 
 ' 'J 
 
 
 ii: 
 
 
 1 . 
 
 m^ _ - , "~- i^r 1 
 
 WM 
 
 ■ 
 
 « 
 
 
 ^ 
 
 1 
 
 IF i^rWKff 
 
 - -^S 
 
 ^ 
 
 ■ 
 
 ^ 
 
 i ■ 
 
 feb. 
 
 L«v ' "'fiir 
 
 - .^^ 
 
 > 
 
 
 
 
 1^ 
 
 1 
 
 ■ "i-f 
 
 
 FIG. 207.— RAPPING THE PATTERN BEFORE SEPARATING THE MOLD. 
 
238 
 
 THE METALLURGY OF IRON AND STEEL 
 
 occurs in the last portion to freeze, and therefore hastening local 
 freezing may be necessary to bring the pipe into the riser or 
 feeder. The latter applies only in iron-casting work in which it 
 is desired to make the outside of a casting very hard. The 
 chilling is usually accomplished by embedding pieces of metal 
 in the sand, against the face of which the casting is poured. This 
 metal is oiled, blackened or "washed," so that it does not stick 
 to the casting. 
 
 Permanent Molds, — ^A great deal of expense in foundry work 
 is due to the fact that a sand mold must be made anew for every 
 
 FIG. 208.— STRIPPING PLATE 
 MOLDING MACHINE. 
 
 FJG. 209.— STRIPPING PLATE 
 MOLDING MACHINE. 
 
 casting, and the subject of permanent molds has occupied the 
 attention of f oundrymen for a great many years without the prob- 
 lem being completely solved. When a casting is knocked out 
 of the mold, the sand is usually knocked out also and its form 
 destroyed. In the rare case of a smooth cylinder, or something 
 of that kind, the casting may be withdrawn without damage 
 except to the face of the sand, and this can sometimes be repaired 
 and swept up anew without re-forming the entire mold. Again, 
 molds for railroad car wheels, which have a metal " chill " all the 
 way round the tread and flange, in order that the cast iron may 
 be white at that point to withstand the grinding action on the 
 rails, have a certain amount of permanency. Finally, molds 
 
IRON AND STEEL FOUNDING 
 
 239 
 
 FIG. 211. 
 
240 THE METALLURGY OF IRON AND STEEL 
 
 carved out of carbon which has been preheated to a very high 
 temperature are said to withstand the action of the melted 
 metal and to last for a large number of castings. 
 
 E. A. Custer has developed a method of casting gray iron in 
 cast iron molds, and such objects as pipe, ornamental cast- 
 ings, etc., are being successfully made in this manner. The 
 molds last a very long time and the principle of the process in- 
 volves arrangements whereby the molds can be opened as soon 
 as the iron has chilled, the cores removed, and the casting taken 
 out of the iron mold to cool more slowly in the air, with the 
 result that the metal, if it has the proper chemical composition, 
 is not chilled but deposits graphite so that the castings may be 
 readily threaded or machined. For certain classes of work, this 
 method would appear to have a great field of usefulness, pro- 
 moting economy, rapidity and a certain degree of independence 
 of labor conditions. It is obvious that the only castings that 
 can be made in this way are ones of which a large 'number 
 are required so that it would pay to make up the necessary 
 molds. The results are also interesting from a scientific stand- 
 point in that they confirm the observation referred to else- 
 where, that graphite separates from cast iron after solidification, 
 as indicated by the circumstance that the iron will be gray 
 if removed from the mold immediately or shortly after solidi- 
 fication, but the carbon will be in the combined condition and 
 .the iron will be white if it is allowed to cool in contact with 
 metal. (See Reference No. 97) . 
 
 Gating of Patterns. — The gate of a mold is the place into which 
 the liquid metal is poured (see Fig. 193) while the sprue extends 
 from the gate to the casting proper. The placing of gate and 
 sprue is goverened by certain principles whose best application 
 can only be learned generally by experience. Referring especi- 
 ally to Fig. 193, we see for example that if the metal were poured 
 into the riser of the casting it would drop upon the core and 
 knock this to pieces. Even if this core were absent, the metal in 
 dropping from such a height onto the sand bottom of the mold, 
 would wash sand from it and make a defective casting because of 
 the cutting of the mold- at this point as well as the probable in- 
 clusion of some of the sand in the solid metal. Furthermore, the 
 metal would spatter on the sides of the mold and frozen drops 
 might accumulate which would afterward form a "cold shut'^ 
 on the side of the casting. It is rarely good practice to pour a 
 
IRON AND STEEL FOUNDING 
 
 241 
 
 casting through a riser, and whenever the metal is poured from 
 a height, it is well to insert a refractory fire brick upon which 
 the stream shall strike. Steel castings are almost always poured 
 from the bottom in a manner in general similar to Fig. 193, and 
 this method is often followed in case of gray iron as well. The 
 general principles of gating may be very briefly summarized as 
 follows: 
 
 (1) The metal must be forced into the mold fast enough to 
 prevent its chilling at any point, so that sprue and gates must be 
 amply large; in the case of a thin casting, for example, this may 
 
 FIG. 212.— ROCK-OVER MOLDING MACHINE. 
 
 necessitate having the metal enter at a number of different 
 points at the side of the mold. 
 
 (2) On the other hand, unnecessary amplitude is to be avoided 
 since the sprues have only the value of scrap. 
 
 '(3) The stream of metal must not wash against the sand in 
 such a way as to cut it. We may exemplify this by an example 
 in steel castings where the stream of metal is usually brought in 
 at the bottom and with a certain tangential direction so that it 
 shall rise in the mold with a rotary motion. 
 
 (4) The metal must not have too far to run. In large or long 
 
 16 
 
242 
 
 Upper head swung back to receive flask. Drag patterns on upper head. 
 Cope patterns on lower head. Yoke swung by power. 
 
 Sand frame on flask. 
 FIGS, 213 AND 214.— VIBRATOR MOLDING MACHINE. 
 
fBffWffff) STEEL FOUNDING 
 
 243 
 
 Mold removed, flask frame raised, showing metbod of drawing cope patterns. 
 
 Mold lowered away, drawing drag patterns. Drag patterns may be returned with absolute 
 
 accuracy. 
 FIGS. 215 AND 216.— VIBRATOR MOLDING MACHINE. 
 
244 
 
 THE METALLURGY OF IRON AND STEEL 
 
 castings this may necessitate gating at several different 
 points : 
 
 (5) As far as possible we must arrange to pour into a reservoir 
 and to have the rate of pouring such that this reservoir shall 
 be kept as nearly full as possible in order that any slag or dirt 
 accidentally getting in at this point, may tend to float to the top 
 of the metal and not be carried into the casting proper. 
 
 (6) The point at which the metal enters is the hottest part 
 of the casting, and this must be borne in mind when considering 
 
 FIG, 217.— VIBRATOR. 
 
 the question of feeding, to avoid shrinksige cavities. Sometimes 
 it is desirable to chill this location in order to artificially oppose 
 these conditions. 
 
 (7) In placing a gate we must remember that as the metal 
 shrinks, ij will draw the sprue towards the casting. The sprue 
 being the hottest part, there may be a tendency for this contrac- 
 tion to tear the metal in two and especially to tear the sprue off 
 at the point where it joins the casting proper, thus making an 
 unsightly defect. 
 
 Gated Patterns, — Where the castings are very small, a large 
 number of them will be made into one pattern, fastened onto a 
 
IRON AND STEEL FOUNDING 
 
 245 
 
 common "gate" through which they are poured, which produces 
 very great economy in molding. (See Fig. 219). 
 
 Molding Machines, — At the present time various types of 
 molding machines are being extensively introduced into foun- 
 dries, in order to save some of the labor or skill required in mold- 
 ing, or both. The simplest form of machine is the "squeezer/' 
 which may be described by reference to Fig. 204. The correct 
 amount of sand is poured into the flask and by means of a long 
 lever the "presser board" is forced down on top of this sand, 
 
 FIG. 218— CORE MACHINE AND CORES. 
 
 squeezing it around the pattern and producing a half mold. In 
 taking the flask off the pattern by hand, damage may be done 
 the sand, and the molder's skill is still required to repair it. 
 
 The stripping-plate machine obviates part of this 'difficulty, 
 however. In this type the half pattern may be pushed up or 
 down through a close-fitting hole in a plate known as the " strip- 
 ping-plate" (see Fig. 208). After the sand has been rammed 
 around the pattern, a lever draws the pattern down through 
 the stripping-plate. As this drawing is mathematically exact, 
 no damage results to the sand and no repairs to the mold are 
 
246 THE METALLURGY OF IRON AND STEEL 
 
 necessary, so that unskilled laborers may be employed for the 
 work. 
 
 A still further extension in the line of machines is the " vibra- 
 tor," whereby the pattern is vibrated an extremely small amount, 
 some 5000 to 30,000 times per minute, during the drawing of the 
 pattern. No stripping-plate is necessary to separate it from the 
 sand, since the vibrator frees it perfectly and without damage. 
 
 In all these molding machines the operation may be conducted 
 either by means of levers, or by a mechanism operated by hy- 
 
 FIGS. 219 AND 220.— MULTIPLE MOLD AND CASTING. 
 
 draulic or pneumatic power, and several hundred patterns may 
 be made per day by one man who is very little, if any, more 
 skilled than a common laborer; and machines are made in which 
 castings weighing from a couple of ounces to several hundred 
 pounds may be molded with great economy. 
 
 Core Machines. — ^There are also on the market several machines 
 for making cores, an example of which is shown in Fig. 218. 
 
 Multiple Molds. — During the past few years a new type of 
 molding, known as "multiple molding," has come into use, in 
 which several flasks are placed in a pile and poured through a 
 common gate of sprue, as shown in Fig. 219. This type of mold- 
 ing saves mold costs, flasks, sand, floor space, and the weight' of 
 metal wasted in sprues, and many difficulties have been over- 
 
IRON AND STEEL FOUNDING 
 
 247 
 
 come, so that the castings are now made accurately to size and 
 good in all respects. 
 
 Contraction. — A bar of cast iron 12 in. long will contract about 
 0.125 in. during solidification and cooling (i.e., it will be about 
 117/8 in. long when cold. See page 328 for further details), 
 while a bar of steel will contract about twice as much. In both 
 cases the contraction in sectional dimensions will not be as great 
 as in length. 
 
 Design of Patterns 
 
 The foregoing description will show what a great financial 
 advantage it is to a purchaser if he designs castings that can be 
 easily molded, and if he can order a large number of castings of 
 exactly the same design. It is certain that a hundred castings 
 of one design can be made with very much greater cheapness than 
 the same number all of different designs, and of this economy the 
 purchaser obtains his full share, because the foundry is glad to 
 encourage such a customer and to make concessions in order to do 
 his work. I cannot recommend too strongly to engineers the 
 practice of making the castings in all 
 similar machines interchangeable, 
 both for the sake of economy and of 
 avoiding some delay and expense in 
 replacement after a breakdown. The 
 correct design of castings is further- 
 more one of the most important 
 branches of engineering work, since 
 the number of castings used is almost 
 one-haK of the total number of pieces 
 used in engineering work, while their 
 
 weight is equal to about one-sixth of the weight of all the iron 
 and steel employed. The following general hints ar^ therefore 
 offered to assist in this design; but each casting is a study in 
 itself, in order that the various desiderata referred to may be 
 obtained. 
 
 To Avoid Checks. — The commonest error in engineering 
 designs of castings is to make the corners too sharp, which makes 
 them very liable to check, because of the crystalline character of 
 iron and steel. This is the more important, because the greatest 
 leverage comes at the corners, which therefore should be made as 
 
 FIG. 221. 
 
248 THE METALLURGY OF IRON AND STEEL 
 
 strong as possible. Metals are crystalline substances and the 
 crystals grow during solidification. As solidification usually 
 extends from the surface inward, the crystals grow in a direction 
 perpendicular to the cooling surfaces. As shown in Fig. .221, 
 this results in a line extending inward from all corners, marking 
 the junction of many crystals. As the junction lines of crystals 
 are not as strong as the crystals themselves, this makes a line of 
 weakness on corners, which is the more marked the sharper the 
 corner is. In case a casting is to be machined, it is much better 
 to put a large fillet in all the corners, even if the rounded metal 
 must be cut away later, as greater strength is obtained in this 
 way. 
 
 The checking of castings comes from the strain produced by 
 the contraction of the metal tending to crush the sand. This is 
 the more intense the greater the distance between the two crush- 
 ing parts, because they must approach each other by an amount 
 exactly proportional to the length of metal between them. It is 
 therefore wise, wherever possible, to avoid long lengths of metal 
 connecting two parts which project into the sand. 
 
 Unequal cooling strains will also cause a check. This may be 
 illustrated by a pulley with thick arms and a thin face. The 
 face will solidify first and therefore yield very little to the sub- 
 sequent contraction of the arms. Moreover, the face, being 
 cooler, will then be stronger and the tendency will be for the arms 
 to tear themselves in two. To avoid this it is common practice 
 to chill the arms either by setting metallic pieces in the mold, or 
 by means of a "water gate. " A water gate is a loose column of 
 coke molded into the sand of the cope, down which water may be 
 poured. 
 
 To Avoid Shrinkage Cavities. — The formation of a "pipe" or 
 shrinkage cavity has already been explained. Such a defect in a 
 casting would be intolerable,, and is commonly avoided by having 
 a reservoir of metal situated above the casting proper and large 
 enough to keep it supplied with molten metal until it has^com- 
 pletely solidified. This reservoir is known as the "riser," or 
 "header" (sometimes merely "head"), or "feeder." The riser 
 is included as a part of the mold when it is made, but is cut off 
 the finished casting and used over again, as scrap. Sometimes 
 castings are so designed by engineers that a heavy section of 
 metal must be molded underneath a thinner section. As the 
 thinner section will solidify first, it cannot "feed" this lower, 
 
IRON AND STEEL FOUNDING 249 
 
 heavy section; and therefore a special form of riser is required, 
 or else the heavier section must be artificially cooled. Iron 
 castings usually do not form shrinkage cavities, but steel castings 
 must be fed carefully to all parts. 
 
 Cupola Melting of Iron for Castings 
 
 Iron for castings is melted either in the cupola or the air fur- 
 nace/ although "direct castings," i.e., castings made from the 
 metal just as it comes out of the blast furnace, are used in many 
 cases, and especially for cast-iron ingot molds at steel-works. 
 For this practice it is necessary to employ men with sufficient 
 expertness to be able to judge by eye the character and the anal- 
 ysis of the liquid iron as it flows from the furnace, because the 
 metal may vary greatly and without warning from one cast to 
 another. Sometimes, also, metal mixers are used, similar to 
 those at steel-works. 
 
 In the cupola (see Fig. 222), a layer of iron rests on top of a 
 bed of white-hot solid fuel. Air is blown aganist this fuel; the 
 oxygen unites with carbon, and the resulting hot gases, as they 
 pass upward, melt the iron. The layer of iron is not deep; it 
 should therefore all be melted in less than 5 or 10 minutes. 
 During this interval the fuel is being consumed also. Now a new 
 layer of fuel rests on the bed to replace what has been burned, and 
 a new layer of iron on top of that. By correct proportioning o^ 
 the alternate layers of fuel and iron, the new layer of iron will 
 occupy exactly the position of its predecessor, this location being 
 the "melting zone" of the furnace. 
 
 Cupola Zones. — The cupola should be so operated that certain 
 well-defined zones of action will be maintained, in order that 
 rapid, hot and economical melting will result and that loss by 
 oxidation be small. If proper conditions prevail all of these 
 desiderata may be obtained together, while, if otherwise, wasteful 
 methods may be accompanied by slow, irregular melting and 
 "dull iron," i.e., iron not sufficiently hot. 
 
 Crucible Zone. — The crucible extends from the bottom of the 
 
 cupola to the level of the tuyeres. The object of this part is to 
 
 form a place in which the iron and slag may collect after they 
 
 Jiave i^ielted and trickled down. If the tap-hole is kept open all 
 
 ^ The air furnace is used in gray cast iron, chilled iron and malleable cast iron foundries 
 and will be discussed in Chapter XIII. 
 
250 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Layer of 
 Iron 
 
 Layer of 
 Coke 
 
 Layer of 
 Iron 
 
 FIG. 222.-IRON CUPOLA. 
 
IRON AND STEEL FOUNDING 251 
 
 the time and the metal allowed to flow out of the cupola and 
 collect in an outside ladle as fast as it melts, the crucible zone 
 will be very shallow, and the tuyeres will be situated not more 
 than 2 to 5 in. above the bottom. If, on the other hand, the 
 crucible is used as a reservoir for a large amount of metal, the 
 tuyeres are placed correspondingly high. Hotter metal may be 
 obtained by collecting the iron in an outside ladle, because it is 
 not blown upon by the cold blast, and using the cupola as a 
 reservoir is not considered good practice except for short inter- 
 vals. 
 
 Tuyere Zone. — The tuyere zone is the place in which the blast 
 comes in contact with, and burns, the red-hot coke. It is the 
 zone of combustion, and all the heat of the operation should be 
 produced in this place. It is of course situated near the tuyeres 
 and wherever the raw blast may come in contact with coke. As 
 there is always a column of coke extending to the very bottom 
 of the cupola, combustion will begin immediately above the 
 reservoir of melted metal. Its upper limit will depend upon the 
 pressure of blast, because the greater the blast pressure, every- 
 thing else being the same, the higher will it extend its zone of 
 combustion. The blast pressure should be such, however, that 
 the top of the tuyere zone, or zone of combustion, should never 
 be more than 15 to 24 in. above the uppermost tuyeres.^ 
 
 Melting Zone. — The melting zone is the space in which all 
 the melting of iron takes place; it is situated immediately above 
 the tuyere zone. During the melting, the iron is supported on a 
 column of coke, which is the only solid material below the melting 
 zone. When each layer or charge of iron enters the melting zone 
 it should be about 15 to 24 in. above the uppermost tuyeres. As 
 fast as it melts it trickles down over the column of coke to the 
 bottom. It takes about five to ten minutes for each layer of iron 
 to melt, however, and during this time the column of coke is 
 burning and sinking. Therefore, the last of the iron will melt at 
 a point about 7 in. lower than the first. Consequently, the melt- 
 ing zone overlaps the upper limit of the zone of combustion. If 
 the layers of iron and coke are properly proportioned to the 
 pressure of blast, each charge of iron will enter the top of the 
 melting zone just before the next previous charge is completely 
 melted at the bottom, and thus a continuous stream of iron will 
 collect in the crucible or run from the tap-hole. Also, the cdke 
 
 ^ There are sometimes two rows of tuyeres in cuDoIas : see Fie. 228. 
 
252 THE METALLURGY OF IRON AND STEEL 
 
 FIGS. 223 TO 227.— FOUNDKY LADLES. 
 
IRON AND STEEL FOUNDING 
 
 253 
 
 burned from the column will be exactly replenished each time 
 by the layer of coke coming down, and the position of the melting 
 zone, which is the important consideration, will be maintained 
 within constant limits. 
 
 FIG. 228. 
 
 FIG. 229. 
 
 The actual position of the melting zone may always be learned 
 when the cupola is emptied, because the iron oxide formed there 
 will corrode the acid lining, which will therefore be cut away 
 somewhat at this point. Corrections may then be made, if 
 
254 
 
 THE METALLURGY OF IRON AND STEEL 
 
 necessary, in the next charge of the cupola. For example, the 
 bed should be so proportioned that it will extend, after the coke 
 is well heated up, to a point about 7 in. above the bottom of the 
 combustion zone. The melting zone itself, as indicated by the 
 cutting of the lining, should not be more than 8 to 12 in. deep. 
 Stack. — The stack extends above the melting zone to the level 
 of the charging door. The function of this part of the furnace is 
 to contain material that will absorb heat and thus prepare itself 
 for the actions at lower levels, and that will also keep the heat 
 down to the melting zone as well as possible. 
 
 FIG. 230.— POSITIVE PRESSURE IMPELLOR BLOWER. 
 
 Tuyeres. — The blast enters the cupola through the tuyeres, of 
 which there is usually one or two rows. The position of the upper 
 row of tuyeres determines the position of the melting zone in the 
 cupola. Two rows of tuyeres give faster melting in the cupola 
 than one row, but cause greater oxidation and the consumption 
 of more fuel on the bed, because the melting zone being higher 
 in the cupola, requires a larger bed to start with. 
 
 Blast. — The blast pressure will depend somewhat upon the 
 size of the cupola, but the present prevailing opinion is in favor 
 of pressure not exceeding a pound, even for the very largest 
 cupolas, and diminishing to half a pound or so for the smaller 
 sizes. Fan-blowers, although cheaper, are giving place to pres- 
 sure blowers, because if the former are opposed by pressure in 
 
IRON AND STEEL FOUNDING 
 
 255 
 
 the cupola stack, they revolve without blowing any wind. The 
 common type of blower used in America is of the two-impellor 
 type, an example of which is shown in Fig. 230. It takes about 
 60 cu. ft. of air to burn a pound of coke, from which may be 
 calculated the size of blower necessary for each cupola, allowing 
 about 50 to 100 per cent, excess for leaks and incomplete 
 combustion. 
 
 FIG. 231. 
 
 Makers of cupolas and. blowers give all the necessary data in 
 their catalogues, but advocate too high blast pressures and vol- 
 umes, for obvious reasons. If the blast volume is too large, or the 
 pressure is too great, the position of the melting zone will be too 
 high. This means that the bed of coke must be larger to reach to 
 the upper level of the melting zone, which is wasteful. It also 
 means that the melted iron will have a greater height to drop 
 through. It therefore oxidizes more, corrodes the cupola lining 
 more, and consequently causes more waste of iron and more slag. 
 
256 
 
 THE METALLURGY OF IRON AND STEEL 
 
 The volume of blast is the most important consideration, but this " 
 is difficult to measure, so the pressure is the substitute. 
 
 Loss. — The loss in melting will average about 2 to 4 per cent. 
 It is made up of the silicon burned and the iron oxidized and 
 carried away in the slag. There are other aources of loss in the 
 foundry, such as a second loss of metal remelted — the sprues, 
 risers, etc., which go back to the cupola in the form of scrap; 
 metal spilled during pouring (which may amount to as much as 
 5 or 6 per cent, more), etc. In some foundries it is customary to 
 pass the used floor sand through a magnetic concentrator, in 
 order to recover the pellets of iron spilled during pouring, and 
 important economy is sometimes obtained in this way. The 
 total loss, that is, the difference in weight between pig iron bought 
 and castings made, will probably be about 7 to 8 per cent, of the 
 weight of the iron bought. 
 
 Scrap Used. — Scrap pig iron is often mixed with new pig iron 
 for the manufacture of castings, both for the sake of economy and 
 because the scrap iron has a somewhat closer grain or texture, 
 which increases the strength of the mix- 
 ture. The amount of scrap used will 
 depend upon the materials to be manu- 
 factured. Cast-iron pipe is usually made 
 without scrap, this industry amounting 
 to between 500,000 and 800,000 tons 
 per year in the United States alone. 
 Stove foundries, on the other hand, use a 
 very large amount of scrap as a rule, and 
 jobbing foundries, in general, would 
 probably use an average of 30 to 40 per 
 cent, of outside scrap, besides the gates, 
 sprues, bad castings, etc., made in their 
 own foundries. The total production of gray-iron castings in 
 the United States will represent about 75 per cent, pig iron and 
 25 per cent, bought scrap. 
 
 Cupola Run. — The campaign of an ordinary foundry cupola 
 is only three or four hours long. As a general thing, the kindling 
 is started about noon and allowed to burn with a natural draft 
 until shortly after one o'clock, when the breast is closed and the 
 blast put on. Metal is then received until four or five o'clock in 
 the afternoon, when the last -charge is melted. The supports are 
 then pulled out from underneath the door closing the bottom of 
 
 FIG. 232.— MAGNETIC 
 CONCENTEATOR. 
 
IRON AND STEEL FOUNDING 257 
 
 the cupola, and the sand bottom, slag, coke, etc., left in the cupola 
 is allowed to drop and is quenched with water. In order to 
 allow plenty of room for the " drop " to fall, the cupola is usually- 
 elevated above the foundry floor. 
 
 Fuel, — ^The cupola is the cheapest furnace for melting because it 
 affords direct contact between metal and fuel. The amount of 
 fuel will be from one-fourth to one-twelfth of the weight of the 
 iron melted, the former figure prevailing where exceedingly hot 
 metal is desired — as, for example, for very small castings for 
 malleable cast-iron work — and the latter figure where the melting 
 is continued for several hours and the metal is not made very hot 
 but is to be poured into large castings. Coke is the commonest 
 fuel, with sometimes a mixture of coke and anthracite for the 
 bed. 
 
 Cupola Charge. — ^In the cupola is first placed shavings and 
 wood, on top of which is placed the bed of coke, which should be 
 large enough to reach 15 to 24 in. above the uppermost tuyeres 
 after the kindling is burned off. On top of this is placed a layer 
 of pig iron about 6 in. thick, then another layer of coke about 7 
 in. thick, another layer of iron, and so on. The actual weight of 
 "bhe coke for the bed and of coke and iron, for each charge will 
 therefore depend directly upon the diameter of the cupola inside 
 the brick lining, which varies from about 32 to 120 in., or even 
 more in some cases. The weight of the coke in each layer will be 
 about one-sixth to one-twelfth of the weight of iron in each 
 layer. The tuyeres and front of the cupola around the tap-hole, 
 known as the ''breast," are left open for an hour or so after the 
 kindling is lighted, in order that the draft may draw air in at that 
 point for combustion. When the kindling is thus burned off anji 
 the bottom coke well lighted; the breast is closed and the wind 
 turned on. It is very necessary that the bed should be well 
 lighted and level. More recently it has become common practice 
 to light the coke bed with a specially designed oil burner, instead 
 of kindlings, which saves time and labor. 
 
 Cupola Melting, — The heat now generated by the combustion of 
 coke begins to melt the iron, and in less than 15 minutes after the 
 wind is put on the metal should begin to run from the open tap- 
 hole. If it takes longer, then the coke bed was too high and waste- 
 ful. In 8 to 10 minutes thereafter the first layer of iron should 
 be £ill melted. Now the second layer of iron lies upon the column 
 of coke, whose top should again be 15 to 24 in. above the upper- 
 
 17 
 
258 THE METALLURGY OF IRON AND STEEL 
 
 most tuyeres. If the layers of coke are too thick there will be a 
 delay in the iron entering the melting zone and the extra coke 
 burned will not have been used to the best account. If the layers 
 of iron are too thick, the last of the layer will melt too near the 
 tuyeres, which will oxidize it excessively and make it cold. This 
 can be observed during the run by noting if the iron runs first hot 
 and then cold. It is very important to watch the flame that 
 comes off the top of the stack in the cupola. When the blast 
 volume is too I'arge this flame will be "cutting," — i.e., oxidizing 
 in character. Too great oxidation may also be observed if 
 sparks of burning iron are projected from the slag-hole. If the 
 layers of iron and coke are both too thick, there may be a correct 
 relation between the weights of the two, but both of the irregular- 
 ities mentioned above will be observed. It is probably not pos- 
 sible to have the alternate layers of fuel and iron too thin if the 
 correct relation between them is maintained.^ Of course, if 
 very hot iron is required it will be necessary to have thicker layers 
 of coke, and slower melting must be expected. 
 
 Chemical Changes. — As the iron drops down over the coke, it 
 absorbs sulphur, the exact proportion depending chiefly upon the 
 relative amount of coke and iron used and the per cent, of sulphur 
 in the coke. It will vary from 0.02 to 0.035 per cent, of the iron; 
 that is to say, if the pig iron charged contained 0.08 per cent, sul- 
 phur, there will be from 0.1 to 0.115 per cent, in the castings. 
 The sulphur in the first iron will be higher than in that of the 
 middle of the run, because of the extra amount of coke burned 
 before the iron begins to come from the tap-hole. The last iron 
 will also be somewhat higher in sulphur, because there is a larger 
 loss of metal during the last of the run, when the oxidizing con- 
 ditions are more intense, and therefore a concentration of sul- 
 phur. The best practice is to cut the blast off progressively as 
 there is less stock in the cupola. 
 
 In many foundries it is customary to charge limestone, in the 
 form of oyster shells, marble chippings, or crude limestone, and 
 sometimes with it a little fluorspar (CaFg), into the cupola. The 
 amount of limestone varies greatly, but will average perhaps 1 /2 to 
 11/2 per cent, of the weight of the metal. This limestone fluxes 
 the dirt on the metal and the ash of the coke and carries off some 
 sulphur in the slag. Fluorspar makes a somewhat more liquid slag 
 than limestone alone and the more liquid slag is believed to ab- 
 
 ^ See an able paper by R. Moldeake, No. 96. 
 
IRON AND STEEL FOUNDING 259 
 
 sorb a little more sulphur, and also to make the cupola "drop" 
 more easily, i.e., dump its contents when the campaign is ended, 
 and the bottom is allowed to fall. It also cuts the lining more. 
 
 As the metal melts and falls from the melting zone down in 
 front of the tuyeres, it suffers oxidation, which carries iron oxide 
 into the slag and also burns up silicon. The melted metal there- 
 fore contains from 0.25 to 0.4 per cent, less silicon than the origi- 
 nal pig.^ In other words, if the mixture charged contains 2.25 per 
 cent, of silicon, the castings will contain 1.85 to 2 per cent, silicon. 
 
 Cupola Gases, — The gases coming out of the top of the cupola 
 charge consist principally of nitrogen from the air, while the re- 
 mainder is carbon dioxide — COg — and carbon monoxide — CO — 
 with sometimes a little free oxygen. Free oxygen is evidence of a 
 ''cutting flame" and shows too great oxidation in the melting 
 zone. Such a flame may be recognized without the aid of chem- 
 ical analysis after a little practice by means of the eye. It is 
 "sharper" looking than a richer flame and burns close to the top 
 of the stock. One can identify it definitely by holding an iron 
 rod in it for a while; after the iron becomes red hot it will oxidize 
 much more rapidly in a cutting flame than in a reducing flame. 
 Again, a reducing flame will usuaUy not burn until it becomes 
 mixed with' the air sucked in at the charging door. All the carbon 
 monoxide that goes out of the charge- represents incomplete com- 
 bustion and a waste of heat. It seems to be impossible to prevent 
 this here, however, just as in the blast furnace, whose operations 
 cupola melting resembles in some general respects. Several 
 analyses of cupola gases are given in Table XVIII. 
 
 ^ With good practice it should be no more than . 30 per cent. less. 
 
260 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE A XVIII.— ANALYSIS OF CUPOLA GASES 
 Collected About 3 or 4 Feet Below the Charging Door 
 
 Time elapsed since 
 blast was put on 
 
 Analysis by Volume 
 
 Oxygen 
 
 
 Carbon 
 
 dioxide 
 
 CO, 
 
 Carbonic 
 
 oxide 
 
 CO 
 
 Nitrogen 
 (by differ- 
 ence) N 
 
 Ratio 
 
 CO2 is to CO 
 
 as 1 is to: 
 
 10 minutes. 
 
 1 hour 13 minutes. 
 
 2 hours 17 minutes. 
 
 3 hours 13 minutes. 
 
 4 hours 15 minutes. 
 
 38 minutes . 
 
 1 hour 42 minutes . 
 
 2 hours 50 minutes . 
 
 3 hours 40 minutes . 
 
 38 minutes 
 
 3 hours. 
 
 10 minutes 
 3 hours 10 minutes 
 
 50 minutes 
 
 1 hour 40 minutes 
 
 2 hours 40 minutes, 
 
 3 hours 40 minutes , 
 
 45 minutes . 
 
 1 hour 40 minutes . 
 
 2 hours 40 minutes . 
 
 3 hours 50 minutes . 
 
 4 hours 43 minutes . 
 
 1 hour 
 
 1 hour 53 minutes . 
 
 2 hours 45 minutes . 
 
 3 hours 45 minutes . 
 
 4 hours 45 minutes . 
 
 0.0 
 0.0 
 0.4 
 0.0 
 0.1 
 
 1.8 
 2.8 
 1.9 
 0.1 
 
 0.0 
 2.9 
 
 0.2 
 0.3 
 
 0.0 
 0.0 
 0.0 
 0.0 
 
 0.0 
 0.0 
 0.0 
 0.4 
 1.2 
 
 0.0 
 
 0.0 
 
 0.0 
 
 0.0' 
 
 0.0 
 
 13.8 
 9.5 
 9.2 
 6.7 
 
 7.8 
 
 7.6 
 7.5 
 7.1 
 
 7.2 
 10.2 
 
 5.4 
 
 13.1 
 
 10.3 
 
 7.1 
 
 8.3 
 
 8.2 
 
 6.0 
 
 13.0 
 
 13.0 
 
 8.2 
 
 6.0 
 
 5.1 
 
 9.8 
 
 9.1 
 
 8.8 
 
 '7.5 
 
 7.5 
 
 9.9 
 16.9 
 16.6 
 21.7 
 22.3 
 
 15.5 
 13.3 
 15.8 
 19.0 
 
 14.6 
 14.3 
 
 7.7 
 11.7 
 
 15.4 
 13.5 
 12.1 
 15.0 
 
 12.6 
 11.2 
 20.0 
 22.1 
 21.2 
 
 15.4 
 16.8 
 16.8 
 19.7 
 
 18.7 
 
 76.3 
 73.6 
 73.8 
 71.6 
 69.8 
 
 75.1 
 76.4 
 75.2 
 73.7 
 
 75.2 
 
 77.4 
 
 79.0 
 
 77.7 
 
 77.5 
 78.2 
 79.7 
 79.0 
 
 74.4 
 75.8 
 71.8 
 71.5 
 72.5 
 
 74.8 
 74.1 
 74.4 
 72.8 
 73.8 
 
 Another Cupola 
 
 0.717 
 1.780 
 1.804 
 3.239 
 2.859 
 
 2.04 
 1.77 
 2.225 
 2.64 
 
 1.431 
 2.65 
 
 0.588 
 1.136 
 
 2.17 
 1.626 
 1.475 
 2.50 
 
 0.97 
 
 0.862 
 
 2.44 
 
 3.683 
 
 4.157 
 
 1.57r 
 
 1.846 
 
 1.91 
 
 2.627 
 
 2.500 
 
 30 minutes . 
 
 0.1 
 
 16.7 
 
 7.3 
 
 75.9 
 
 .437 
 
 1 hour 30 minutes . 
 
 0.1 
 
 13.1 
 
 10.7 
 
 76.1 
 
 .817 
 
 2 hours 38 minutes . 
 
 0.4 
 
 11.8 
 
 11.0 
 
 76.8 
 
 .932 
 
 3 hours 20 minutes . 
 
 0.0 
 
 12.8 
 
 7.7 
 
 79.5 
 
 .602 
 
IRON AND STEEL FOUNDING 
 
 261 
 
 Burdening the Cupola, — It should be the duty of the foundry 
 metallurgist or chemist to learn from his records, or other ap- 
 proximations, the amoimt and analysis of all the metal in the 
 yard. The following table will, for example, show a convenient 
 form of this record. 
 
 TABLE A XIX 
 
 Kind 
 
 Weight 
 tons 
 
 Si 
 
 s 
 
 P 
 
 Mn 
 
 500 
 
 0.70 
 
 0.100 
 
 1.50 
 
 0.30 
 
 60 
 
 2.50 
 
 0.025 
 
 0.07 
 
 0.60 
 
 100 
 
 3.00 
 
 0.030 
 
 0.80 
 
 1.25 
 
 150 
 
 1.75 
 
 0.070 
 
 0.30 
 
 0.60 
 
 30 
 
 10.00 
 
 0.040 
 
 0.50 
 
 0.10 
 
 30 
 
 50.00 
 
 0.003 
 
 0.04 
 
 
 100 
 
 1.70? 
 
 0.100? 
 
 1.00? 
 
 0.60? 
 
 300 
 
 1.50? 
 
 0.20? 
 
 1.40? 
 
 0.60? 
 
 100 
 
 .1.50? 
 
 0.20? 
 
 1.40? 
 
 0.60? 
 
 100 
 
 0.10 
 
 0.07 
 
 0.10 
 
 0.60 
 
 Price 
 
 High Sulphur, Southern 
 High Silicon, Bessemer . 
 
 XNo. 1 
 
 No. 3 Foundry. ... 
 
 Ferrosilicon A 
 
 Ferroeilicon B 
 
 Machinery Scrap. 
 
 Miscellaneous Scrap 
 
 Cast-iron Borings., 
 
 Steel Scrap 
 
 $18.00 
 25.00 
 24.00 
 22.50 
 35.00 
 
 105.00 
 19.00 
 15.00 
 11.00 
 13.00 
 
 The price should always be in evidence. It should not be the 
 price at which the material was purchased but the market price 
 at the time the iron is to be used. For instance, if a large amount 
 of high grade pig iron had been contracted for a year previously 
 and if meanwhile the price of pig iron had been rising, the pur- 
 chase price of that pig iron would not represent its present value. 
 From the current numbers of such trade periodicals as the Iron 
 Age and the Iron Trade Review, one can always obtain the pre- 
 vailing prices for the different grades of iron. 
 
 Suppose now with these irons it is desired to burden a 72-in. 
 cupola with a mixture for making heavy hydraulic pumps for 
 which a satisfactory analysis might be 1.60 per cent, silicon, 
 0.70 per cent, phosphorus^ less than 0.10 per cent, sulphur and 
 about 0.50 per cent, manganese. The first step is to calculate 
 the cupola charges, and the chemist knows by experience with 
 this particular cupola that it will lose 0.25 per cent, silicon and 
 0.10 per cent, manganese and it will gain 0.03 per cent, sulphur. 
 The average analysis of the mixture put into the cupola must then 
 be 1.85 per cent, silicon, 0.70 per cent, phosphorus, less than 
 0.07 per cent, sulphur and about 0.60 per cent, manganese. 
 
262 THE METALLURGY OF IRON AND STEEL 
 
 The chemist also knows by calculation that about 5200 lb. of 
 iron will give a layer of the proper thickness in a 72-in. cupola. 
 His problem now is to make such a mixture of the available pig 
 irons that their collective weight will be 5200 lb. and their aver- 
 age analysis as given above. Moreover if he is a good metallur- 
 gist he must aim at using as large an amount as possible of the 
 cheapest materials. 
 
 He first considers the steel scrap, but knows he cannot use 
 very much of this because too much coke would be required for 
 getting iron of the requisite fluidity, but he estimates that 5 per 
 cent, (say 200 lb.) will not increase harmfully the fuel necessary. 
 This figure therefore comes at the top of his list (see Table XX). 
 
 Next he considers the use of machinery scrap, because he knows 
 that his miscellaneous scrap and borings are too uncertain in 
 analysis to be used in a mixture which must give pretty strong 
 and non-porous castings. One thousand pounds of machinery 
 scrap would be about 20 per cent, of his mixture, and he knows 
 from experience that this is a fairly satisfactory proportion for 
 scrap, so that figure goes down second in Table XX. The low 
 price of the High Sulphur Southern Pig tempts him, but he real- 
 izes that he must offset the use of this material by some high 
 silicon low sulphur iron. And in casting about for such a one he 
 naturally considers first the X No. 1. He cannot use much of 
 this either because of its high manganese and it seems reason- 
 able to mix an equal amount of these two; the only question is. 
 How much of this mixture will the cupola stand? To get an idea 
 of this, he first calculates their average analysis and finds it to 
 be 1.85 per cent, silicon, 0.065 per cent, sulphur, 1.15 per cent, 
 phosphorus, and 0.78 per cent, manganese. Evidently th6 phos- 
 phorus is the only element in this mixture that gives him difficulty. 
 Indeed, if that were not high he could make almost his whole 
 charge up of these two irons and the scrap. The phosphorus in 
 this mixture is 0.45 per cent, higher than that of his desired 
 mixture. Therefore he knows that he must use a good deal of 
 No. 3 Foundry iron to bring this element down. The phosphorus 
 in the No. 3 Foundry iron is about as much below the desired 
 phosphorus as that in the mixture of the High Sulphur Southern 
 and the X No. 1 is above it. He must not forget, however, that 
 he has already used 1000 lb. of machinery scrap containing 
 probably 1 per cent, of phosphorus. Therefore he must use a 
 correspondingly larger amount of No. 3 iron to offset this also. 
 
IRON AND STEEL FOUNDING 263 
 
 As a first estimate he therefore considers using 800 lb. of High 
 Sulphur Southern, 800 lb. of X No. 1, and 2400 lb. of No. 3 
 Foundry— that is, once and a half as much No. .3 as the mixture 
 of the other two. But a little reflection tells him that this mix- 
 ture is going to be too low in silicon, because the mixture of High 
 Sulphur Southern and X No. 1 only gave us 1.85 per cent, silicon, 
 while the No. 3 Foundry and the machinery scrap are both below 
 that. There are then three ways open to him. He may use a 
 little Ferrosilicon A or he may pound up a little Ferrosilicon B 
 and dissolve it in the ladle of iron or he may use a little High 
 Silicon Bessemer iron. Either of these methods would do, but 
 the writer would prefer to use the High Silicon Bessemer because 
 this will have the effect of cutting the sulphur and phosphorus 
 down and the expense is practically the same. (It requires such a 
 small amount of ferrosilicon to give the desired silicon in the mix- 
 ture thg-t the expense of using it is very small, in spite of its price.) 
 Consequently we put down the weights shown in the second col- 
 umn of Table XX, and we now figure out the weight of silicon, 
 sulphur, phosphorus, and manganese in the mixture by the meth- 
 ods indicated there, and the average percentage of each element. 
 The latter figures show us that the silicon is too low, and a simple 
 calculation shows us that' we need 5 lb. more in the total weight 
 of silicon. We can get this by increasing the amount of either 
 High Silicon Bessemer or of X No. 1, and correspondingly de- 
 creasing the No. 3 Foundry. The High Silicon Bessemer has 
 0.75 per cent, more silicon than the No. 3, so it would take (5 lb. 
 •^0.75 per cent. =) about 650 lb. change to makeup the difference 
 in this way. The X No. 1 has 1.25 per cent, more silicon than the 
 No. 3, so it would take (5 lb. -=-1.25 = ) 400.1b. change to make 
 up the difference in this way. We naturally would prefer to use 
 the latter, being cheaper, and if we think we can stand all that 
 extra manganese in our castings we probably will do so; if not, 
 we will have to use altogether 1000 lb. of High Silicon Bessemer 
 and only 1400 lb. of No. 3 Foundry. We then make up a new 
 table similar to Table XX and figure out the average analysis as 
 before. It should now come about right. 
 
264 
 
 THE METALLURGY OF IRON AND STEEL 
 
 X 
 X 
 
 
 
 
 , ® 
 
 
 
 
 
 
 
 
 N 
 
 1 
 
 
 g 1 
 
 c^ 
 
 o -^ c 
 
 C 
 
 ^ 
 
 
 
 1 § 
 
 
 CO (N O (N C^ 
 
 1 -^ 
 
 
 
 
 
 
 1—1 
 
 CO o 
 
 1 « 
 
 
 
 
 
 
 
 
 . 
 
 
 CO 3 
 
 (N 
 
 O O >* 
 
 c 
 
 C^ 
 
 ) 05 t- 1 
 
 
 o En 
 
 
 
 
 
 
 
 
 CD 
 
 
 ^ 2 
 
 O O CS 
 
 CO CO c 
 
 ) -* 
 
 ^ 
 
 P^ a, 
 
 
 1-H T-1 
 
 
 
 CO o 
 
 ■+^ 
 
 
 
 
 
 
 
 
 
 
 tie 
 
 S 
 
 
 
 
 
 
 
 
 1-1 
 
 ^5* 
 
 ^ 
 
 -^ 
 
 c 
 
 c 
 
 TJ^ 
 
 o 
 
 c 
 
 00 t^ 
 
 ^ 
 
 r— 1 
 
 
 o oc 
 
 (M 
 
 ■<* 
 
 1- 
 
 CD O 
 
 
 ^ 
 
 C 
 
 F-^ O C 
 
 1—1 
 
 c 
 
 CO o 
 
 
 
 
 
 
 
 
 
 « 
 
 
 g 
 
 CSl 
 
 c 
 
 CC 
 
 o 
 
 o 
 
 c 
 
 00 ^ 
 
 
 .2 
 
 c 
 
 l> 
 
 iC 
 
 ■^ 
 
 lO 
 
 G 
 
 -^ t: 
 
 
 r;^ 
 
 
 ^ 
 
 
 (N 
 
 CC 
 
 
 Ol . 
 
 
 CQ 
 
 
 
 
 
 
 
 
 -... 1 
 
 
 . (u 
 
 
 p- 
 
 
 
 
 
 
 
 
 
 
 c 
 
 c 
 
 c 
 
 »o 
 
 o 
 
 e^. 
 
 
 
 
 
 CC 
 
 cc 
 
 CO 
 
 (N 
 
 CO 
 
 CC 
 
 
 
 
 
 '^ bO 
 
 <= 
 
 c 
 
 c 
 
 1-H 
 
 o 
 
 C 
 
 
 
 
 
 
 e^ 
 
 
 
 
 
 
 
 
 
 m ^ 
 
 c 
 
 C 
 
 o 
 
 o 
 
 o 
 
 l> 
 
 
 
 
 
 So^ 
 
 »H 
 
 C 
 
 lO 
 
 X 
 
 -. CO 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^i. 
 
 c 
 
 iH 
 
 T-i 
 
 o 
 
 o 
 
 o 
 
 
 
 
 >s 
 
 
 
 
 
 
 
 
 
 
 
 'S^ 
 
 
 
 C^ 
 
 
 
 
 
 
 
 
 < 
 
 %^ 
 
 o 
 
 O 
 
 o 
 
 1— ( 
 
 T- 
 
 o 
 
 CO 
 
 o 
 
 o 
 
 
 
 
 
 
 3 
 
 o 
 
 c 
 
 o 
 
 o 
 
 o 
 
 o 
 
 
 
 
 
 OQ 
 
 
 
 
 
 
 
 
 
 
 rt 
 
 
 "N 
 
 
 
 
 
 
 f 
 
 
 
 o 
 
 o 
 
 c 
 
 o 
 
 o 
 
 lO 
 
 o 
 
 
 
 
 
 ;S ^ 
 
 I— 1 
 
 l> 
 
 I> 
 
 o 
 
 t^ 
 
 iC 
 
 
 
 
 
 M 
 
 o 
 
 1—1 
 
 o 
 
 CO 
 
 rH 
 
 IM 
 
 
 
 
 
 4^ 
 
 
 
 
 
 
 
 
 
 
 
 -E • 
 
 
 c 
 
 o 
 
 
 o 
 
 O 
 
 c 
 
 ' 
 
 
 
 SPr^ 
 
 
 ^ 
 
 ^ 
 
 
 
 O 
 
 c 
 
 
 
 
 
 
 f*^ 
 
 oc 
 
 cc 
 
 e^ 
 
 ■^ 
 
 C*3 
 
 
 
 
 
 r- 
 
 
 
 (N 
 
 
 IC 
 
 
 
 
 
 a 
 
 
 a 
 
 ■4-: 
 
 3 
 
 t 
 
 P 
 
 1 
 
 « 
 
 c 
 
 c 
 
 
 i 1 
 
 ? 
 
 3 
 3 
 
 In 
 
 
 
 1 
 
 
 : 1 
 
 3 s 
 
 D ^ 
 
 i t^ 
 
 C"- 
 
 c 
 
 { t2 
 
 1 i 
 
 1 ^ 
 
 3 
 
 H 
 
 3 
 
 3 
 
IRON AND STEEL FOUNDING 265 
 
 Comparative Cupola Practice 
 
 In Table XXI are shown some figures on Comparative Cupola 
 Practice which confirm in a striking way the rules laid down 
 above, and add new light on the subject. 
 
 Fuel. — The first lesson we learn from the table is that a 
 mixture of coal and coke and inferior coke give slow melting and 
 a poor fuel ratio. Indeed, the work of the cupola using these 
 grades of fuel is so far inferior to the others that I have separated 
 them in Table XXI and omitted them from all my calculations. 
 After a careful study of the figures, we are strengthened in the 
 opinion that a mixture of coal and coke has nothing to recommend 
 it except a deceptive first cost. 
 
 Tuyere Ratio. — The next most striking evidence produced by 
 the figures is the relation between the tuyere area and the speed 
 of melting: if we average up the iron melted per minute in the 
 cupolas whose area is less than 6.56 times the tuyere area we 
 obtain a figure of 22.56 lb.; if we get the corresponding figure 
 for the cupolas with lesser proportionate tuyere area, we obtain 
 18.57 lb. Indeed so striking is the relation that there is only 
 one exception, namely, cupola No. 8, and we need not look far 
 for a reason for the slow melting in this cupola. It is evidently 
 due to the short height of stack which causes the iron to reach the 
 melting zone before it has been sufficiently preheated. A large 
 proportionate tuyere area evidently means that the wind will pass 
 through the tuyeres with less resistance and a lower velocity. 
 That is to say, we will get more wind and it will not be driven so 
 much to the center of the cupola. It may be said to be established 
 for the types of practice here exhibited that the collective area 
 of the tuyeres should not be less than one-sixth, and preferably 
 not less than one-quarter, of the horizontal area of section of the 
 cupola. 
 
 Height of Stack. — There is also an important relation between 
 the speed of melting and the height of the charging door above 
 the tuyeres divided by the diameter of the cupola. The average 
 speed of melting of the cupolas, where this ratio is greater than 
 2.5, is 24.12 lb. per minute, while the average speed of melting 
 of those whose ratio is less than 2.5 is only 19. 15 lb. per minute. 
 There are only two exceptions to this rule: (1) Cupola No. 6 is a 
 fast melter, but this is doubtless due to the large proportionate 
 tuyere area. (2) Cupola No. 3 is a slow melter, but this is doubt- 
 
266 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXL— COMPARATIVE CUPOLA PRACTICE 
 
 Diameter of cupola, inches 
 
 Height of tuyeres from sand 
 bottom, inches 
 
 Height of charging door above 
 tuyeres, inches 
 
 Height of charging door above 
 tuyeres divided by diameter... 
 
 Number of tuyeres 
 
 Size of tuyeres, inches, vertical 
 X horizontal 
 
 Area of tuyeres, square inches.. 
 
 Cupola area is how many times 
 tuyere area 
 
 Diameter of blast pipe, inches.... 
 
 Blast pressure, oimces, 20 min- 
 utes after start 
 
 Class of work made. 
 
 Lined up how often? Months. 
 Weight of fuel on bed, pounds. 
 Weight of iron on bed, pounds... 
 Weight of fuel in charges sub- 
 sequent to the bed, pounds. . . 
 Weight of iron in charges sub 
 
 sequent to the bed, pounds... 
 
 Total weight of fuel, pounds 
 
 Total weight of iron, pounds. . . . 
 Ratio of fuel on bed, above 
 . tuyeres, to iron on bed, 1 is to 
 Ratio of fuel to iron, later 
 
 charges 
 
 One pound fuel melts how many 
 
 pounds iron 
 
 Kind of fuel used / 
 
 Fuel weighed or measured 1 
 
 Height of fuel bed above tuyeres, 
 
 inches 
 
 Thickness of fuel, charges after 
 
 the bed, inches 
 
 Thickness of iron charges after 
 
 the bed, inches 
 
 27 
 
 12 
 
 106 
 
 4.0 
 6 
 
 4x6 
 144 
 
 3.97 
 8 
 
 Job 
 
 350 
 700 
 
 90 
 
 400 
 2,000 
 8,000 
 
 3.2 
 4.4 
 
 4.0 
 
 Coke 
 
 W 
 
 20 
 
 8.4 
 
 3.3 
 
 35 
 
 120 
 
 3.4 
 
 6 
 
 5x5 
 
 170 
 
 5.662 
 16 
 
 16 
 
 Plate 
 
 6 
 650 
 1,300 
 
 125 
 
 1,300 
 
 1,850 
 
 14,000 
 
 2.8 
 10.4 
 
 7.5 
 
 Coke 
 W 
 
 2V 
 
 7.0 
 6.5 
 
 42 
 
 11 
 139 
 
 3.3 
 
 6 
 
 4x6 
 144 
 
 9.62 
 14 
 
 Light 
 Job 
 
 6 
 650 
 1,400 
 
 100 
 
 1,000 
 
 2,150 
 
 16,000 
 
 3.5 
 10.0 
 
 7.6 
 
 Coke 
 
 M 
 
 15 
 
 3.8 
 3.5 
 
 44 
 12 
 109 
 2.5 
 
 336 
 
 4.53 
 15 
 
 9 1/2 
 
 Boiler 
 & Plate 
 
 6 
 
 1,300 
 3,000 
 
 175 
 
 2,000 
 
 3,075 
 
 27,000 
 
 3.2 
 11.4 
 
 8.8 
 
 Coke 
 
 W 
 
 33 
 
 6.1 
 
 6.3 
 
 54 
 7.5 
 103.5 
 
 1.9 
 
 6 
 
 4^x13 
 348 
 
 6.58 
 17 
 
 13 
 
 Stove 
 Plate 
 
 6 
 
 1,400 
 3,000 
 
 300 
 
 3,000 
 
 4,800 
 33,000 
 
 3.7 
 10.0 
 
 6.9 
 
 Coke 
 
 W 
 
 24 
 
 6.9 
 
 6.3 
 
 1 Two rows of tuyeres. 
 
IRON AND STEEL FOUNDING 267 
 
 TABLE XXL— COMPARATIVE CUPOLA PRACTICE.— Con/mwed. 
 
 Time before iron comes after 
 
 blast is on, minutes 
 
 Time to melt each iron charge 
 
 after the bed, minutes 
 
 Total iron melted per minute, 
 
 pounds 
 
 Total iron melted per minute, per 
 
 square foot cupola area .... 
 
 7 
 
 3.25 
 123 
 30.98 
 
 15 
 8.5 
 155 
 23.20 
 
 5 
 
 5.7 
 175 
 18.19 
 
 15 
 9 
 225 
 21.30 
 
 10 
 10 
 300 
 18.87 
 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 11 
 
 Diameter of cupola, inches 
 
 54 
 
 56 
 
 58 
 
 60 
 
 60 
 
 72 
 
 Height of tuyeres from sand 
 
 
 
 
 1 
 
 
 
 bottom, inches 
 
 14 
 
 25 
 
 16 
 
 2 
 
 12 
 
 24 
 
 Height of charging door above 
 
 
 tuyeres, inches 
 
 113 
 
 141 
 
 92 
 
 112 
 
 114 
 
 142 
 
 Height of charging door above 
 
 
 
 
 
 
 
 tuyeres divided by diameter. . . 
 
 2.1 
 
 2.5 
 
 1.6 
 
 1.9 
 
 1.9 
 
 2.0 
 
 Number of tuveres 
 
 6 
 
 12 
 
 6 
 
 8 
 
 6 
 
 
 Size of tuyeres, inches, vertical 
 
 
 
 
 
 
 
 X horizontal 
 
 10x7 
 420 
 
 6x12 
 864 
 
 7x12 
 548 
 
 4ix7i 
 270 
 
 7x10 
 420 
 
 
 Area of tuyeres, square inches. . . 
 
 706 
 
 Cupola area is how many times 
 
 
 
 
 
 
 
 tuyere area 
 
 5.45 
 
 2.85 
 
 4.83* 
 
 10.47 
 
 6.73' 
 
 5.76 
 
 Diameter of blast -pipe, inches. . . 
 
 18 
 
 24 
 
 20 
 
 18 
 
 18 
 
 24 
 
 Blast pressure, ounces, 20 min- 
 
 
 
 
 
 
 
 utes after start 
 
 14 
 
 16 
 
 16 
 
 8 
 
 8 
 
 13i 
 
 
 Boiler 
 
 
 Pipe 
 
 
 Sani- 
 
 Elec- 
 
 Class of work made • 
 
 & Radi- 
 ator 
 
 Rolls 
 
 Ft'g& 
 Job 
 
 Plate 
 
 tary & 
 Plate 
 
 trical 
 
 Lined up how often? Months. . . 
 
 10 
 
 8 to 10 
 
 7 
 
 6 
 
 12 
 
 6 
 
 Weight of fuel on bed, pounds. . . 
 
 2,000 
 
 3,000 
 
 2,000 
 
 1,700 
 
 1,800 
 
 2,500 
 
 Weight of iron on bed, pounds. . . 
 
 6,000 
 
 9,000 
 
 6,000 
 
 6,000 
 
 5,000 
 
 7,000 
 
 Weight of fuel in charges subse- 
 
 
 
 
 
 
 
 quent to the bed, pounds 
 
 450 
 
 450 
 
 400 
 
 200 
 
 350 
 
 600 
 
 Weight of iron in charges subse- 
 
 
 
 
 
 
 
 quent to the bed, pounds 
 
 4,000 
 
 4,500 
 
 4,000 
 
 2,000 
 
 4,000 
 
 7,600 
 
 1 Two rows of tuyeres. 
 
 (For averages, see next page.) 
 
268 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXI.—COMPARATIVE CUPOLA PRACTICE.— Confewe^i 
 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 11 
 
 Total weight of fuel, pounds 
 
 Total weight of iron, pounds 
 
 Ratio of fuel on bed, above 
 
 tuyeres, to iron on bed, 1 is to . . 
 Ratio of fuel to iron, later 
 
 charges 
 
 8,500 
 64,800 
 
 4.3 
 
 8.8 . 
 
 7.5 
 
 Coke 
 
 M 
 
 32 
 
 10.4 
 8.4 
 20 
 11 
 360 
 22.64 
 
 12,000 
 100,000 
 
 5.0 
 
 10.0 
 
 8.3 
 Co£e 
 W 
 
 40 
 
 9.6 
 
 8.8 
 
 5 
 
 12 
 
 370 
 
 21.60 
 
 4,400 
 30,000 
 
 5.0 
 
 4.1 
 
 6.8 
 
 Coke 
 
 W 
 
 14^ 
 8.0 
 7.3 
 
 15 
 
 12 
 333 
 
 18.14 
 
 6,600 
 49,500 
 
 3.8 
 
 10.0 
 
 7.5 
 Coke 
 
 . M 
 
 30 
 3.8 
 3.4 
 
 1 
 5.5 
 367 
 18.80 
 
 6,600 
 50,000 
 
 4.7 
 
 11.4 
 
 7.6 
 
 Coke 
 
 W 
 
 22 
 6.5 
 6.8 
 
 15 
 
 11 
 360 
 
 18.40 
 
 12,700 
 
 127,500 
 
 5.4 
 
 11.7 
 
 One pound fuel melts how many 
 pounds iron 
 
 10.0 
 
 Kind of fuel used i. 
 
 Solvay 
 
 Fuel weighed or measured 
 
 Height of fuel bed above tuyeres, 
 inches 
 
 Coke 
 
 W 
 
 6 
 
 Thickness of fuel, charges after 
 the bed, inches 
 
 7.8 
 
 Thickness of iron charges after 
 the bed, inches 
 
 9 
 
 Time before iron comes after 
 bla«t is OTij Tnirmtpifi 
 
 15 
 
 Time to melt each iron charge 
 after the bed, minutes 
 
 Total iron melted per minute, 
 pounds 
 
 13.5 
 
 567 
 
 Total iron melted per minute, per 
 
 20.06 
 
 Al 
 
 A2 
 
 A3 
 
 A4 
 
 Averages 
 
 Diameter of cupola, inches 
 
 Height of tuyeres from sand 
 
 bottom, inches 
 
 Height of charging door above 
 
 tuyeres, inches 
 
 Height of charging door above 
 
 tuyeres divided by diameter . 
 
 Number of tuyeres 
 
 Size of tuyeres, inches, vertical 
 
 X horizontal 
 
 32 
 
 11 
 
 133 
 
 4.1 
 6 
 
 4|x6 
 
 42 
 
 12 
 
 132 
 
 3.1 
 6 
 
 4x10 
 
 44 
 9.5 
 
 84.5 
 
 1.9 
 
 8 
 
 2^x11 
 
 48 
 
 14 
 
 103 
 
 2.1 
 6 
 
 4x6 
 
 12. 6^ 
 116 
 2.52 
 
 1 Including Al. A2, A3, A4. 
 
 ^ Excluding Al, A2, A3, A4. 
 
IRON AND STEEL FOUNDING , 269 
 
 TABLE XXI.— COMPARATIVE CUPOLA PRACTICE.— Continued. 
 
 
 Al 
 
 A2 
 
 A3 
 
 A4 
 
 Averages 
 
 Area of tuyeres, square inches.. 
 
 Cupola area is how many times 
 
 tuyere area 
 
 162 
 
 4.96 
 12 
 
 16 
 
 Gas 
 
 Engine 
 
 6 
 450 
 1,000 
 
 110 
 
 1,000 
 1,250 
 8,000 
 
 3.5 
 
 8.7 
 
 6.4 
 
 Light 
 Coke 
 
 M 
 21 
 
 7.3 
 
 6 
 10 
 
 7.7 
 57 
 10.2 
 
 240 
 
 5.77 
 
 14 
 
 12 
 
 Job 
 
 1,000 
 4,500 
 
 250 
 
 2,000 
 
 1,935 
 
 12,000 
 
 8.6 
 
 8.0 
 
 6.2 
 
 Coal& 
 Coke 
 
 M 
 11 
 
 9.5 
 
 7 
 15 
 
 198 
 
 7.68 
 12 
 
 13 
 Pumps 
 & Job 
 
 3 
 
 1,000 
 1,000 
 
 no 
 
 1,500 
 
 2,040 
 
 14,800 
 
 1.6 
 
 13.6 
 
 7.2 
 
 Coal& 
 Coke 
 
 W 
 
 25 
 3.9 
 4.7 
 
 15 
 9.5 
 164 
 
 15.53 
 
 144 
 
 12.56 
 16 
 
 8 
 
 Medium 
 
 Light 
 
 6 
 1,300 
 4,000 
 
 200 
 
 2,000 
 
 4,100 
 
 32,000 
 
 5.0 
 
 10.0 
 
 7.8 
 
 Coal& 
 Coal 
 
 M 
 22.5 
 
 5.8 
 
 5.3 
 
 7 
 13.8 
 145 
 11.54 
 
 
 6. 
 
 56^ 
 
 Diameter of blast pipe, inches . 
 Blast pressure, ounces, 20 min- 
 utes after start 
 
 
 12 
 
 I 
 
 Class of work made 
 
 Lined up how often? Months. 
 
 Weight of fuel on bed, pounds. 
 
 Weight of iron on bed, pounds . 
 
 Weight of fuel in charges sub- 
 sequent to the bed, pounds... 
 
 Weight of iron in charges sub- 
 sequent to the bed, pounds... 
 
 Total weight of fuel, pounds. . 
 
 
 
 
 
 
 
 Total weight of iron, pounds.. . 
 Ratio of fuel on bed, , above 
 
 tuyeres, to iron on bed, 1 is to 
 Ratio of fuel to iron, later 
 
 charges 
 
 
 4 
 9 
 7 
 
 22 
 7 
 6 
 
 11 
 9. 
 
 2 
 
 5 
 
 One pound fuel melts how 
 m.any pounds iron 
 
 S^ 
 
 Kind of fuel used .... V 
 
 
 Fuel weighed or measured 
 
 Height of fuel bed above 
 
 tuyeres, inches 
 
 Thickness of fuel, charges after 
 
 the bed, inches 
 
 5^ 
 0^ 
 
 Thickness of iron charges after 
 the b*ed, inches 
 
 4* 
 
 Time before iron comes after 
 blast is on, minutes 
 
 Time to melt each iron charge 
 after the bed, minutes 
 
 Total iron melted per minute, 
 pounds 
 
 3^ 
 46^ 
 
 Total iron melted per minute, 
 per square foot cupola area. . 
 
 21 
 
 .112 
 
 ^Including Al, A2, A3, A4. 
 
 ^Excluding Al, A2, A3, A4. 
 
270 THE METALLURGY OF IRON AND STEEL 
 
 less due to the very small proportionate'tuyere area. With these 
 exceptions the rule is universal and a comparison of different 
 cupolas one with another only strengthens it; for example, 4 with 
 8, the proportionate tuyere area being nearly the same; also 2 
 with 6, etc. A comparison of 3 with 9 is an apparent exception r 
 which is perhaps explained by the low tuyeres in No. 9, 
 
 Blast Pressure. — The average speed of melting of the cupolas 
 with more than 12 oz. blast pressure is only 20.75 lb. per minute, 
 while that with less than 12 oz. is 21.53. This relation is not 
 so striking as to establish a rule, especially as a single cupola 
 (No. 1) throws such a large influence into the low-blast column. 
 Nevertheless, the evidence is sufficiently striking to discredit 
 the theory that higher blast pressure necessarily gives faster 
 melting. Indeed, if cupolas 3, 9, and 10 had a larger tuyere area, 
 we should expect an average result very favorable to low blast 
 pressure. 
 
 Height of Fuel Bed, — ^The original height of fuel bed is no 
 criterion with which to figure as it in many cases is raised or 
 lowered during the first few minutes of melting, and thereafter 
 occupies some other position. It is the thickness of the fuel and 
 iron charges after the bed which is the important consideration 
 and, as already observed, these should be regulated to about 
 6 to 8 in. for fuel and 6 in. for iron irrespective of the diameter of 
 the cupola. In case hot iron is desired the layer of fuel should be 
 at the upper limit of thickness and vice versa. 
 
 Speed of Melting.— The speed of melting is very important, be- 
 cause everything else being equal the faster the melting per square 
 foot of cupola area the greater will be the efficiency of opera- 
 tion of the cupola, and therefore of economy under the given 
 conditions. 
 
 Melting Steel For Castings 
 
 Steel castings are made in (1) acid open-hearth furnaces, (2) 
 basic open-hearth furnaces, (3) small Bessemer converters of 
 special design, and (4) crucibles. The castings are made from 
 the metal just as it comes from the steel furnace, and the processes 
 are mentioned above in the order of relative production. 
 
 Open-hearth Furnaces. — ^The making of steel for castings is 
 practically the same as making steel for ingots, except that 
 foundry furnaces are of smaller size, varying on the average 
 
IRON AND STEEL FOUNDING 271 
 
 from 5 to 30. .tons. In some cases furnaces smaller than this 
 are used, but it is generally believed that circumstances rarely 
 warrant this, since the expense of running the small furnaces 
 is large in proportion. The chief difference in practice is that 
 the temperature for steel-casting work is hotter than when 
 ingots are made. Therefore furnace repairs are higher and 
 the life shorter. In ordinary open-hearth foundry work the 
 purification is continued to the point where the steel contains 
 about 0.18 to 0.28 per cent, carbon after recarburizing, while the 
 silicon will be usually 0.20 to 0.30 per cent. 
 
 Acid versus Basic Open-hearth Steel. — In steel castings it is 
 necessary to have somewhat lower phosphorus and sulphur than 
 in ingots, because the metal is not to receive the beneficial effect 
 of mechanical work, and therefore must be purer in order to 
 have a good degree of strength and ductility. Consequently, if 
 we use an acid steel-making process, we must start with very low 
 phosphorus and low sulphur pig iron, which is costly and becomes 
 more so each year in America. For this reason the Bessemer and 
 the acid open-hearth steel-making processes are more expensive 
 for casting work than the basic open-hearth. The result is a 
 present rapid increase in the use of basic open-hearth steel in 
 America as well as in Germany, with a probability that in a few 
 years it will be the predominant process for this purpose. This is 
 in spite of the fact that basic steel has very serious disadvantages, 
 chief among which are the amount of oxygen contained in it at 
 the end of the process and the difficulty of keeping the desired 
 amount of silicon in it during teeming, both of which conditions 
 increase the liability to blow-holes, which are especially objec- 
 tionable in castings, as there is no opportunity of their being 
 welded up and as castings may have to be discarded on this 
 account after a good deal of expensive machine-work has been 
 done. On this account the acid open-hearth process has long 
 held the predominant position in the steel-casting industry. In 
 fact, this is the only place where the acid open-hearth process 
 now finds important employment on a large scale in this 
 country. 
 
 Bessemer versus Open-hearth. — The open-hearth furnace gives 
 a large amount of steel at long intervals, which is very incon- 
 venient for foundry work, because the molds necessary to take all 
 the metal must be stored upon the foundry floor until the heat is 
 ready to pour, and then those who are to do the teeming must 
 
272 THE METALLURGY OF IRON AND STEEL 
 
 interrupt their other work for half an hour or more for this pur- 
 pose. Even where the foundry is large enough to have many 
 furnaces, there is no surety that they will come out at regular and 
 short intervals, because the operation in one may be delayed. 
 Another disadvantage of the open-hearth process is that, in order 
 to be economical, it must be operated continuously day and 
 night, which also is inconvenient for foundry work. Further, it is 
 not possible to get the metal as hot as desired without great 
 damage to the furnace, which is subjected to a higher tempera- 
 ture than the metal. Lastly, since hot metal is desirable for all 
 castings except those of very large size, it is usually necessary to 
 tap all the metal from the furnace at once, and recarburize it at 
 once, which prevents castings of special analysis being made, un- 
 less ordered in very large quantities. Nowadays, a great many 
 nickel-steel castings are made in open-hearth furnaces, but this 
 requires nice calculations, sa that the castings molded shall be 
 just equal to the capacity of the heat, and usually results in a 
 certain amount of scrapping of high-class metal. 
 
 All these objections are avoided in making castings in Besse- 
 mer converters, but they too have their great disadvantages, 
 chief among which is greater co^t. The latter is due principally 
 to the amount of waste in the side-blown converters used for this 
 purpose, and the greater cost of the pig iron used, which must be 
 low in phosphorus and sulphur. Small converter plants are so 
 very cheap — costing less than $5000 for apparatus — that a great 
 many iron-foundries in the United Sates are putting them in as 
 an adjunct to their cupola process, in order that they may be able 
 to make steel castings at will. The amount of capital tied up is 
 so small, and there being no important expense in starting and 
 stopping the converter as often as desired, they can do this 
 economically. 
 
 Tropenas Converter. — The largest number of converters for 
 steel-casting work are of the Tropenas type, in which the wind en- 
 ters the vessel from seven tuyeres on the side, and the converter 
 is tipped in such a manner that the streams of air are deflected 
 onto the top of the bath. The impurities are oxidized as in the 
 regular Bessemer process, except that the action is not quite 
 so rapid, and the carbon is burned to carbon dioxide instead of 
 carbon monoxide, which generates a much larger amount of heat; 
 
 C+ = C0 (generates 29,160 calories). 
 C + 20 = C02 (generates 97,200 calories). 
 
IRON AND STEEL FOUNDING 
 
 273 
 
 ,— D 
 
 The pig iron used for these converters runs frequently above 
 2 per cent, of silicon, and this, together with the formation of 
 carbon dioxide, results in very hot and fluid steel, which can be 
 poured into castings of almost any small size. The blows usually 
 last about 15 to 20 minutes, and the loss is from 17 to 20 per cent, 
 of the weight of pig iron charged into the cupola in which it is 
 melted. The vessel can be started up and stopped with very 
 little expense, and this advantage over the open-hearth, together 
 with the small amount of capital 
 necessary to build the converter plant 
 and the other conditions already men- 
 tioned, has caused many of these con- 
 verters to be installed in America, 
 and several in France, England, and 
 other countries. 
 
 The great disadvantages of the 
 Tropenas converter are the waste and 
 the cost for making repairs. Slight 
 patching can be done through the 
 mouth, but there is a hole in the front 
 of the converter shell, closed by a 
 movable steel plate, through which 
 the operator can dig his way into the 
 interior to perform the necessary re- 
 pairs. As the lining in the neighbor- 
 hood of the tuyeres^is usually worn 
 out in less than twenty blows, this 
 costly method of lining is a serious 
 drawback. 
 
 Long Tuyere Converter. — Next to the Tropenas, the greater 
 number of converters at work in America are of the Long-Tuyere, 
 or Stoughton, type, devised by the writer. The bottom part of 
 this vessel is attached to the trunnion-ring by a method similar 
 to that used for regular Bessemer bottoms, and may be removed 
 with great ease, thus cheapening and facilitating repairs. More- 
 over, the chief repairs are in the bottom part at the mouths of the 
 tuyeres, and therefore the lining of the upper part does not have 
 to be relined completely for several months, although slight patch- 
 ing is necessary every twenty-five to thirty blows, when the 
 bottom is changed. This converter is arranged to have but one 
 row of tuyeres, discharging the blast immediately at the surface of 
 
 IS 
 
 FIG. 233.— SECTIONS OF LONG- 
 TUYERE CONVERTER. 
 
274 
 
 THE METALLURGY OF IRON AND STEEL 
 
 the metal, and the lining on the tuyere side is thicker in order that 
 the tuyeres may be increased in length, which decreases the loss 
 of metal during the process. The excessive loss in the side-blown 
 converters is due chiefly to the spitting, which is very large, 
 especially when the tuyeres have become worn away to a short 
 length and the streams of air are badly directed and set up in- 
 terfering currents. 
 
 Another cause of the excessive loss is the large amount of slag 
 formed, because more iron is oxidized, and this corrodes the lining 
 very rapidly. Part of the iron is oxidized at the mouths of the 
 tuyeres, and another part is oxidized when the violent agitation 
 of the bath, which occurs in all of these converters, especially dur- 
 ing the boil, throws the metal up into the stack, where it meets 
 free oxygen. The loss in the Long-Tuyere modification is 14 
 to 16 per cent., including cupola loss. 
 
 Sizes Used. — Most of the small converters have a capacity of 
 about 2 tons, because this is economical and well proportioned to 
 the capacity of an ordinary foundry. There are some 4- and 5- 
 ton converters, however, and some of less than 1 ton capacity. 
 Sizes less than 2 tons are costly to operate in proportion to their 
 output. The blast pressure usually employed is 3 to 5 lb., with 
 an average of about 3 3/4 lb. 
 
 TABLE XXII,— ANALYSES OF SIDE-BLOWN CONVERTER-GASES. 
 
 4 
 
 Time after Beginning 
 of Blow 
 
 Analyses — ^Per 
 cent. 
 
 Calculations from 
 Analyses — ^Per cent. 
 
 0) 
 
 'a 
 
 CO 
 
 co^ 
 
 
 
 N 
 
 and 
 H 
 
 Total 
 
 
 N^ 
 
 
 
 enter- 
 ing 
 
 Burning 
 
 Si and Fe 
 
 and Mn 
 
 1 
 
 2 
 3 
 4 
 
 4 min., flame starts . . 
 
 10 min,, boiling 
 
 12 min., shortening^. . 
 17 min., after first drop 
 21 min., end of blow. . 
 
 0.0 
 
 0.3 
 
 0.4 
 
 10.7 
 
 8.2 
 24.3 
 
 8.8 
 13.0 
 
 1.1 
 0.4 
 0.2 
 0.2 
 
 90.7 
 75.0 
 90.6 
 76.1 
 
 7.1 
 18.3 
 
 6.8 
 15.8 
 
 88.7 
 73.0 
 88.6 
 74.1 
 
 23.6 
 19.4 
 23.5 
 
 19.7 
 
 16.5 
 1.1 
 
 16.7 
 3.9 
 
 1 By difference, H being estimated as 2 per cent. 
 
 ^ I.e., just before the first drop. There are two drops to the fiame in this operation, the 
 second marking the end. 
 
X 
 
 THE SOLUTION THEORY OF IRON AND STEEL 
 
 Scientifically considered, all of the members of the iron and 
 steel series are alloys of iron and carbon. Therefore a study of 
 the general theory of alloys leads to important information upon 
 iron and steel. According to the authoritative definition, "a 
 metallic alloy is a substance possessing the general physical prop- 
 erties of a metal, but consisting of two or more metals, or of 
 metals with non-metallic bodies, in intimate mixture,. solution, or 
 combination with one another, forming, when melted, a homo- 
 geneous fluid. " It will be seen from this that the essential fea- 
 ture of an alloy is that, when melted, it shall form a homogeneous 
 fluid. In plain language, this means that, when melted, the 
 different components are dissolved in one another. Melted 
 alloys, therefore, come under the general head of solutions. In 
 fact,- the great bulk of our alloys, and especially of iron and steel, 
 are produced by first dissolving the melted components and then 
 allowing them to freeze. The laws governing this freezing, or 
 solidification, have been known only a few years, and if this new 
 knowledge has made great revolutions in physical chemistry, it 
 has led to no less important discoveries in regard to the nature of 
 iron and steel. 
 
 Solid Solutions. — ^^uppose first we have two metals that are 
 soluble in each other when liquid, and also when solid. In other 
 words, the metals of the. alloy will be just as completely dissolved 
 in each other after solidification as before. They will then form a 
 "solid solution, " and a solid solution bears practically the same 
 relation to a liquid solution as a solid pure metal does to the same 
 metal when liquid. For example, gold and silver dissolve in each 
 other when liquid, and also when solid, in any proportion. Con- 
 sequently any solution of these metals will cool to the freezing- 
 point and then solidify without there being any important change 
 (from the metallurgical or practical standpoint) in the relations of 
 the two metals after the freezing. The reason that these solid 
 solutions form in any proportion is that the two metals crystallize 
 
 275 
 
276 THE METALLURGY OF IRON AND STEEL 
 
 alike. It is, perhaps, a new thought to the reader, but it is never- 
 theless true, that a metal forms a crystal whenever it solidifies. 
 Furthermore, each metal has a particular, general shape which its 
 crystals assume, and there is almost no force powerful enough to 
 prevent them from taking that shape in preference to any other. 
 
 Tiny as the crystals sometimes are — often requiring the highest 
 powers of the microscope to reveal them — their crystalline forces 
 are very powerful. If, therefore, two metals do not form like 
 crystals, they cannot solidify in solution, i.e., in the same crystal, 
 but freezing must be accompanied by separation. 
 
 By European metallurgists solid solutions are often called 
 "mixed crystals, " or "isomorphous mixtures, " but this termin- 
 ology is objected to because the relation of two substances when 
 dissolved is far more intimate than any mixture possibly could be. 
 In a mixture,* the microscope will always be powerful enough to 
 distinguish the different components, but a solution always ap- 
 pears like a simple uniform body. Furthermore, the properties 
 of a mixture are intermediate between the properties of its com- 
 ponents, but a solution — either liquid or solid — has some prop- 
 erties which are different from any of the properties of either of 
 its components. Again, the components of a solution are held 
 together by chemical forces, while the components of a mixture are 
 either not held together at all, or only because of close mechanical 
 association. In brief, a solution has some of the characteristics of 
 a chemical compound, and differs from such a compound chiefly 
 because the latter must always be composed of definite amounts 
 of each component, and in some multiple of their atomic weights, 
 while a solution may contain widely varying amounts of each 
 component. 
 
 Freezing of Solid Solutions. — When metals form solid solu- 
 tions, the solutions freeze more or less like pure metals. In Fig. 
 240. I have shown graphically the freezing of all the alloys 
 of gold and silver, extending from no silver (i.e., 100 per cent, 
 gold) to no gold (100 per cent, silver). This figure has been 
 drawn from results obtained by experiment. The proportion of 
 silver is shown by the abscissae, or the horizontal distance away 
 from the axis OCj and the temperature is shown by the ordinates, 
 or the vertical distance away from the axis 00, In other words, 
 any point between these two axes represents by one axis distance 
 the composition of the alloy in silver and gold, and by the other, 
 the temperature at which the alloy is at the moment. For 
 
THE SOLUTION THEORY OF IRON AND STEEL 
 
 277 
 
 example, the point marked a is a certain distance from the axis 
 OCj which shows that the alloy in question contains 50 per cent, 
 of silver; it is also a certain distance above the axis OOj which 
 shows that the alloy is now at a temperature of 1100°C.(2012°F.)- 
 
 Since there is nothing in the alloys but silver and gold, the 
 percentage of gold in each is 100 minus the percentage of silver. 
 The horizontal distances therefore show the percentage of gold as 
 well as of silver, and this is given in the second line of figures 
 under the table. 
 
 Suppose we have a solution of 50 per cent, of silver and 50 per 
 cent, gold at a temperature of 1100° C. It is represented at the 
 point a, and is now liquid. When it cools to about 1035° C, it 
 
 1200°C, 
 
 1100° 
 A 
 
 1000° 
 
 ■ 
 
 900" 
 
 
 
 L 
 
 quid 4-^^oy 
 
 
 b 
 
 
 
 
 --==: 
 
 
 
 Liq 
 
 Solid 
 
 a 
 
 ^SoJid 
 Alloy 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 ■...^^^ 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 10^ Silver 
 100^ Gold 
 
 30% 
 70^ 
 
 50% 
 50^ 
 
 1Q% 
 
 2192 F. 
 
 2012" 
 
 1832" 
 B 
 
 1662" 
 
 go;s loo^stiTOT 
 
 10% 
 
 FIG. 240.-FREEZING CURVE OF THE GOLD-SILVER ALLOYS. 
 
 begins to freeze. It does not all freeze at the same time, but first 
 some solid crystals freeze out and we have a mixture of solid 
 crystals with liquid solution. The more the mass cools the more 
 solid crystals there will be, each crystal being a solid solution of 
 gold and silver. It is not until we reach a temperature of -985° C, 
 however, that the last liquid freezes, and then we have a solid 
 solution of gold and silver, the two solid metals now bearing 
 practically the same relation to each other chemically that they 
 did when liquid. 
 
 It will be noticed that when this alloy cooled from the point a 
 to 1035°, it met the line A E B. This is the line that represents 
 the beginning of freezing for all the gold-silver alloys. It has 
 been drawn after many experiments have been made to show 
 where it lies in the diagram. It will furthermore be noticed that 
 when this alloy cooled to 985°, it met the line A D B, This is the 
 line that represents the completion of freezing for all the gold- 
 silver alloys, and its position has been learned from many ex- 
 periments- 
 
278 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Take now the alloy containing 60 per cent, silver and 40 per 
 cent, gold, at a temperature of 1150° C. — the point 5. As this 
 cools to a temperature of 1025° C, it meets the line A E B and 
 freezing begins in the same way as before. Solid crystals form 
 more and more as the alloy cools further, until we meet the line 
 A D B dX di> temperature of 980° C. At this point the last liquid 
 freezes. 
 
 The Freezing of Alloys of Lead and Tin 
 
 1. Suppose now, on the other hand, we have two metals which 
 are soluble in each other when melted, but not when solid. Evi- 
 dently there cannot be the same results as those described in the 
 case of the gold-silver alloys. Such a series is found in the alloys 
 of lead and tin, whose freezing is shown graphically in Fig. 241. 
 
 c. sso" 
 
 002 F. 
 
 . 100 80 80 70 60 60 40 SO 20 10 
 
 FIG. 241.— THE FREEZING OF ALLOYS OF LEAD AND TIN. 
 
 Here again I have shown the percentage of lead and tin by the 
 abscissae, and the temperature by the ordinates. And, again, the 
 diagram has been drawn from results obtained by experiment. 
 The following facts are not based upon reasoning or logic, but 
 are to be accepted because it is known that the different actions 
 described actually took place upon trial. 
 
 2. Consider first a solution containing 83 per cent, of lead and 
 at a temperature of 300° C. This will be represented by the point 
 n in Fig. 241, and will be a solution of lead in tin, still liquid, 
 
THE SOLUTION THEORY OF IRON AND STEEL 279 
 
 although about 26° below the melting-point of lead itself. If 
 cooled, the solution will remain liquid until it reaches a tempera- 
 ture of about 275°, which brings it, we see, to the line A B. This 
 is the lowest temperature to which it will go and retain that much 
 lead in solution. If cooled any more, the lead will begin to 
 precipitate and of course as much as is precipitated will immedi- 
 ately solidify, being already well below its melting-point. The 
 lead comes out as crystals of solid lead, which remain mixed with 
 the mass of the molten solution, but form no longer a part of it 
 chemically.^ The solution, i.e., the part still liquid, becomes en- 
 riched in tin in proportion as the lead precipitates. Leaving this 
 alloy here for the moment, let us next consider one with less lead. 
 
 3. Now take an alloy with 67 per cent, of lead and 33 per cent, 
 of tin at a temperature of 250° C. This will be represented by the 
 point h in Fig. 241. We have again liquid lead dissolved inliquid 
 tin, although the alloy is 76° below the melting-point of pure lead. 
 Suppose this alloy cools until it reaches a temperature of about 
 240° C, where it meets the line A B. It will not cool below that 
 temperature and retain all the lead in solution. If we cool it a 
 few degrees, lead separates out and becomes mechanically mixed 
 with the liquid solution. 
 
 4. Consider next an alloy of 55 per cent, lead dissolved in 45 
 per cent, of tin at a temperature of 250° C. As this cools to a 
 temperature slightly below 225°, it again meets the line A B and 
 crystals of lead separate out and solidify. And so on : any solu- 
 tion, when it cools to a temperature where it meets the line A B, 
 reaches its limit of solubility; it cannot cool more and retain all of 
 the lead in solution; but if it does cool, then some of the lead 
 must be precipitated.^ 
 
 5. To sum up the preceding paragraphs, we may then say that 
 the less lead we have in solution (within limits to be afterward 
 defined) the lower the temperature can go without any of it being 
 precipitated. Or, in other words, the less the lead in solution the 
 lower the temperature will go before any freezing begins. This 
 knowledge is the result of experiment, and it has been shown that 
 
 ^ It is a fact that when the lead is precipitated from the solution, it carries with it a few 
 traces of dissolved tin, but for the present we neglect this slight impurity for the sake of 
 simplicity, and consider that pure lead separates. 
 
 ^ Any lead that is precipitated must of course freeze, because the temperature is already 
 belov7 the freezing-point of lead. Conversely, any lead that freezes must be precipitated 
 from the alloy, because we know as a matter of experiment that frozen lead will not retain 
 tin in solution (omitting, of course, the few traces of tin retained by the lead and which, for 
 the sake of simplicity, we omit in the discussion). 
 
280 THE METALLURGY OF IRON AND STEEL 
 
 the line A B in Fig. 241 represents the equilibrium between the 
 amounts of lead in the different solutions and the temperature to 
 which each will cool before any lead is precipitated, or, in other 
 words, the amount of lead that saturates the solution at each 
 temperature. 
 
 6. With this knowledge let us consider again the first solution 
 — containing 83 per cent, of lead. We have stated that when 
 this solution was cooled below 275° C, some lead was precipi- 
 tated. Have we any evidence now as to how much lead would be 
 precipitated with each unit drop in temperature? Evidently we 
 have, for we know how much lead is normally in the saturated 
 solution at each temperature; this evidence is given to us by the 
 line A B, For example, assume that the solution containing 83 
 per cent, of lead is cooled to 240° C; how much lead will be left in 
 solution, and therefore how much will have been precipitated? 
 From paragraph 3 we know that 67 per cent, of lead remained in 
 solution down to a temperature of 240° C. and that there was an 
 equilibrium between the amount of lead and this temperature, or, 
 in other words, the solution is saturated with lead at this point. 
 Therefore it would be reasonable to suppose that the solution 
 which started with 83 per cent, of lead would retain exactly 67 
 per cent, of lead by the time the temperature of 240° is reached. 
 That this reasoning is correct is proved by experimental evidence. 
 
 7. Now let us cool the same alloy to a temperature of 225°; 
 how much lead will this retain in solution and how much would be 
 in a precipitated form? We have already seen (paragraph 4) 
 that an alloy containing 55 per cent, of lead will retain all of that 
 lead in solution until it falls to a temperature of 225°. Therefore 
 it is reasonable to suppose that the cooling alloy we are considering 
 will retain just 55 per cent, of lead and no more by the time its 
 temperature has fallen to 225°, and this is in fact the case. In 
 short, each so'ution will always retain enough lead to keep it in 
 a saturated state after it once becomes saturated. 
 
 8. We have learned above that every point on the line A B 
 represents the maximum amount of lead that can be retained in 
 one of these solutions at that temperature. Therefore as soon as 
 any solution meets the line A B, it follows along this line as it 
 cools and lead precipitates in amounts proportional to the tem- 
 perature, so that the solutions will always be saturated with lead. 
 
 9. It is therefore evident that if we start with any of these 
 liquid alloys containing an amount of lead represented on the 
 
THE SOLUTION THEORY OF IRON AND STEEL 281 
 
 diagram to the left of the point B, lead will precipitate from it 
 during cooling and the amount of lead in solution will continually 
 decrease, so that the composition of the alloy at the different 
 temperatures will be represented by a point traveling down the 
 line A B, Therefore, in every such case we shall reach finally a 
 solution with 31 per cent, of lead and 69 per cent, of tin (the 
 proportions represented by the abscissa of the point B) when 
 the temperature has fallen to 180° C. (the ordinate of the point 
 B), Mixed with the solution at this temperature will be the 
 amount of lead which has separated during cooling and which 
 will depend upon the percentage in the original solution. (If 
 this is not clear on the first reading, a little thought will make it 
 so, especially if the reader follows in Fig. 241 each action I have 
 described, step by step.) 
 
 10. We have considered up to now only the solutions con- 
 taining large amounts of lead and from which lead is precipitated 
 on cooling. Let us now consider a solution containing 90 per 
 cent, of tin at a temperature of 225° C. (point dj Fig. 241). It is 
 still liquid, although below the normal melting temperature of 
 both lead and tin. This will cool until a temperature of about 
 210° is reached and it meets the line C B. If it cools any more, 
 tin will be precipitated and will of course freeze, because it is 
 already below its normal freezing-point (231° C). 
 
 11. Consider next an alloy containing 85 per cent, of tin at a 
 temperature of 225° (point e in Fig. 241) . This will cool to about 
 200° C, where it meets the line C B, but if it cools any further, 
 tin will be precipitated and will freeze. 
 
 12. In other words, the line C B represents the conditions of 
 equilibrium between the temperature and the amount of tin that 
 will be retained in solution, just as the line A B represented the 
 equilibrium between the temperature and the amount of lead 
 that would be retained in solution. That is to say, every point on 
 the line C B represents the amount of tin that will saturate a 
 solution at that temperature. 
 
 13. Returning then again to the alloy containing 90 per cent, 
 of tin, how much tin will have been precipitated by the time the 
 alloy cools to, let us say, 200° C? Obviously 5 per cent, of tin 
 would have been precipitated, leaving a solution containing 85 
 per cent, of tin, because we have already seen that 85 per cent, of 
 tin will saturate a solution at 200° C. 
 
 14. Whatever solution we may have had to start with, pro- 
 
282 THE METALLURGY OF IRON AND STEEL 
 
 vided there was always more than 69 per cent, of tin, as soon as 
 that solution met the line C B it would travel down this line, pre- 
 cipitating tin progressively in proportionate amounts such that 
 the tin left in the solution at the varying temperatures would 
 correspond to the ordinates of the line C B, In other words, this 
 solution follows the line C B, Therefore, in all such alloys we 
 shall finally arrive at a solution having 31 per cent, of lead and 69 
 per cent, of tin (the abscissa of the point B) when the temperature 
 has fallen to 180° C. This solution is known as the "eutectic " 
 solution, and with it will be mixed precipitated tin in amount 
 depending upon the amount of tin in the original solution, 
 
 15. We may now sum up paragraphs 9 and 14 by saying that 
 whatever solution of lead and tin we have to start with, we will 
 always have one containing 31 per cent, of lead and 69 per cent, 
 of tin by the time the temperature has fallen to 180°, and with 
 this solution will be mixed some precipitated tin or precipitated 
 lead, as the case may be (unless, of course, we started with exactly 
 31 per cent, of lead and 69 per cent, of tin). 
 
 Now let us consider the further cooling of these alloys. What 
 will happen, however, to the solution containing 31 per cent, of 
 lead and 69 per cent, of tin, which we have called the "eutectic" 
 solution? It is evident that when this solution cools below 180°, 
 it must cross the point Bj and therefore both the lines A B and 
 B C a,t once. Obviously, it cannot cross either of these lines 
 without. precipitating lead or tin. In point of fact, on crossing 
 both of the lines at the same time, it precipitates at once all the 
 remaining lead and all the remaining tin, and therefore completes 
 the decomposition of the solution and the solidification. 
 
 The lead and tin separate in tiny solid crystals or flakelets, 
 which arrange themselves in a parallel banded structure similar 
 to that shown in Fig. 251, page 294. This structure is known as 
 the "eutectic" structure. The term "eutectic alloy" means 
 etymologically "well-melting alloy, " because it remains melted 
 longer than any other alloy of the same metals, and every 
 solution, the components of which are soluble in each other in the 
 liquid state and insoluble in the solid state, will form a eutectic 
 solution in the way I have described. This applies not only to 
 solutions of metals in each other, but to solutions of metals in 
 liquids which are afterward frozen, and even of salts in water, etc. 
 
 Freezing-point Curves. — The lines A B and C B are often 
 spoken of as the "upper freezing-point curves" of the alloys, be- 
 
THE SOLUTION THEORY OF IRON AND STEEL 283 
 
 cause any alloy which cools to this line will then commence to 
 freeze. The line D E is often called the "lower freezing-point 
 curve, " because this line represents the temperature of freezing 
 of the eutectic of the series, and we have already shown that 
 every alloy in the series automatically forms a eutectic by 
 "selective" precipitation; therefore every alloy in the series will 
 not be entirely solid until it reaches the temperature at which 
 the eutectic solidifies/ which is always the same. 
 
 Cooling Curves. — There are certain thermal changes which 
 accompany the chemical changes I have outlined in the preceding 
 paragraphs. These thermal changes are of importance, because 
 it is by means of them that we are usually able to obtain the first 
 evidence of the precipitation of excess metal, the formation and 
 solidification of a eutectic, etc. Consider the alloy containing 
 83 per cent, of lead and 17 per cent, of tin, at 300° C, and let us 
 observe by means of a thermometer or pyrometer the rate of 
 cooling. At first the thermometer will fall pretty fast, but when 
 we reach 275°, where the line A B is met, the rate of fall is sud- 
 denly retarded. It thus becomes evident to us that some event 
 counteracts the fall in temperature. What this event is we learn 
 from microscopic evidence, and, as has already been explained, 
 it is the precipitation of lead. This explanation might have been 
 expected, because the precipitation of lead at a temperature 
 below its normal freezing-point would of course be accompanied 
 by freezing and, during the freezing, the metal would liberate its 
 latent heat of fusion and thus oppose the cooling of the mass as a 
 whole. The rate of fall of temperature is, moreover, retarded all 
 the way down to 180°, because lead is being continuously precipi- 
 tated as the solution travels down the line A B. When we reach 
 180°, the fall of temperature is not only retarded, but actually 
 ceases; in some cases the temperature may rise slightly. This 
 change is due to the large amount of latent heat of fusion liberated 
 by the rapid freezing of the eutectic at one temperature. This 
 arrest continues until the eutectic is entirely solid, after which 
 the rate of fall becomes rapid again and proceeds without impor- 
 tant change until the - atmospheric temperature is reached, 
 because now we have nearly the cooling of a solid alloy. 
 
 These changes are represented diagrammatically in Fig. 242, 
 in which the abscissEe show the time, in minutes, from the begin- 
 
 1 Some metallurgists prefer not to draw the line DE, but to represent the freezing of the 
 eutectic merely by the point B. 
 
284 
 
 THE METALLURGY OF IRON AND STEEL 
 
 hing of the cooling, and the ordinates show temperatures. The 
 change in direction at the point a shows the retardation due to 
 the precipitation of the lead, while the long horizontal part at h 
 shows that for several minutes the temperature was not falling 
 at all, because the eutectic was freezing. 
 
 Next, experimenting upon the alloy with 67 per cent, of lead, 
 we find that it cools rapidly until it meets the line A B, when the 
 rate of fall is retarded continuously until we reach 180°, where 
 again an arrest (or perhaps an actual rise) of the temperature is 
 observed, after the completion of which the fall in temperature 
 becomes rapid again and proceeds normally. 
 
 Similar thermal changes are observed in the alloy containing 
 55 per cent, of lead, but at temperatures of 225° and 180°. Now, 
 
 572E 
 
 8odc 
 
 
 
 
 
 
 
 
 
 
 
 a 
 
 ^ 
 
 .^iv*,^ 
 
 
 
 
 
 
 
 
 ax) 
 
 
 
 <ii^ 
 
 p^ 
 
 
 
 
 
 
 
 
 
 
 X 
 
 V • ,— 
 
 b El 
 
 tectic 
 
 
 
 
 
 \/ 
 
 fM 
 
 ezing 
 
 \ 
 
 lflO° 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 n 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 212 
 
 10 Minutes 
 
 20 Minutes 
 
 FIG 242. 
 
 -FREEZING CURVE OF AN ALLOY CONTAINING 83 PER CENT. 
 . OF LEAD AND 17 PER CENT. OF TIN. 
 
 by plotting .the upper changes in the several solutions (i.e., at 
 275° C, at 240° C, and at 225° C), the position of the line A Bis 
 determined, and by plotting the lower changes (z.e., at 180° C. in 
 each case) the line D E is determined. 
 
 In studying the solutions of ,the series rich in tin, similar 
 thermal changes are observed. Consider the alloy containing 
 90 per cent, of tin and 10 per cent, of lead; this will cool rapidly 
 until a temperature of 210° C. is reached, when a retardation will 
 occur and will persist until the temperature 180° is reached. By 
 this time the excess tin will all have been precipitated and the 
 residual solution^ will have travelled down the line C 5 to the 
 
 1 These residual solutions are technically known as " mother liquors " or " mother metals," 
 because it is out of them that the solid metal is being *'boni." 
 
THE SOLUTION THEORY OF IRON AND STEEL 
 
 285 
 
 point B, Thereupon an arrest (or an actual reversal) in the rate 
 of cooling will occur until the eutectic has solidified, after which 
 the cooling will proceed at a rapid rate. In the case of the alloy 
 containing 85 per cent, of tin, the first retardation will begin at 
 200°, and then an arrest at 180°. By plotting the points at 210° 
 and 200°, we obtain the position of the line C B,^ and by plotting 
 the two points at 180° we obtain further points to determine the 
 line D E. 
 
 eo 
 
 70 
 
 80 
 
 90 
 
 100 Per- 
 
 40 
 
 30 
 
 SO 
 
 10 
 
 Cent 
 
 
 
 Tia 10 ao 30 40 60 
 
 I^ad 100 90 80 70 60 60 
 
 FIG. 243.— FREEZING-POINT AND STRENGTH CURVE OF THE LEAD-TIN 
 
 ALLOYS. 
 
 Properties of Eutectics, — As already said, not only the lead-tin 
 alloys, but every alloy of which the components arc soluble when 
 liquid, and insoluble (or nearly so) when solid, will form eutectics, 
 and it happens that this is the case with the great majority of 
 our metallic alloys. For this reason the properties of the eutectic 
 are very important in all metallurgical work. It will be seen 
 that a eutectic is not a chemical compound in atomic proportions. 
 For example, in the case of the lead-tin alloys, the point B 
 
 1 It should be understood that in actual experimental determination of linea corresponding 
 i^ AB, CB and DE, in any series of alloys, not a few, but a large number of different solu- 
 tions are studied, and a great xuany retardation points are found before the lines are dramu 
 
286 THE METALLURGY OF IRON AND STEEL 
 
 comes at 31 per cent, of lead and 69 per cent, of tin merely be- 
 cause the line A B crosses the line C 5 at this ordinate, and this 
 has no relation to atomic ratios. If the melting-point of tin hap- 
 pened to be higher, or if it did not precipitate from lead so fast 
 upon cooling, the line C B would cut the line A 5 at a different 
 point, and therefore the composition of the eutectic would be 
 different. 
 
 The structure of the eutectic is also very important. The 
 majority of eutectics have a structure similar to the banded form 
 shown in Fig. 251. The tiny crystals in this structure are inter- 
 mingled very intimately, and this close association has a bene- 
 ficial effect on the strength of the mixture. In Fig. 243 we may 
 see how the curve showing tensile strength rises to a maximum 
 at the point corresponding to the eutectic of the series of lead-tin 
 alloys. 
 
 The Freezing of Iron and Steel 
 
 1. "When the iron and steel alloys are liquid, they are com- 
 posed of liquid carbon dissolved in liquid iron, and their freezing- 
 point curves are shown in Fig. 244, p. 287. It will be noticed 
 that the lines in this figure bear a close similarity to those in Fig, 
 241. This similarity is real, and the laws governing the freezing of 
 this series of alloys are very similar to those governing the freezing 
 of the lead-tin alloys. There is a eutectic of this series when the 
 line A B crosses the line C B, and the components of this eutectic 
 are 95.7 per cent, iron and 4.3 per cent, carbon. In the study of 
 the iron-carbon alloys, however, we have to take into account a 
 solid solution which forms. That is to say, we must remember 
 that iron, if it contains as much at 2.2 per cent, of carbon, never 
 separates from the liquid state without carrying 2.2 per cent, of 
 carbon with it in solid solution.^ It will be remembered that 
 when lead separated from the liquid solution, it carried with it a 
 small amount of tin as an impurity, and that when tin separated, 
 it carried with it a small amount of lead as an impurity; but we 
 disregarded these traces of impurity for the sake of simplicity in 
 outlining the laws of solutions. In the case of iron-carbon alloys, 
 however, the carbon carried out with the iron in solid solution, 
 
 ^ This solid solution may consist of carbon dissolved in the iron, or of iron carbide dis- 
 solved in the iron. We do not know definitely which, but it is not necessary to discuss this 
 question just yet. If it contains less than 2.2 per cent, carbon, it retains all of it in solid 
 solution. 
 
THE SOLUTION THEORY OF IRON AND STEEL 
 
 287 
 
 substantially as an impurity, is too important in its effect upon 
 the material to permit us to neglect it. 
 
 2. The line X Y therefore divides the diagram in Fig. 244 into 
 two parts. Everything to the left of this line freezes as a solid 
 solution and the laws are similar to the freezing laws of the gold- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1600°C 
 A 
 
 
 b 
 
 
 
 
 
 
 
 
 
 
 2&i21^ 
 
 ^ 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 1*10 
 
 ^ 
 
 s 
 
 ^ 
 
 d 
 
 
 
 e 
 
 
 
 
 
 2552* 
 
 
 ^ 
 
 ts 
 
 ^ 
 
 - 
 
 
 
 
 
 
 
 1200° 
 
 
 
 ^*^ 
 
 \ 
 
 
 
 
 ^^ 
 
 
 
 C 
 
 2193'' 
 
 
 
 
 \ 
 
 \ 
 
 a 
 
 
 
 
 
 \B^ 
 
 ^^ 
 
 
 
 EutJ 
 
 ctic fre 
 
 ezes 
 
 
 n 
 
 
 1000° 
 
 
 
 
 
 
 
 
 
 
 
 
 isaa"* 
 
 
 
 
 
 
 
 
 
 
 
 
 800° 
 
 
 
 
 
 
 
 
 
 
 
 
 U72° 
 
 
 
 
 
 
 
 
 
 
 
 
 600° 
 
 
 
 
 
 
 
 
 
 
 
 
 wst 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1( 
 
 lOiiro 
 
 a S 
 
 L 36 Carb 
 95<Iron 
 
 on i 
 
 
 .r 
 
 ; 
 
 i% 
 
 i 
 
 i% 
 
 6^ 
 95^ 
 
 Carbon 
 Iron 
 
 FIG. 244.-THE FREEZING OF ALLOYS OP IRON AND CARBON. 
 
 silver alloys. Everything to the right of the line X Y freezes 
 selectively, according to the same laws as those given in the case 
 of the lead-tin alloys. It is because of this difference in the freez- 
 ing of the alloys that the line Z F is arbitrarily considered as the 
 dividing line between steel and cast iron. That is to say, all the 
 alloys with less than 2.2 per cent, carbon are defined as steel, and 
 all with more than 2.2 per cent, carbon are defined as cast iron. 
 
288 THE METALLURGY OF IRON AND STEEL 
 
 3. Freezing of Steel. — All the steels freeze as solid solutions. 
 Let us consider a solution of 99.5 per cent, iron and 0.5 per cent, 
 carbon at 1650° C. This will be represented by the point h in Fig. 
 244. This cools until it meets the line A B, and now it commences 
 to solidify. For a few degrees of temperature it is part liquid and 
 part solid, but by the time it has fallen to a temperature where it 
 meets the line Aa, ithas become entirely solid, and it is now a 
 homogeneous solution of 0.5 per cent, of carbon in iron. Con- 
 sider next a solution containing 99 per cent, of iron and 1 per cent, 
 of carbon, at the point c in Fig. 244. When this cools to the 
 temperature where it crosses the line A Bit commences to solidify 
 and it is in a partly liquid and partly solid condition until it 
 crosses the line Aa^ upon which solidification is completed and it 
 now becomes a homogeneous solution of, 1 per cent, carbon in 
 iron. The satne actions take place with an alloy containing 98.5 
 per cent, iron and 1.5 per cent, carbon (the point d in Fig. 244), 
 and also in the case of 98 per cent, iron and 2 per cent, carbon. 
 In all these alloys we finally arrive at a solid solution of carbon in 
 iron.^ To this solid solution the name of "austenite" is given, and 
 this name applies no matter how much or how little carbon is in 
 solid solution. In other words, all steels are in the condition of 
 austenite as soon as their solidification is complete. 
 
 4. Freezing of Cast Iron, — The freezing of cast iron is shown 
 by the diagram to the right of the line X Y, and if we should con- 
 sider this part as a separate diagram, then it would be similar to 
 Fig. 241, the freezing of the lead-tin alloys. There is one differ- 
 ence to be borne in mind, however: along the line A Bin Fig. 241 
 we had a selective precipitation of lead; along the line A Bin Fig. 
 244 we have a selective precipitation, not of pure iron, but of iron 
 containing 2.2 per cent, of carbon. In other words, this entity, 
 consisting of a solid solution of iron with 2.2 per cent, of carbon, 
 behaves as if it were an elemental substance. It is sometimes 
 called " 2.2 per cent, austenite. " Suppose we have, for example, 
 a liquid solution of 2.5 per cent, of carbon in iron at a temperature 
 of 1400° C. This will be represented by the point e in Fig. 244. 
 The liquid solution will cool until it reaches a temperature of 
 about 1320° C. At this point there will begin to precipitate the 
 entity of which I have spoken, namely, iron containing 2.2 per 
 
 1 The formation of the solid solution from the liquid solution is discussed in detail in 
 Professor Howe's *'Iron, Steel and Other Alloys," but it requires too much spacie to be 
 discussed here. For our purpose it is sufficient to know that when freezing la completed, 
 we have a homogeneous solid solution of all of the carbon in all of the iron. 
 
THE SOLUTION THEORY OF IRON AND STEEL 289 
 
 cent, carbon in solution. This precipitation will cause the 
 liquid solution to be impoverished in iron, and it will therefore 
 move to the right in the diagram as the temperature falls; or, in 
 other words, it will travel down the line A B. By the time the 
 temperature of 1 135° C. is reached, a large amount of 2.2 per cent, 
 austenite will have precipitated, and the small amount of liqiiid 
 solution left will be at the point Bj that is, the eutectic point, 
 where there is 4.3 per cent, carbon. With further cooling the 
 eutectic will cross the point B, and therefore will complete its 
 precipitation and its freezing. It decomposes into crystals, a 
 part of which are of tiny flakelets of the austenite before men- 
 tioned and the other part, crystals of graphite — i,e,, carbon. 
 
 5. A similar result will be obtained in the case of a 3 per cent, 
 liquid solution of carbon in iron. This will cool until it reaches 
 1280° C, where the 2.2 per cent, austenite will precipitate, de- 
 creasing the residual solution in iron so that it travels down the 
 line A B and finally reaches the point B, after which this eutectic 
 solution at the point B will precipitate as before. 
 
 6. We may sum up paragraphs 4 and 5 by saying that any 
 solution of iron containing more than 2.2 per cent, and less than 
 4.3 per cent, of carbon will consist, after freezing, of a eutectic 
 together with a certain amount of previously precipitated 
 austenite (consisting of iron with 2,2 per cent, of carbon in solid 
 solution) . 
 
 7. What will occur in case the solution contains more than 4.3 
 per cent, of carbon?^ Take, for example, a liquid solution con- 
 taining 4.7 per cent, of carbon at a temperature of 1200° C. This 
 will cool until a temperature of about 1170° *is reached and the 
 line C B h met. As cooling proceeds to lower temperatures, 
 carbon (i.e., graphite) precipitates out and the liquid solution 
 remaining moves to the left in the diagram. That is to say, it 
 travels down the line C B. When the temperature 1135° is 
 reached, so much graphite is precipitated that the residual solution 
 is now of the eutectic proportions. .With further cooling, this 
 eutectic breaks up as before, consisting thereafter of crystals of 
 graphite and of austenite with 2.2 per cent, of carbon. 
 
 8. Summary, — To sum up, then, all the solutions of iron and 
 carbon containing less than 2,2 per cent, of carbon will consist, 
 after solidification, of a solid solution of iron and carbon having 
 
 . ^ It 13 very seldom that solutions contain more than 4.3 per cent, of carbon, because this 
 much carbon does not readily diasolve in iron. 
 19 
 
290 THE METALLURGY OF IRON AND STEEL 
 
 the same chemical composition as the original liquid solution, and 
 being a homogeneous solution of one in the other. There can be 
 no eutectic form if there is not more than 2.2 per cent, carbon. 
 All the solutions with more than 2.2 per cent, of carbon will con- 
 sist, after solidification, of a eutectic together with a certain 
 amount of previously precipitated graphite or of the previously- 
 precipitated 2.2 per cent, austenite mentioned. 
 
 Changes on Heating. — In the foregoing discussion we have been 
 considering only the changes that occur during the cooling of 
 the metal. These changes are simply reversed on heating. 
 
 We have seen that every alloy of iron and carbon contains, 
 after solidification, a varying amount of the solid solution of iron 
 and carbon. For example, if we started with 1 per cent, of 
 carbon, then, immediately after solidification, we would have a 
 s61id solution of 1 per cent, carbon in iron; if we started with 1.5 
 per cent, of carbon, we would have a solid solution containing 1.5 
 per cent.; if we started with 2.2 per cent, of carbon, then we 
 would have a solid solution of 2.2 per cent, carbon. Even if we 
 start with more than 2.2 per cent, carbon, then the alloy, after 
 solidification, will consist partly of a solid solution containing 2.2 
 per cent, of carbon and partly of graphite. 
 
 Now, what becomes of these solid solutions which make up a 
 part or a whole of the cast iron and steel alloys when they freeze? 
 Does the carbon remain in solid solution down to the atmospheric 
 temperature, or does it precipitate, or does it undergo some other 
 change? We find by experiment that the solid solutions do not 
 survive, but precipitate at a lower temperature, and the laws 
 governing the decomposition of these solid solutions are similar 
 to the laws governing the decomposition of liquid solutions — 
 for example, the lead-tin solutions. In short, we have another 
 series of curves showing the selective precipitation of the con- 
 stitutents of these solid solutions, and the only difference between 
 the nature of these curves and the lead-tin curves is that these 
 represent changes taking place in the solid state, while the lead- 
 tin curves represent changes taking place in the liquid state. 
 
 Nature of the Solid Solution. — In the footnote on page 286 I 
 called attention to the fact that the solid solution of iron and 
 carbon might be a solution of pure carbon in iron, or a solution of 
 a carbide of iron in iron, for example, of FegC in iron. Several 
 authorities hold this view,^ while other maintain that the solution 
 
 1 Indeed, one authority believes that there are probably several different carbides (FejC, 
 FeaC, FeC) dissolved in the iron. 
 
THE SOLUTION THEORY OF IRON AND STEEL 
 
 291 
 
 is of elemental carbon in iron. The question is of more academic 
 than practical interest. The important thing is that, when the 
 solution decomposes, it is a carbide of iron which precipitates. 
 Those who believe that the solid solution is composed of elemental 
 carbon and iron explain the precipitation of the carbide by main- 
 
 1600C 
 
 1400 
 
 1200 
 
 1000 
 
 800 
 
 M 
 
 600 
 
 ^ 
 
 
 100 1^ Iron 
 
 1{£ Carbon 
 
 2913r- 
 
 1633 
 
 1478** 
 K 
 
 356 
 
 4« 
 
 5?£ Carbon 
 95?^ Iron 
 
 FIG. 245.-DECOMPOSITION CURVES OF THE SOLID SOLUTIONS OF IRON 
 AND STEEL. (ALSO KNOWN AS THE CRITICAL POINTS.) 
 
 taining that when the carbon separates from solution, it immedi- 
 ately unites with iron and forms a carbide, usuUy FegC. With 
 this explanation I shall hereafter, for simplicity's sake, discuss 
 the solid solutions as if they were FegC in iron. 
 
 The Decomposition of the Solid Solutions. — The curves of de- 
 composition of the solid solutions are shown in Fig. 245. The line 
 GO Sis the line upon which there is selective precipitation of pure 
 
292 THE METALLURGY OF IRON AND STEEL 
 
 iron.> To this pure iron the name of " f errite " has been given by 
 Professor Howe, and this name meets with universal acceptance. 
 
 Consider first a solid solution containing 0.40 per cent, of 
 carbon at a temperature of 800° C. This will be at the point h in 
 Fig. 245. It will cool until it reaches a temperature of about 
 780° C: upon which ferrite will begin to precipitate. As the 
 temperature continues to fall more and more ferrite precipitates, 
 which impoverishes the solid solution in iron and causes it to 
 travel down the line S, By the time the temperature has 
 reached 690°, the solid solution has reached the point Sj corres- 
 ponding to 0.90 per cent, carbon. 
 
 Consider next an alloy containing 1.60 per cent, carbon at 
 1000° C. : this will be at the point k. It will cool until it reaches 
 a temperature of about 970°; at which carbide of iron (Fe^C) will 
 begin to precipitate. This precipitation continues as the tem- 
 perature falls, constantly decreasing the amount of carbon in 
 the solid solution, which therefore travels down the line E S 
 until, at 690°, it reaches the point Sj where there is 0.90 per cent, 
 carbon. 
 
 A similar precipitation will take place with all of the solid 
 solutions of iron and carbon: if they contain less than 0,90 per 
 cent, carbon, they will begin to. precipitate out ferrite when they 
 fall to the line G S. If they contain more than 0.90 per cent, 
 carbon, they will begin to precipitate carbide of iron when they 
 meet the line E S, In either case the residual solid solution will 
 travel down the line G S, or else E S, until it reaches the point 
 S when the temperature has fallen to 690° C; we will then have 
 some solid solution left containing 0.90 per cent, carbon, and 
 mixed with this some previously precipitated ferrite or cementite, 
 as the case may be. 
 
 Eutectoid. — The alloy containing 0.90 per cent, carbon is 
 known as the "eutectoid alloy," a name invented by Professor 
 Howe to indicate that the formation of this alloy, which results by 
 selective precipitation of the solid solution, is similar to the forma- 
 tion of the well-known eutectics of liquid solutions. When this 
 eutectoid solid solution cools below 690° C, it is completely de- 
 composed into its constituents, ferrite and cementite. These 
 constituents precipitate in tiny flakelets, which arrange them- 
 selves in the banded structure already familiar to us as the 
 structure of eutectics. A magnified view of the structure is 
 shown in Fig. 247, while Fig. 248 shows the magnified structure 
 
THE SOLUTION THEORY OF IRON AND STEEL 293 
 
 of a piece of steel composed of a eutectoid together with previ- 
 ously precipitated ferrite; and Fig. 249 shows the structure of a 
 steel consisting of the eutectoid with previously precipitated 
 cementite. 
 
 It will be evident that there will be some of the eutectoid in 
 every piece of iron or steel, for even the cast irons .contain, after 
 solidification, a certain amount of solid solution which precipi- 
 tated either while the liquid alloy was traveling down the line A B, 
 or during the freezing of the liquid eutectic (containing 4.3 per 
 cent, of carbon), or both. The formation and characteristic of 
 this eutectoid are therefore of very great importance. Its pres- 
 ence in iron and steel was known long before the theories that I 
 have outlined in this chapter had begun to be understood, and 
 the name of " pearlite " was given to it because, under certain cir- 
 cumstances, it has the appearance of mother-of-pearl. The 
 name pearlite is only used to designate the eutectoid after its 
 complete decomposition and the separation of the ferrite and 
 cementite into the banded structure shown in Fig. 250. 
 
 Changes on Heating. — As in the case of the liquid solutions, the 
 changes already described above in connection with the cooling of 
 the solid solutions are reversed on heating. The temperatures at 
 which the reverse changes occur on heating, however, are some- 
 what higher than the corresponding, change on cooling, unless 
 both cooling and heating be very slow. In other words, the 
 chemical changes seem to lag behind the temperature changes. 
 With extremely slow heating and cooling, the lag is but slightly 
 - perceptibla, and the various changes and reversals take place at 
 almost the same temperature, but with the rates of cooling 
 usually employed in research laboratories at steel works, there 
 is a difference of about 20° between the temperatures of the 
 changes in the different directions. It is customary to plot the 
 curves as in Figs. 241,- 244, showing temperatures on cooling, and 
 that method has been adopted here. 
 
 The Complete Roberts-Austen, Roozeboom Diagram 
 
 We may now take the diagrams of Figs. 244 and 245 and 
 combine them into one diagram which shall represent all these 
 changes in the heating and cooling of iron and steel. Such a 
 diagram is shown in Fig. 252. Nevertheless, this diagram does 
 not by any means tell the whole story; there are many changes 
 
294 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 246.— EUTECTIC OF COPPER 
 
 AND SILVER. 
 
 (William Campbell.) 
 
 FIG. 247.— PEARLITE EUTECTOID 
 
 OF IRON AND CARBON. 
 
 (F. Osmond.) 
 
 FIG. 248.— PEARLITE AND FERRITE. 
 250 diameters. 
 
 FIG, 249.— PEARLITE AND CEMEN- 
 
 TITE. 
 
 250 diameters. 
 
 FIG. 250.— PEARLITE. FIG. 251,— PEARLITE. 
 
 250 diameters. 1000 diameters. 
 
 Picric Acid.— Large crystalliisation due to long overheating. Polished in Relief. 
 
THE SOLUTION THEORY OF IRON AND STEEL 
 
 295 
 
 whose nature is not yet understood, but evidence of which is 
 conclusive. We must leave most of these to treatises devoted 
 to this special branch of iron and steel metallurgy, on account of 
 lack of space here, but one important modification must be in- 
 cluded because of its great usefulness in controlling the properties 
 of cast iron; this modification is the effect of silicon on the changes 
 occurring in freezing. 
 
 
 
 
 
 
 
 
 
 
 
 
 leoo'b 
 
 
 
 
 
 
 
 
 
 
 
 ^>^ 
 
 
 
 
 
 
 
 
 
 
 mx? 
 
 X 
 
 s. 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 \, 
 
 
 \ 
 
 ^ 
 
 
 
 
 
 lawf 
 
 
 
 \ 
 
 \ 
 
 
 ^ 
 
 
 •"^ 
 
 
 c 
 
 
 
 
 \ 
 
 \ 
 
 
 
 X 
 
 ^^^ 
 
 ^-^ 
 
 
 V 
 
 
 
 
 
 
 100(f 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 F 
 
 SMf 
 M 
 P 
 
 600" 
 
 \ 
 
 
 / 
 
 
 
 
 
 
 
 
 ;< 
 
 s 
 
 / 
 
 V 
 
 
 
 
 
 
 
 
 ] 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 2552' 
 
 2198° 
 D 
 
 
 lOQ^IroiL 
 
 in Carboa 
 
 2* 
 
 d% 
 
 m 
 
 1472" 
 
 K 
 
 1112" 
 
 5!( Carbon 
 95^ Iron 
 
 FIG. 252.-THE FREEZING AND SOLID DECOMPOSITION CURVES OF 
 THE IRON ALLOYS. 
 
 Effect of Silicon on the Freezing of Cast Iron, — The changes 
 shown in Figs. 244 and 252 refer to the freezing of normal gray 
 cast iron used for iron castings and cooled moderately slowly as in 
 the practice of founding, etc. Under such conditions there is a 
 formation of austenite with graphite, the austenite subsequently 
 decomposing into ferrite and iron carbide (cementite) j as de- 
 scribed. Indeed, when there is some 2 per cent, or so of silicon 
 
296 
 
 THE METALLURGY OF IRON AND STEEL 
 
 present as there often is in foundry irons, the austenite decomposes 
 in large part into ferrite and graphite, instead of ferrite and 
 cementite, so that such irons, when cooled to atmospheric tem- 
 perature, will consist largely of ferrite and graphite, with only a 
 small amount of combined carbon. It is this freedom from much 
 cementite that causes iron castings to machine so readily. In 
 short, silicon tends to produce a precipitation of graphite instead 
 
 Cast-iron 
 
 Molten Metal 
 
 400 
 300 
 200 
 100 
 
 (rt^.^^^^'" Graphite 
 ■fyi jf^^^'^"" contlnoee to freeze. 
 .Zj^^^jM^ecEic here freezes 
 
 Solid 
 
 /Austenite + eutectic Graphite Graphite + eutectic Austenite 
 
 K 
 
 Legend: 
 Graphite -Austenite dia^rram 
 
 Cementite -Austenite diagram*' 
 
 o 
 
 _L. 
 
 _!_ 
 
 _L- 
 
 _1_ 
 
 _L_ 
 
 _1_ 
 
 TJl 
 D 
 
 CarbonO 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 6.67 
 Iron 100 99.5 99.0 98.5 98.0 97.5 97.0 96.5 96.0 95.5 95.0 94.5 94.0 98.5 93.33 
 
 Percentage Composition 
 
 FIG. 253.— THE STABLE AND METASTABLE IRON-CARBON DIAGRAM. 
 
 of cementite. This effect of silicon operates not only during the 
 decomposition -of the eutectoid, but also during freezing. 
 
 In other words, when we experiment in the laboratory with 
 pure iron and carbon, free from silicon and other impurities — a 
 kind of iron, be it noted, that does not exist commercially — the 
 products of freezing are not austenite and graphite, but austenite 
 and cementite. Such an iron, when cooled to atmospheric 
 temperature, will consist of eutectoid plus cementite, and no 
 graphite will be found in it. This result is represented by the 
 
TPIE SOLUTION THEORY OF IRON AND STEEL 297 
 
 dotted lines of Fig. 253. When cooling pure irons, free from 
 silicon, etc., the product will be according to these dotted lines, 
 and the word cementite should be substituted at every point 
 where graphite occurs. The full lines, on the other hand, 
 represent the action when commercial irons are cooled. 
 
 Effect of Rate vf Cooling. — None of the changes represented by 
 these diagrams are instantaneously completed. This enables 
 us to control them to a limited extent. For example, in the case 
 of a cast iron (unless the silicon be very high) we can prevent 
 the formation of any graphite by cooling the liquid metal very 
 rapidly, as, for instance, by pouring it into cold water. The 
 result will be that most of the carbon will remain dissolved in the 
 iron as it was in the liquid state, while the remainder will be in 
 the form of cementite. In the case of steel, we may largely 
 prevent the separation of ferrite and cementite from the austenite 
 by cooling the steel rapidly from the temperature where austenite 
 exists, i.e.j above the lines G S a. 
 
 Other Lines in Fig. 253. — Certain retardations in the cooling 
 curves would indicate that there was a line extending all the way 
 across this dagram at about 775° C, and another at 600° C. 
 
 Roberts- Austen. — ^The diagram of Fig. 252 is often known as the 
 Roberts-Austen diagram, after Sir William Roberts-Austen, be- 
 cause the cooling curves that located the lines were first deter- 
 mined in his laboratory.* 
 
 Upton's Diagram. — One of the latest theories to help clarify 
 the equilibrium of the iron carbon alloys is that of Upton, and he 
 has collected his hypotheses into the diagram shown in Fig. 254. 
 It will be seen that this diagram involves the acceptance of three 
 different carbides of iron. There are certain suppositions in this 
 diagram which are not correct, but it has gone a long way toward 
 explaining several of the observed phenomena and as such should 
 be familiar to all metallurgists. It does away with the theory 
 of a stable and metastable eutectic, and assumes that the eutectic 
 consists of austenite plus graphite, which solidifies a*t 1145° C, 
 but that, at about 1095° C. the compound FegC is formed. If 
 there be less than 3.46 per cent, of total carbon present, the for- 
 mation of this Fe^C will use up all the graphite and the alloy 
 
 ' See "Proceedings of the Institution of Mechanical Engineers" (England), 1897, Fourth 
 Report of the Alloys Research Committee, Plate 2. See aho "Le Fe et I'Acier au Point de 
 Vue de la Doctrine des Phases," H. W. Bakhuia-Roozeboom, in Zeitschrift fur physikaliscke 
 Chemie, vol. xxxiv, p. 437: French translation, "Contribution k TEtude des Alliagea," pp. 
 327-386, Paris, 1901, 
 
298 
 
 THE METALLURGY OF IRON AND STEEL 
 
 will contain only gamma iron and Fefi in solution. If there be 
 more than 3.46 per cent, of total carbon present then only the 
 excess of carbon over 3.46 per cent, will appear in the form 
 of graphite. 
 
 However, the recombinations of graphite and formation of 
 Fefi is a very slow reaction indeed, taking place only to a limited 
 degree in any industrial cooling and being therefore of theoretical 
 rather than of practical application, except that the tendency 
 
 12 3456789 
 
 FIG. 254.— UPTON'S DIAGRAM OF THE IRON-€ARBON ALLOYS. 
 
 for this action to occur explains certain hitherto puzzling re- 
 actions, such as for instance, the reabsorption of temper carbon 
 when malleable cast iron is over-annealed. 
 
 On further cooling the Fefi breaks up, and FegC is formed, 
 which would result in the absorption of more graphite if the cool- 
 ing were sufficiently slow, but_ again the extreme -slowness of this 
 reaction causes it to take place completely only under such con- 
 ditions as prevail in the laboratory, as a rule. For further dis- 
 cussion of this very interesting theory, the reader is referred to 
 the original article in the Journal of Physical Chemistry, Vol. 
 XII, No. 7, October, 19Q8, pp, 507 to 549, 
 
XI 
 
 THE CONSTITUTION OF STEEL 
 
 The properties of steel depend upon the chemical composition 
 of its constituents as well as upon their size and relation to one 
 another. Enough has been said to show that steel is not a 
 simple homogeneous union of iron with varying proportions of 
 carbon, silicon, manganese, etc.; but is built up of individual 
 crystals somewhat in the same way as crystalline rocks are 
 formed — granite, for example. But while the crystals of granite 
 are generally visible to the naked eye, and its structure may 
 therefore be determined by a more or less cursory examination, 
 the structure of steel is visible only by means of the microscope 
 and after careful polishing, sometimes followed by chemical treat- 
 ment to differentiate between the various grains. Nevertheless, 
 the structure of steel is of great importance, and in some cases, 
 perhaps, is even more so than the chemical composition. 
 
 The Micro-constituents of Steel 
 
 In this chapter I shall speak only of slowly cooled steel except 
 where I have indicated the contrary. We have already learned 
 that slowly cooled steel must necessarily contain ferrite and 
 cementite, resulting from the decomposition of the solid solution 
 of iron and carbon. There are also other constituents which are 
 found under the microscope, or separated by chemical analysis, 
 or in both ways. These latter constituents are ' compounds of 
 iron with various other impurities, such as iron sulphide, iron 
 phosphide, and iron silicide; or of two impurities with each other, 
 such as manganese sulphide. 
 
 Ferrite. — ^Ferrite is theoretically pure iron, and especially iron 
 free from carbon. It is weak as compared with several of the 
 other constituents, having a tensile strength of about 45,000 to 
 50,000 lb. per sq. in; it is also very soft and ductile, re- 
 sembling copper in these properties, and has, furthermore, a high 
 degree of malleability. It has a very high electrical conductivity 
 as compared with the other constituents of iron and steel, and 
 
 299 
 
300 THE METALLURGY OF IRON AND STEEL 
 
 about one-seventh the conductivity of copper and silver, the best 
 conductors known. (Copper and silver are of nearly the same 
 conductivity.) Its magnetic force is the highest of any known 
 substance, its magnetic permeability is high, and its hysteresis 
 low. It crystallizes in the isometric system. 
 
 Ferrite is an important constituent of all steels and the pre- 
 dominant one in all the low-carbon steels. The industrial prod- 
 uct approaching nearest to pure ferrite is wrought iron, if we 
 disregard the slag, which, being mechanically mixed with the 
 mass, does not appreciably alter its chemical and physical be- 
 havior. It is for this reason that wrought iron is so useful where 
 a soft and ductile material is necessary, as in boilers for instance; 
 or where high electrical conductivity is demanded, as in telegraph 
 wire; or a high degree of magnetism, as in the cores of electro- 
 magnets. The wrought iron made in Sweden, and known as 
 '* Norway iron," is greatly preferred for this latter purpose, on 
 account of its purity. 
 
 Under the microscope ferrite may be distinguished from ce- 
 mentite by its softness. If steel containing these two constituent, 
 be polished on damp, rough parchment, or on chamois skin 
 stretched over a soft background (as wood), the ferrite will wear 
 away below the carbide and appear in intaglio. The same effect 
 will be obtained by Osmond's ''polish attack."^ Ferrite is also 
 distinguished from carbide of iron by the fact that, after being 
 subjected to the brief action of certain reagents, such as 2 per 
 cent, nitric acid, or ordinary commercial tincture of iodine, the 
 ferrite is seen in darker grains and the carbide in bright thin 
 plates, unattacked by the reagent. When the two are intimately 
 associated in minute grains, as in pearlite, the carbide appears 
 bright and the ferrite dark, because eaten away below the surface 
 by the reagent. 
 
 Allotropic Modifications. — ^There is one peculiarity of pure iron, 
 or ferrite, which deserves special attention, namely, its ability 
 to assume different allotropic modifications at different temperar- 
 tures. The nature of aUotropism has been explained in Chap. 
 XX). To the allotropic modification of iron the names of "aldha," 
 ''beta," and "gamma" have been assigned. The alpha modifica- 
 tion, is the one existing at atmospheric temperatures, and it is 
 familiar to all who make or use iron. If this be heated, however, 
 it undergoes a sudden change at about 760° C. (1390° F.). This 
 
 1 See page 447. 
 
THE CONSTITtJTION OF STEEL , 301 
 
 change is evidenced by an absorption of heat and the circum- 
 stance that the iron loses almost entirely its power to attract a 
 magnet; that is to say, it becomes about as non-magnetic as 
 lead, copper, etc. The change in magnetism is accompanied by 
 a change in electrical conductivity, in specific heat, and in other 
 properties. In short, the iron has changed in many of its prop- 
 erties "without undergoing any alteration in chemical composi- 
 tion. This new allotropic modification of iron is known as beta 
 iron. 
 
 If, now, beta iron be further heated to a temperature of about 
 890° C. (1634° F.), it again changes several of its properties and 
 becomes what is known as gamma iron. Gamma iron differs 
 from beta iron, especially in electrical conductivity and in orys- 
 talline form. Ferrite crystallizes always in the cubic system, and 
 Osmond^ and Stead^ have studied the variations of form assumed 
 by it and by its alloys with carbon. Osmond especially has 
 studied the crystallography of the gamma, beta, and alpha modi- 
 fications of the pure metal. Gamma iron does not crystallize 
 isomorphously with either beta or alpha iron, which crystallize 
 identically in cubes; but it assumes all the combinations of the 
 cube and octahedron, and, in the latter form would be isomor- 
 phous with carbon in the diamond form. Therefore, in the 
 isomorphous mixtures (i.e., solid solutions) of iron and carbon 
 one would expect to find some carbon in the diamond form, which 
 has indeed been accomplished by Osmond. This is used as an 
 argument in favor of the belief that the solid solution is with 
 carbon, not with cementite; for Osmond has shown^ that 
 cementite does not assume any crystalline form which would 
 be isomorphous with ferrite. Beta and alpha iron do not 
 crystallize isomorphously with either carbon or cementite, 
 which accords with the observed tendency of ferrite to begin 
 to separate from the solid solution at the same temperature at 
 which it changes from gamma to beta. 
 
 If ferrite is in the gamma form, and say at 1000° C. (1832° 
 F.), it will undergo upon cooling the reverse of the changes which 
 have already been described. That is to say, at about 890° C, 
 (1634° F.) it will change from gamma to beta iron; at 760° 
 (1390° F.) it will change from beta to alpha iron, in each case re- 
 ceiving again the properties which it had before heating. In 
 other words, the change from one allotrophic form to another is 
 
 1 See No. 110. 2 gee No. 181. ^ See No. 181 
 
302 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Dqocu 
 
 | < ■ -B ° > [CBntLsradej4-- 
 
 Evolutlon — 
 Of Heat -K 
 
 Sodden Evolution 
 or Heat 
 
 G 
 
 a reversible change, taking place in one direction on cooling and 
 in the opposite direction on heating. The change on cooling 
 takes place at a slightly lower temperature than on heating. 
 This is not because it is other than a true reverse action, but be- 
 cause the change in either direction is neces- 
 sarily slow and lags a little behind the 
 temperature. The amount of this lag will 
 vary directly with the speed of heating and 
 cooling. 
 
 We never get pure gamma iron at any 
 temperature unless we start with iron free 
 from carbon, because gamma ferrite never 
 separates from the solid solution. If, how- 
 ever, we have a solid solution of iron with, 
 let us say, 0.2 per cent, of carbon, then this 
 solid solution will begin to precipitate ferrite 
 at a temperature of about 830° C. (1524° F.). 
 This ferrite will be, of course, in the beta 
 form and wiU be non-magnetic until the mass 
 cools below 760° (1390° F.), when the pre- 
 viously precipitated ferrite will change from 
 the beta to the alpha form." Any ferrite 
 which separates from solid solution after 
 that temperature is reached will separate in 
 the first instance in the alpha form. For 
 example, if we have a solid solution, con- 
 taining, say, 0.7 per c6nt. carbon, this will 
 commence to precipitate ferrite below 760° C, 
 (about 720°), and the ferrite will be in the 
 alpha form.^ 
 
 Cementite, — The carbide of iron is, next to 
 ferrite, the most important constituent of 
 steel, and practically all of the carbon is present in this form.^ 
 Cementite is very hard and brittle, scratching glass with ease 
 and flying into pieces under a blow. It crystallizes usually in 
 thin flat plates, which are large in size (sometimes up to 1/8 in. 
 
 1 Some maintain that the ferrite always precipitates as gamma ferrite, then immediately 
 changes to beta and next to alpha ferrite. Others maintain that when the solid solution 
 is cooled near the line G S,\t changes from a solid solution of gamma iron into a solid 
 solution of beta iron, and then into a solid solution of alpha iron, from which alpha iron 
 then precipitates. (See No. Ill and page 378.) 
 
 ^ There is not wanting evidence in favor of several carbides being present, such aa FeC, 
 FeaC, FesC, FeeC, etc., but FeaC is the one usually considered. 
 
 FIG. 260.— C OGLING 
 CURVE OF PURE 
 IRON. 
 
THE CONSTITUTION OF STEEL 303 
 
 in diameter) when there is much cementite present. It is at- 
 tacked by reagents less than most of the other constituents and 
 is in this way distinguished under the microscope. It is a little 
 difficult to distinguish, by microscopic evidence alone, between 
 steel consisting of pearlite with a slight excess of cementite and 
 steel consisting of pearlite with a slight excess of ferrite. The 
 practised eye can usually tell; but a chemical analysis readily 
 distinguishes, since steel with less than 0.8 per cent, carbon will 
 have excess ferrite over pearlite, and that with more than 0.9 
 per cent, will have excess cementite. 
 
 Cementite contains 6.6 per cent, of carbon, -or roughly, is one- 
 fifteenth carbon. We may therefore tell the amount of cementite 
 in any steel by multiplying the amount of carbon by fifteen.^ 
 Cementite may be separated from steel by electrolysis.^ It is 
 magnetic at ordinary temperatures, but not above 700° C. (1292° 
 
 F.). 
 
 The carbon united with iron in cementite has been given 
 various names, such as " cement carbon," or " carbon of cementa- 
 tion'' (because of its prominent appearance in cemented steels), 
 and "carbon of the normal carbide," "annealing carbon'' (be- 
 cause all the carbon of well-annealed steels will usually be present 
 as cementite). 
 
 Manganiferous Cementite, — Manganese forms a carbide having 
 the formula MugC. This is isomorphous with FogC, and we often 
 find the two carbides together in one crystal. The name cemen- 
 tite is still applied to this crystal, although it must be recognized 
 that a part of the iron has been replaced by manganese. The 
 formula for the compound is usually written (FeMn)3C. The 
 amount of manganese in these crystals is very variable, running 
 , almost all the way from nothing to 90 per cent. As manganese 
 has an atomic weight almost the same as that of iron (Mugg, Fogg), 
 one weight of manganese will replace almost exactly an equal 
 weight of iron in the crystal. The peculiarity of the manganif- 
 erous cementite is that the crystals of free cementite are liable to 
 be larger, especially when the proportion of manganese is large, 
 and are seemingly harder and more difficult to machine. 
 
 Manganese Sulphide. — Manganese and sulphur unite to form 
 manganese sulphide, having the formula MnS, and this compound 
 
 ^ In other words, steel containing . 5 per cent, of carbon will contain 7 . 5 per cent, 
 cementite; steel containing 1 per cent, carbon will contain 15 per cent, cementite; etc. 
 2 See No. 112. 
 
304 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 261.~FERRITE, PURE IRON 
 
 (ELECTROLYTIC). 
 1000 diametera. Etched with picric acid. 
 
 FIG. 262.— PEARLITE CRYSTALS. 
 
 Surrounded by ferrite. 250 diameters. 
 
 Etched with HNO3. 
 
 FIG. 263.— BIG CEMENTITE CRYSTAL 
 
 IN PEARLITE. 
 
 250 diameters. PoUshed in relief. 
 
 FIG. 264.— CRYSTALS OF MANGA- 
 
 NIFEROUS CEMENTITE. 
 50 diameters. Etched with nitric acid. 
 
 FIG. 265.— ELONGATED BUBBLE OF 
 
 MANGANESE SULPHIDE.' 
 
 250 diameters. Unetched. 
 
 FIG. 266.— EUTECTIC OF FesP AND 
 
 IRON. 
 1000 diameters. Etched with picric acid. 
 
THE CONSTITUTION OF STEEL 305 
 
 is found in all steels. Indeed, all of the sulphur will be found 
 in this combination, provided there is enough manganese in the 
 steel to unite with it. It is necessary to have more than the theo- 
 retical amount of manganese for this purpose, however, because 
 unless there is a surplus present, the attraction of the manganese 
 for the sulphur does not seem to be always sufficient to catch it all. 
 Steel should therefore always contain about four times as much 
 manganese as sulphur, because it is advantageous to have the 
 sulphur all in the form of manganese sulphide. 
 
 Manganese sulphide is seen under the microscope as a dove- 
 gray substance before the polished material is etched with any 
 reagents. It is usually segregated in round drops, which are 
 sometimes, if large in size, seen to be elongated by the rolling 
 or hammering of the material (see Fig. 265). 
 
 Manganese tends to make the crystals of steel smaller, which 
 is advantageous, but makes the metal more liable to crack in 
 heating, and still more so in cooling suddenly from a red heat. 
 
 Iron Sulphide. — The bulk of the sulphur not united with the 
 manganese will be found in the form of iron sulphide, FeS. This 
 iron sulphide is more brittle than, manganese sulphide, and in- 
 stead of coalescing in drops, it spreads out in webs or sheets. It 
 is therefore .very weakening to the steel, because the area of 
 weakness is more extensive than the tiny spots of manganese 
 sulphide. Steel containing iron sulphide is liable to show poor 
 tensile test and low ductihty. It is at the rolling temperature, 
 however, that iron sulphide produces the greatest weakness, 
 because at this point it is in a liquid form and therefore has 
 practically no adhesion to the crystals of steel, which are liable 
 to break along the planes or meshes of sulphide. The same is 
 true, to a less extent, of the effect of manganese sulphide, which 
 is also in a liquid or pasty condition at the rolling temperature; 
 but the extent of its damage is not so great on account of its 
 drawing together in drops. ^ These facts explain the well-known 
 beneficial effect of manganese in counteracting the damage due to 
 sulphur in iron and steel. 
 
 Iron Phosphide. — Iron forms at least one phosphide, having 
 the formula FegP,- and this phosphide forms with iron a series of 
 alloys, of which the eutectic contains 64 per cent, of FegP (10.2 
 per cent, of phosphorus). Even 1 per cent, of phosphorus will 
 make the melting-point of iron very much lower, and it is for this 
 
 1 See No. 113. 
 20 
 
306 THE METALLURGY OF IRON AND STEEL 
 
 reason that foundry irons are often desired with a high content of 
 phosphorus. Even where there is a smaller amount of phosphorus 
 there will be some of the phosphorus eutectic formed, and this 
 remains in a molten condition for some time after the bulk of the 
 steel has solidified. This liquid eutectic- tends to migrate to the 
 spaces between the crystals, where it remains after solidification 
 and forms a very brittle network, which naturally makes the 
 whole mass more or less brittle. For these reasons phosphorus 
 is the greatest source of brittleness in steel, and especially brittle- 
 ness under shock, and is thus a great enemy to the engineer and 
 other users of the material. 
 
 Besides forming the eutectic, as I have described, phosphorus 
 tends to produce coarse crystallization in steel, and this makes 
 it both weak and brittle. It is a fact observed many times that 
 the embrittling effect of phosphorus on steel is much less when 
 the steel is very low in carbon, and as the carbon rises the brittle- 
 ness caused by phosphorus rises. This and other effects of 
 phosphorus have been explained by Prof. J. E. Stead in two 
 very able papers.* He has shown that a little phosphorus will 
 dissolve in ferrite and that then the eutectic which produces 
 the brittleness will not form, but as the carbon in the steel in- 
 creases, it precipitates the phosphorus from the ferrite solid 
 solution and therefore causes the eutectic to form. Hence, the 
 more ferrite and the less cementite in steel, the less will be the 
 brittleness produced by phosphorus. 
 
 Iron suicides. — There seem to be three or more silicides of 
 iron, but the one having the formula FeSi seems to be found 
 most commonly in steel. It appears to increase very slightly 
 the strength of steel, and also, to a limited extent, its hardness. 
 The chief importance of silicon, however, as already pointed out, 
 is in promoting soundness. 
 
 Iron Oxides. — Oxygen occurs in steel in at least three forms: 
 CO, FeO and FejOg. In either form its presence is very harmful, 
 producing brittleness in both hot and cold steel, besides causing 
 the liability to blow-holes already discussed. There is proba-bly 
 no constituent more harmful to steel than oxygen, and unfortu- 
 nately the chemists are far behind in respect of not yet having 
 found a satisfactory method of accurately determining small traces 
 of this gas. The effect of oxygen is somewhat similar to that of 
 sulphur, and, in common parlance, makes the steel "rotten," 
 
 ^ See No. 184 and No. 115. 
 
THE CONSTITUTION OF ST;i:Ji:L 307 
 
 Nitrogen and Hydrogen. — Both nitrogen and hydrogen occur 
 in steel, and one of the theories to explain the superiority of 
 crucible steel is based upon the relative freedom of this material 
 from these two gases. The amount of nitrogen and hydrogen 
 present is usually very small. Hydrogen dissolves very easily 
 in iron at a high temperature, but is evolved in part as the metal 
 cools. In order to obtain entire freedom, however, it is necessary 
 to heat and cool several times in vacuo. 
 
 The Strength of Steel 
 
 The properties of steel most commonly desired are strength 
 and ductility. Unfortunately there is more or less incompati- 
 bility between these two. That is to say, as the strength of steel 
 increases, the ductility usually decreases; and, conversely, as the 
 ductility increases, the strength usually decreases. There are 
 other properties of steel which are likewise of importance, either 
 because they are desired, or the reverse. Among these we shall 
 especially discuss hardness, brittleness, electric conductivity, 
 magnetic permeability, magnetic hysteresis, permanent magnet- 
 ism, and weldability. 
 
 Pure iron has a tensile strength of about 45,000 lb. per sq. 
 in. and a compressive strength of about 80,000 lb. per sq. in. 
 These are increased by several of the ordinary impurities found 
 in steel, but the most important strengthener is carbon, because 
 this will increase its strength with the least decrease in ductility. 
 
 Carbon. — Each increase in carbon (cementite) gives an 
 increase in tensile strength until we reach a maximum of about 
 0.9 to 1 per cejit. of carbon (13.5 to 15 per cent, cementite). 
 With further increase in cementite there is a decrease in tensile 
 strength. It is probable that the reason for this maximum 
 tensile strength at approximately the eutectoid ratio of the steel 
 is due largely to the small crystallization and the intimate mix- 
 ture of the crystalline constituents when the steel is at, or near, 
 the eutectoid proportions.^ With less cementite the grains of 
 pearlite are surrounded by a network of ferrite; with more ce- 
 mentite, the grains of pearlite are surrounded by network of 
 cementite; and both of these networks have a weakening 
 effect upon the material by decreasing the adhesion of the 
 
 * The still greater increase in strength when the two constituents of the pearlite (ferrite 
 and cementite) are even more intimately associated is explained on page 377. 
 
308 
 
 THE METALLURGY OF IRON AND STEEL 
 
 crystals. The relation between carbon and tensile strength is 
 shown graphically in Fig. 267. 
 
 H. H. Campbell and W. R. Webster have studied exhaustively 
 the effect of the different impurities upon the tensile strength of 
 steel. The latest word on this subject has been said by Campbell 
 who gives^ the following rule for the effect of each 0.01 per cent, 
 of carbon on acid and on basic open-hearth steel. Starting with 
 40,000 lb. per square inch for pure steel, each 0.01 per cent, of 
 
 170,000 
 
 140,000 
 
 110,000 
 
 80,000 
 
 50,000 
 
 Percentage of Carbon 
 0.10 0.20 0.80 0.40 0.50 0.G0 0.70 0.80 0.90 ^■**'' 1.10 1.20 1.30 1.40 1.50 1.G0 1.70 1.30 1.90^'"™ 
 
 
 RELATION BETWEEN THE PHYSICAL 
 
 CHARACTERISTICS OF STEEL AND THE 
 
 PROPORTIONS OF FERRITE AND FeaC 
 
 
 
 ^"^x 
 
 
 
 ""^ 
 
 
 ^-^ 
 
 --• 
 
 
 \/ 
 
 ^^^5J 
 
 
 
 
 s. 
 
 -A 
 
 \ 
 \ 
 
 
 
 
 
 
 / 
 
 
 \. 
 
 "^^Uy^l 
 
 L4CtlIity 
 
 
 
 loose 
 
 80^ 
 
 mi 
 
 m 
 
 2H 
 
 SEFeaCO 5 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 ^FesC 
 
 StFerrlte 100 95 
 
 90 
 
 85 
 
 80 
 
 75 
 
 TOSfiFerclte 
 
 FIG. 267.^From Howe, "Iron, Steel, and other Alloys." 
 
 carbon will increase the strength by 1000 lb. per square inch in 
 the case of acid open-hearth steel, and by 770 lb. per square 
 inch in the case of basic open-hearth steel. Because the color 
 method of determining the amount of carbon does not show all 
 the carbon present, the figures given above must be replaced by 
 1140 lb. and 820 lb. for each 0.01 per cent, of carbon, when the 
 color test is used. We have no data to determine the strengthen- 
 ing effect of carbon in Bessemer and crucible steel, but it is 
 probable that a little lower value than that given for basic open- 
 hearth steel would be used for Bessemer, and a little higher 
 
 1 Page 391 of No. 2. 
 
THE CONSTITUTION OF STEEL 309 
 
 value than that given for acid open-hearth steel would be used 
 for crucible steel. 
 
 Silicon. — The effect of silicon on strength is probably very 
 small in the case of rolled steel. In the case of castings, how- 
 ever, an important increase in tensile strength may be obtained 
 by increasing the silicon to 0.3 or 0.4 per cent. This practice 
 results in practically no decrease in ductility, but it is necessary 
 to supply larger risers on account of the deep piping that will be 
 produced. It is probable that the beneficial effect of silicon in 
 this case is due very largely to its producing soundness. 
 
 Sulphur. — H. H. Campbell says that the effect of sulphur on 
 the strength of acid and basic open-hearth steel is very small. 
 It is probable, however, that this statement is only true when 
 the sulphur is in the form of manganese sulphide, because the 
 effect of iron sulphide would be to lower the strength and the 
 ductility of the material. The worst effect of sulphur is undoubt- 
 edly its production of ^'red-shortness" and the liability to cause 
 checking during rolling, or, in the case of a casting, during 
 cooling. 
 
 Phosphorus. — Campbell states that each 0.01 per cent, of 
 phosphorus increases the strength of the steel by 1000 lb. per 
 square inch. It should be observed, however, that this increase 
 of strength is measured by the resistance of the material to 
 stresses slowly applied and that it ceases with 0.12 per cent, 
 phosphorus and is reversed. In the case of vibratory stresses and 
 sudden shocks, phosphorus is probably the most harmful of the 
 elements, so that it is undesirable to increase the strength of steel 
 by means of this element. This is the more true because of the 
 brittleness produced by phosphorus, for an increase of strength 
 obtained through this medium is accompanied by a much greater 
 decrease in ductility than when the same increase in strength is 
 obtained through the medium of carbon. 
 
 Manganese. — The beneficial effect of manganese on tensile 
 strength begins, according to the same authority, only when the 
 manganese is above 0.3 or 0.4 per cent. With less manganese 
 than this, "in the case of open-hearth and Bessemer steels, the 
 presence of some other condition, probably iron oxide, masks the 
 effect of man'ganese. It will be remembered that when 
 the manganese is low, open-hearth and Bessemer steels are 
 harmfully charged with oxygen. Furthermore, the effect of 
 manganese is dependent upon the amount of carbon present. 
 
310 
 
 THE METALLURGY OF IRON AND STEEL 
 
 In acid open-hearth steel each 0.01 per cent, of manganese 
 (beginning at 0.4 per cent.) increases the strength 80 lb. per 
 square inch when the carbon is 0.1 per cent., but each increase of 
 0.01 per cent, of carbon increases the strengthening effect of 
 manganese by 8 lb. So that, for example, if we have an acid 
 open-hearth steel containing 0.4 per cent, of carbon, then each 
 0.01 per cent, of manganese will increase the strength by 320 lb. 
 per square inch. In the case of basic steel each 0.01 per cent, of 
 manganese (beginning at 0.3 per cent.) increases the strength 
 130 lb. per square inch when the carbon is 0.1 per cent., and so 
 on to 250 lb. per square inch when the carbon is 0.4 per cent., for 
 each additional 0.01 per cent, of carbon increases the strengthening 
 effect of manganese on basic steel by 4 lb. (See Table XXVI.) 
 FormulcB. — Campbell gives the following formulae for the 
 strength of acid and basic open-hearth steels: 
 
 For acid steel: 40,000 + 10000+1000P+xMn + R = ultimate strength. 
 For basic steel: 41,500 + 7700 + 1000P+yMn + R = ultimate strength. 
 
 In these formulae, C = 0.01 per cent, carbon (determined by com- 
 bustion), P = 0.01 per cent, phosphorus, Mn = 0.01 per cent, 
 manganese, and R = a variable, depending upon the heat treat- 
 ment which the steel has received. For X and Y see Table XXVI. 
 
 TABLE XXVI.- 
 
 -EFFECT OF EACH 0.01 PER CENT. MANGANESE 
 ON OPEN-HEARTH STEEL 
 
 Percentage of Carbon 
 
 On Acid Steel 
 
 X 
 
 On Basic Steel 
 Y 
 
 Lb. per sq. in. 
 
 Lb. per sq. in. 
 
 0-05 
 
 80^ 
 120 
 160 
 200 
 240 
 280 
 320 
 360 
 400 
 440 
 480 
 
 1102 
 130 
 
 0.10 
 
 0-15 
 
 0.20 
 
 170 
 190 
 210 
 230 
 250 
 
 0-25 
 
 0.30 
 
 0-35 
 
 0.40 
 
 0-45 
 
 0.50 
 
 
 
 0.55 
 
 
 0.60 
 
 
 
 
 1 Begininng only with . 4 per cent, of manganese. 
 '• Beginning only with . 3 per cent, of manganese. 
 
THE CONSTITUTION OF STEEL 311 
 
 Copper. — Copper, up to at least 1 per. cent., does not appear 
 to have any important effect upon the strength or ductility of 
 low- and medium-carbon steels, but increases the brittleness of 
 steel containing 1 per cent, carbon. When the sulphur is more 
 than 0.05 per cent, copper appears to make the steel roll less 
 easily and above 0.5 per cent, of copper appears to make high- 
 carbon steel draw less easily into wire. 
 
 Arsenic, — Some steels contain arsenic, which does not appear 
 to have any effect when it is below 0.17 per cent., but any larger 
 quantity raises the tensile strength and decreases the ductility 
 to a very important extent. 
 
 Oxide of Iron. — All Bessemer and open-hearth steels contain 
 more or less oxide of iron, there probably being more in basic steel 
 than in acid, and more in Bessemer steel than in basic open- 
 hearth steel. It is probable that this oxide of iron has not any 
 very marked effect upon strength, as Campbell quotes some steels 
 containing quantities larger than usual whose strength is good. 
 No data are given as to the ductility, however, and it is probable 
 that oxide of iron has a deleterious effect upon this quality. 
 
 Hardness and Brittleness op Steel 
 
 As a general thing the hardness and brittleness of steel increase 
 together. The chief commercial application of this property is 
 in such articles as railroad rails and car-wheels, bevel- and spur- 
 gears, axles and bearings, the wearing parts of crushing machin- 
 ery, etc.^ To produce hardness in these materials carbon is the 
 chief agent used, because it gives the maximum hardness with 
 the least brittleness. It is for this reason that railroad rails are 
 now made up to 0.7 per cent, carbon, and although this material 
 is somewhat brittle and breakages occasionally occur, the high 
 carbon is demanded in order that the head of the rail may be 
 durable. Phosphorus increases both the hardness and the 
 brittleness, especially under shock. Manganese likewise increases 
 hardness, and especially the kind of hardness which makes it 
 more difficult and more expensive to machine the metal. With 
 more than about 1 per cent, of manganese the steel becomes 
 somewhat brittle. When the content rises above 2 per cent, the 
 steel is so brittle as to be practically worthless. In this connec- 
 
 * I purposely omit here the consideration of such hardness as that produced in springs, 
 cutting tools, etc., by heating the steel to a bright-red heat and plunging into water, as 
 this wiU be discussed in Chapter XIV. 
 
312 THE METALLURGY OF IRON AND STEEL 
 
 tion a curious phenomenon is observed, in that a still further 
 increase in manganese produces a reversal of its influence, and 
 when we have more than 7 per cent, the metal is not only 
 extremely hard and practically impossible to machine commer- 
 cially, but becomes, after heat treatment, very ductile and tough. 
 This will be considered more fully in Chapter XV. 
 
 Silicon. — Silicon makes the steel slightly harder, but appar- 
 ently without increasing its brittleness, unless we have more 
 than 0.5 or 0.6 per cent. 
 
 Electric Conductivity of Steel 
 
 The purer the material and the nearer it is to f errite, the better 
 will be its electric conductivity; therefore only wrought iron, or 
 the softest and purest forms of steel, are used generally for wire 
 for electric conduits.^ The case is somewhat complicated when 
 we come to third rails for electric railroads, which have now 
 become a very important industrial product, because so pure a 
 material will be very soft and will rapidly wear away under the 
 abrasion of the contact-shoes. To increase the hardness of these 
 rails with the least' decrease in electric conductivity, it is best to 
 avoid nickel and manganese, which decrease electric conductivity 
 in greater proportion than the other elements, and to obtain the 
 hardness as much as possible from phosphorus, and not from 
 carbon, because phosphorus will give the greatest amount of 
 hardness with the least decrease in the purity of the iron. As 
 high phosphorus steels are difficult to roll, however, the section 
 of the rail chosen should be as free as possible from sharp corners 
 and thin flanges, in order that the tearing action in rolling may 
 be as slight as possible. 
 
 Magnetic Properties of Steel 
 
 Alpha ferrite is the chief magnetic constituent of iron and steel, 
 and therefore the greater the amount of this constituent present, 
 the greater will be the magnetic force and magnetic permeability 
 and the less its magnetic hysteresis. On this account the cores 
 of electromagnets, the armatures of dynamos, etc., are commonly 
 made of Swedish wrought iron, which is the purest commercial 
 form of iron made.^ 
 
 1 Omitting, of course, the use of copper. 
 
 ^ A silicon alloy of iron with a double-heat treatment discovered by R. A. Hadfield, and 
 having a very high magnetic force and permeability, will be discussed under the head of 
 "AUoy Steels." 
 
THE CONSTITUTION OF STEEL 
 
 313 
 
 Ferrite- has no permanent magnetism, but immediately loses 
 its magnetic force when it is out of contact with a magnet, or, in 
 the case of cores of electric magnets, when the electric current is 
 cut off. Permanent magnets are therefore made of a high-carbon 
 steel (1 per cent.), whose strength and permanency are increased 
 if about 5 per cent, of tungsten is present. This steel is heated 
 above the critical temperature and hardened in water, after which 
 it is magnetized by causing an electric current to flow around it 
 for a short time. Steel so treated will retain the magnetic force 
 for many years. Osmond has explained thepermanent magnetism 
 of steel in the following very ingenious manner: Each molecule 
 of alpha ferrite is believed to have a north and a south magnetic 
 
 
 FIG 268. 
 
 ^ O C»^C^<^C»<»C*c^l=P C» (=» ^» ^» 
 £* d* £9 ^ i£V c:» c* c9 e9 (9 £» <B ^ ^ cm 
 
 €* ^ c ^ c» cv c c9 <^ i:^ €M <^ ^ <9 ^ o 
 
 C» O O c» a»C»cvC»c>^ <:% ^ 1:3 Cte o ^> 
 
 FIG. 269. 
 
 pole, which, in the ordinary unmagnetized condition of the iron, 
 will be oriented in many different directions, as shown in Fig. 268. 
 When this piece of iron in placed in the magnetic field, however, 
 the molecules arrange themselves in accordance with the lines of 
 magnetic force, with their north poles all in one direction and 
 their south poles all in the opposite direction, thus making the 
 whole piece of iron a magnet. As soon as the magnetic force is 
 removed, however, the molecules all return to their original 
 orientation, and the whole piece of iron loses its magnetism. We 
 have already learned that it is only the alpha molecules which 
 have north and south magnetic poles, and if the steel consists 
 entirely of beta or gamma molecules, it is not capable of becom- 
 
314 THE METALLURGY OF IRON AND STEEL 
 
 ing magnetic. According to Osmond's theory, when steel is 
 cooled rapidly from above the critical temperature, the shortness 
 of the time taken in reaching the atmospheric temperature is such 
 that only a part of the molecules are able to change to the alpha 
 allotropic form, and the remainder are retained in the gamma 
 form. This retention is assisted by the 1 per cent, of carbon 
 present, which acts as a brake to make the change slower. When, 
 now, this hardened steel is subjected to the magnetic force, the 
 alpha molecules orient themselves with their north poles all in one 
 direction; but when the magnetic force is removed there is a 
 certain number of gamma molecules present to interfere with the 
 free movement of the alpha molecules and prevent them from 
 returning to the original unoriented position. This explanation 
 implies that the magnetic force is sufficient to overcome the 
 resistance of the gamma molecules and force the alpha molecules 
 into a magnetic position, but that the force of the alpha mole- 
 cules tending to return to their original position is not so great. 
 The welding of steel and the effect of different elements upon 
 it will be discussed in Chapter XIV, 
 
XII 
 THE CONSTITUTION OF CAST IRON 
 
 Practically all the cast iron which is not purified is used for 
 making iron castings, so that a study of the constitution of cast 
 iron resolves itself into a study of iron castings. The difference 
 between cast iron and steel is that the former contains less iron 
 and more impurities, especially carbon, silicon, phosphorus, and 
 occasionally sulphur and manganese. The advantages of cast 
 iron, and the reason it is used as much as it is, are its fluidity, 
 lesser amount of shrinkage when cooling from the molten state, 
 relative freedom from checking in cooling, and the ease with which 
 very different properties are conferred upon it at will. Its dis- 
 advantages are its weakness and lack of ductility and malleability. 
 The last-named deficiency renders it practically impossible to 
 put any work upon cast iron; hence it can never be wrought to 
 shape and must always be used in the form of castings. Its most 
 important advantages are probably its ready fusibility, which 
 makes it so easy to melt and cast, and its cheapness. 
 
 Graphite. — All of the characteristic qualities of cast iron are 
 due to the presence of the large amount of impurities in it. These 
 impurities are the same in kind as the impurities in steel, and 
 differ only in amount, with the single exception of graphite. 
 This constituent is almost never found in steel, or is found in a 
 very small number of cases, those cases being confined to the 
 high-carbon steels, the amount of which is very small in com- 
 parison. In cast iron, however, graphite is the largest and one 
 of the most important constituents. It occurs in thin flakes, in 
 sizes varying from microscopic proportions to an eighth of a 
 square inch in area,^ disseminated through the body of the metal 
 and forming an intimate mechanical mixture, a magnified 
 section of which is shown in Fig. 279. Each flake of graphite 
 is composed of smaller flakes, buUt up somewhat like the sheets of 
 mica with which all are familiar, but with very little adhesion 
 
 ^ In rare and unusual Cases the flakes of graphite may be as much as an inch and a half 
 to two square inches in area, but practically never so large in the commercial cast irons, 
 
 315 
 
316 THE METALLURGY OF IRON AND STEEL 
 
 between the small component flakes, so that the sheet of graphite 
 may be split apart with very little force. Graphite is very 
 light in weight, having a specific gravity of only about 2.25 as 
 compared with a specific gravity of 7.86 for pure iron; conse- 
 quently although the percentage of graphite by weight is only 
 4 per cent, or less of the iron, its percentage by volume may be, 
 in normal cases, as much as 14 per cent. This may readily be 
 seen by noting the amount of space occupied by the graphite 
 flakes in Fig. 279. 
 
 Combined Carbon.— y^e have already discussed the solidifica- 
 tion and slow cooling of cast iron, and it will be remembered that 
 all the carbon which does not precipitate as graphite forms first 
 as austenite, which later decomposes into ferrite and cementite. 
 In short, all the carbon in slowly-cooled cast iron will ultimately 
 be found partly in the form of graphite and partly in the form 
 of cementite. The carbon of cementite in cast iron commonly 
 goes under the name of ''combined carbon," but it must be 
 remembered that cementite is the constituent which gives the 
 observed effects. Unless the cooling is very slow, some carbon 
 will be in the austenitic or dissolved form. 
 
 White Cast Iron. — In slowly-cooled white cast iron the carbon, 
 amounting often to 3 or 4 per cent., will all be in the form of ce- 
 mentite, which will therefore form 45 to 60 per cent, of the mate- 
 rial; consequently white cast iron will possess largely the properties 
 of cementite. It is very hard and brittle, being machined only 
 with the greatest difficulty and with special kinds of cutting 
 tools, and resisting wear by abrasion very effectively. It is so 
 brittle as to be readily broken by the blows of a hammer, and is 
 weak because of the presence of very large plates of cementite, 
 which adhere but slightly to one another. Consequently white 
 cast iron has few uses and is employed usually only as a hard 
 surface on the outside of gray-iron castings. 
 
 Gray Cast Iron. — Gray cast iron may have about the same 
 total amount of impurities present as white cast iron, the only 
 difference then being that the carbon is partly or wholly pre- 
 cipitated as graphite. The gray ' color of a freshly broken 
 fracture, from which this material receives its name, is due al- 
 together to the graphite present, for this constituent is so weak 
 that the iron breaks chiefly through its crystals, which are rent 
 asunder, leaving one part sticking to each side of the fracture. 
 The weakness of gray cast iron as compared with steel is thus 
 
THE CONSTITUTION OF CAST IRON 317 
 
 readily understood, since there is but a small proportion of 
 metallic surface to be broken and the graphite splits so easily. 
 An interesting experiment is to take a freshly broken surface of 
 gray pig iron and brush one-half of it for some time with a stiff 
 brush. In this way the adhering crystals of graphite are par- 
 tially removed and we get a surface which is almost as white as 
 the fracture of white cast iron. This shows clearly that the gray 
 color is due altogether to the graphite and that the metallic part 
 is as silvery white as iron itself. The prevalence of the gray 
 color also shows how completely fracture takes place through 
 the graphite crystals. 
 
 Gray cast-iron castings are by far the more important, and 
 the study of their constitution is the chief object of this chapter. 
 These castings usually contain 2 per cent, or more of graphite 
 and less than 11/2 per cent, of combined carbon. It will be ob- 
 served that this limit of combined carbon is also the range found 
 in steel. Furthermore, it will be observe^ that the graphite is 
 not a chemical component of the metallic body, but is mechanic- 
 ally mingled with it. In this sense, therefore, we may consider 
 gray cast iron as a very impure steel, ^ mechanically mixed with 
 graphite, and upon this reasoning the study of its constitution 
 becomes much simpler, ^ as we may study first the properties of 
 the metallic part, and next that of the graphite, and so be able 
 to foretell to some extent the properties of the mixture. Indeed, 
 the properties of the metallic part are already understood pretty 
 well from our knowledge of the constitution of steel, and there 
 is no new constituent or new condition except the larger amounts 
 of silicon and phosphorus, which are of minor importance, be- 
 cause their effect is collectively far less than the weakening and 
 embrittling effect of the graphite. Even though we had a very 
 pure metallic constituent, the strength and ductility of this por- 
 tion would not be siifficient to prevent the mass as a whole break- 
 ing at a small load and without exhibiting any practical ductility,' 
 because of the weakening effect of the crystals of graphite. In 
 other wordS; it is the carbon which is the great factor in determin- 
 
 * The silicon in gray cast iron, is i^ually between 0.75 per cent, and 3 per cent., or, let 
 U3 say, ten times that in steel, while the phosphorus is usually from 0.5 to 1.5 per cent. , 
 or, again, about ten or more times that in steel. The sulphur varies greatly, but is not 
 infrequently as high as 0.15 to 0.2 per cent. Manganese is an exception and is usually 
 no higher in cast-iron castings than in steel. 
 
 ^ This theory of the constitution, which meets with very favorable acceptance in many 
 quarters, was independently evolved by J. E. Johnson, (American Machinist, 1900), 
 and H. M. Howe (Trans. A. 7. M. E., 1901, vol. xxxi, pp. 318-339). 
 
318 THE METALLURGY OF IRON AND STEEL 
 
 ing the properties of cast iron, for this may be either all graphitic, 
 or all combined, or part in both conditions. 
 
 Total Carbon, — By running the blast furnace very hot, we 
 may extend the saturation point of the iron for carbon and thus 
 get a slightly higher total carbon. This is not a very potent 
 influence, however, for we seldom have total carbon more 
 than 4.5 per cent., or less than 3.25 per cent. This control, such 
 as it is, may be exercised either during the manufacture of the 
 pig iron or during the remelting in the cupola, because in the 
 latter furnace the liquid iron is in contact with coke and 
 will absorb carbon up to its saturation point at the existing 
 temperature. 
 
 Rate of Cooling. — A far more potent effect on the properties of 
 the metal is the transfer of carbon from the graphitic into the 
 combined form, or vice versa, by rapid or by slow cooling from 
 the molten condition. It will be remembered that the carbon is 
 always dissolved in the iron when the mass is in a molten condi- 
 tion, that is, when it is above the lines AB and BC in Fig. 253. 
 As we cool from the molten state, graphite precipitates, but 
 this cooling must be very slow indeed for this normal chemical 
 change to take place completely. If, therefore, we cool with 
 great rapidity, as, for example, by pouring the iron into a 
 metallic mold which "chills" it, or by some other form of 
 artificial rapid cooling, we may prevent the precipitation of 
 graphite, by denying the time necessary for the chemical re- 
 action, and obtain a metal in which all the carbon is in the com- 
 bined form,^ i.e., white cast iron. It is also evident that, by a 
 rate of cooling intermediate between this rapid rate and the slow 
 rate which permits the precipitation of the normal amount of 
 graphite, we may obtain an intermediate amount of carbon in 
 the graphitic form. Rapid cooling is a very important means 
 of "chilling" the surfaces of gray-iron castings, whereby we may 
 have a relatively soft gray iron in the interior of each article and a 
 hard surface. For example, chilled-iron rolls are made in this way 
 (see page 187), and also American railroad car wheels^ which are 
 cast against an iron chill (see page 336), giving nearly an inch 
 depth of white iron around the tread and flange where the metal 
 
 1 The term combined here is used to include chemical combination or solution. 
 
 ^ In other important railroad countries it is more usual to have the car wheels made of 
 steel, as it is believed that the iron wheels are not sufficiently strong and ductile. The 
 manufacture of pressed-steel car wheels is Increasing in America; nevertheless, the chilled 
 cast-iron wheels seem to give very good service. 
 
THE CONSTITUTION OF CAST IRON 319 
 
 is to suffer abrasion in grinding over the rails, while the web and 
 bore will be of gray cast iron, because cooled more slowly in the 
 sand part of the mold, and thus will be less brittle, and better 
 able to withstand the shocks of service and may be machined 
 easily. 
 
 The Effect of Carbon on Cast Iron 
 
 The nature and constitution of gray cast iron is far more 
 difficult to understand than that of steel, and even greater is the 
 difficulty of predicting the effect of any change in composition or 
 in constituents. The chief reason for this complexity is that a 
 change in any one of the constituents of gray cast iron is liable 
 to effect changes in several others as well. The simplest example 
 of this is in the case of the carbon; we have total carbon, graphite, 
 and combined carbon, and if we change any one of these three, 
 we must change either one or both of the other two, and it makes 
 a great deal of difference which. Indeed, we almost never change 
 the amount of graphite without making the reverse change in the 
 amount of combined carbon, and vice versa. Thus a very loose 
 system of speaking of these matters has come into vogue among 
 foundrymen. For instance, it is very common to hear a foundry- 
 man say: "In order to soften your iron, increase the graphite"; 
 but what he really means is: "In order to soften your iron, 
 decrease the combined carbon." He knows that the one change 
 usually follows from the other, and he speaks of it in this way, 
 regardless of the fact that graphite can be increased (i.e., by 
 increasing the total carbon and leaving the combined carbon the 
 same or a little higher), and yet the iron will not be made any 
 softer, but may even be harder. 
 
 Graphite and ■ Shrinkage, — :The most important effect of 
 graphite on cast iron, aside from causing weakness, is in decreas- 
 ing the shrinkage. The reason for this will be understood when 
 we consider what happens when cast iron solidifies. It will be 
 remembered that when the eutectic forms, the cast iron breaks 
 up into alternate plates of graphite and austenite. This separa- 
 tion of graphite from solution is the birth of a new constituent, 
 and this constituent occupies space, so that there is an expansion 
 of the mass as a whole in proportion to the amount of graphite 
 that separates. , If, therefore, we pour liquid cast iron into a 
 mold, which is, of course, entirely filled at the moment when the 
 iron begins to solidify, the first action that takes place after the 
 
320 
 
 THE METALLURGY OF IRON AND STEEL 
 
 beginning of solidification is an expansion, due to the separation 
 ot graphite. The expansion continues for several moments until 
 the chemical precipitation is completed, after which the metal 
 begins to contract, as all metals do in cooling from a high temper- 
 ature; but the preliminary expansion has been so great that the 
 ultimate shrinkage may be only about one-half what it otherwise 
 
 2 
 
 I 
 
 "8 i 
 
 r 
 
 10 
 
 12 
 
 U 
 
 ?§■ 
 
 :?;: 
 
 ii 
 
 mm 
 
 & 
 
 B? 
 
 ^ 
 
 So 
 
 1^^ 
 
 FIG. 275.— SHRINKAGE CURVES. 
 
 would have been. We can thus control this shrinkage by con- 
 trolling the amount of the expansioji, through varying the 
 graphite. This point will be more readily understood by refer- 
 ring to Fig. 275, which is taken from an article by Prof. Thomas 
 Turner of England.^ 
 
 Explanation of Fig. 275. — The point marks the position 
 occupied by the end of the bars at the moment of solidification. 
 It will be seen that in the case of copper the metal contracts con- 
 
 ^ Journal of the Iron arid Steel Institute, No. 1, 1906, page 57. 
 
THE CONSTITUTION OF CAST IRON 321 
 
 tinuously from this point, as shown by the continuous drop of the 
 curve. In the case of white cast iron, the metal contracts con- 
 tinuously until we reach a certain point (which is at a temperature 
 of about 665*^ C), when a momentary arrest of the shrinkage 
 takes place, after which the metal again contracts. This arrest 
 is common to all cast iron and steel and marks the decomposition 
 of austenite at the point Sj Fig. 252, page 295. Now, see what a 
 difference there is in the case of gray cast iron, which does not 
 shrink immediately after freezing, but expands very appreciably, 
 as shown by the rise in the curve. This expansion is due to the 
 graphite that is being expelled from the metal and occupies space 
 between the particles of iron. 
 
 Again, in the case of the Northampton iron, which is high in 
 both graphite and phosphorus, the expansion is very long- 
 continued, and the metal cools to almost a black heat before the 
 bar has shrunk again to the size it had when first cast. This 
 expansion is again due to the separation of carbon, and is 
 assisted apparently by the phosphorus keeping the iron in a 
 semifluid condition for a long time and thus allowing the graphite 
 more easily to separate and make place for itselt. Here, too, we 
 have an explanation why phosphoriferous irons fill every crevice 
 of the molds so perfectly. Being in a pasty condition for some 
 time, and continually expanding, the semifluid metal is forced 
 into the tiniest crevices of the molds, filling all the corners with 
 astonishing sharpness. 
 
 It is evident that any increase in graphite, whether caused by 
 an increase in total carbon or by a decrease in combined carbon, 
 will produce less shrinkage. The extent of this may be judged by 
 noting that gray cast iron expands so much in solidifying that no 
 contraction cavity or pipe is formed, such as occurs in the case of 
 steel. If the iron is only slightly gray, or if it is a very large 
 section of metal, then a slight spongy place may be formed in the 
 center, which is the nearest approach to a shrinkage cavity that 
 is normally found in most iron castings. 
 
 Graphite and Porosity.— It is also evident that an increase 
 in graphite, whether produced by an increase in total carbon or 
 by a decrease in combined carbon, will increase the porosity of 
 the casting, which is often disadvantageous, as in the case of 
 hydraulic cylinders or other receptacles for holding liquids under 
 pressure. The separation of much graphite usually is accom- 
 panied by large-sized graphite crystals, and therefore the crystals 
 
 21 
 
322 THE METALLURGY OF IRON AND STEEL 
 
 of the mass, as revealed by the fracture, appear large and the 
 grain is said to be "open." 
 
 Graphite and Workability, — When we come to consider the 
 effect of graphite upon the softness or workability of the cast 
 iron, it is evident that we must consider it in relation to other 
 things; for if we increase the graphite by increasing the total 
 . carbon, then we increase the workability of the metal only in so 
 far as the graphite acts as a lubricant for the tool that is doing 
 the cutting. Evidently the tool will have no difficulty in cut- 
 ting through the soft flakes of graphite; the chief resistance to it 
 will be given by the metallic part of the mixture. Though the 
 lubricating effect of the graphite undoubtedly helps the tool, its 
 presence evidently cannot increase the actual softness of the 
 metallic body with which it is mixed. But, on the other hand, 
 if we increase the graphite by decreasing the combined carbon,' 
 then we have not only increased the amount of lubricant, but 
 we have, in addition, increased the softness of the metallic part 
 of the mixture by reducing the proportion of cementite in it. 
 
 Graphite and Strength, — Everything else being equal, it is 
 evident that the more graphite we have in cast iron, the weaker 
 it will be, for we have already shown that gray cast iron breaks 
 by the ready splitting apart of the flakes of graphite. Thus, if 
 we make no change in the combined carbon, but increase the total 
 carbon of our castings, and consequently the graphite, we should 
 expect a corresponding decrease in strength, and this is in fact 
 found to occur. When, however, we increase the graphite and at 
 the same time decrease the combined carbon, we may or may not 
 get an increase in strength, this depending altogether on how 
 much combined carbon there was before and after the change. 
 For example, if we had 3 per cent, of combined carbon iind 1 per 
 cent, of graphite in a casting, that casting would be weak be- 
 cause of too high combined carbon. To decrease this and in- 
 crease the gi'aphite would have a beneficial effect on strength. 
 On the other hand, if we had 1 per cent, of combined carbon 
 and 3 per cent, of graphite, to decrease the combined carbon 
 and increase the graphite would have a detrimental effect on 
 strength. We must therefore consider the question of strength 
 from a much larger viewpoint than by considering any one con- 
 stituent alone. 
 
 Combined Carbon and Shrinkage. — Combined carbon has very 
 little effect on the shrinkage of cast iron except in so far as it 
 
THE CONSTITUTION OF CAST IRON 323 
 
 changes the graphite. That is to say, if by increasing the com- 
 bined carbon we decrease the graphite, we will get an increase in 
 shrinkage, and vice versa. 
 
 The Effect of Silicon, Sulphur, Phosphorus, and Man- 
 ganese ON Pig Iron 
 
 The constitution of cast iron is, furthermore, very complicated 
 because of the double influence of silicon, sulphur, phosphorus, 
 and manganese. Each of these elements has a direct influence 
 upon the properties of the material, which is in general similar to 
 its influence upon steel. For example, silicon produces freedom 
 from oxides and blow-holes and makes the iron more fluid; man- 
 ganese counteracts the effect of sulphur and increases the diffi- 
 culty of machining the material; sulphur makes the metal very 
 tender at a red heat, and therefore liable to checking if put under 
 strain during this period. For example, if a casting in shrinking 
 tends to crush the sand, this strain will be more liable to break it 
 in case the^ sulphur is high. Sulphur also makes solidification 
 take place more rapidly, and causes blow-holes and dirty iron. 
 Phosphorus makes the metal very fluid and reduces its melting- 
 point. It also makes it more brittle under shock, especially 
 when cold, and produces a fusible eutectic, a photomicrograph 
 of which is shown in Fig. 266. Phosphorus and sulphur increase 
 the tendency to segregate. 
 
 Furthermore, the various compounds of the impurities with 
 iron and with each other, which we find in steel, are also found in 
 cast iron. Indeed, some of them are far more important in the 
 latter than in the former, because the amount of the impurities is 
 greater. This is especially true of manganese sulphide and iron 
 sulphide, for the sulphur in cast iron is often large in amount, 
 while the manganese is often intentionally small on account of the 
 difficulty which this element produces in the machining of the 
 casting. Therefore we are even more liable to find iron sulphide 
 itf cast Tron than in steel. 
 
 But the direct effect of these impurities is usually far less 
 important than their indirect effect, namely, their influence upon 
 the carbon. After all, it is the carbon which is the chief factor in 
 controlling the most important properties of the cast iron, and 
 we may vary this either by increasing or decreasing the total 
 amount, or else leaving the total amount the same, by increasing 
 the graphite and decreasing the combined carbon, or vice versa. 
 
324 THE METALLURGY OF IRON AND STEEL 
 
 It is the ease with which we may vary the amount or the condition 
 of the carbon, and therefore the properties of the iron, that is one 
 of the most important advantages of the materiah But, strangely 
 enough, although it is easy to keep this control, it can never 
 be accomplished in a direct way. It will be remembered (see 
 pages 29 to 32), that the blast-furnace manager can vary the 
 amount of silicon and sulphur in his pig iron at will; that he has 
 only a small control over the manganese and practically none 
 over the phosphorus or carbon, but that the metal always 
 saturates itself with this latter element; and it has also been 
 seen (see page 318) that the limit of this saturation is small. 
 
 Silicon. — By means of his control over the silicon and sulphur, 
 the metallurgist exercises indirectly his most important control 
 over the condition of the carbon; for silicon acts as a precipitant of 
 carbon, driving it out of combination and into the graphitic form, 
 so that with about 3 per cent, of silicon, slow cooling and very 
 low sulphur and manganese, we may obtain a cast iron in which 
 almost none of the carbon is in the form of cementite. That is to 
 say, the presence of this amount of silicon acts so strongly that 
 it may partially prevent the formation of austenite during solidi- 
 fication, and also cause graphite to precipitate instead of cemen- 
 tite at 690° C. (1275° F.) when the eutectoid decomposes (point 
 S, Fig. 252, page 295), so that it decomposes into ferrite and 
 graphite instead of ferrite and cementite. The maximum pre- 
 cipitation of graphite seems to occur with about 2.5 to 3.5 per 
 cent, of silicon. With each increase of silicon up to that point 
 (the amount of sulphur, the rate of cooling, and other influential 
 conditions remaining the same), we get an increase in the amount 
 of graphite precipitation, but when the amount of silicon exceeds 
 about 3 per cent, it seems to reverse its effect, and each addition 
 of silicon thereafter causes an increase in the proportion, not of 
 graphite, but of combined carbon. At this point large amounts 
 of various iron silicides (FegSi, FegSig, etc.), make their appearance. 
 Then the color of a freshly broken fracture begins to be bright 
 like a mirror, in contradistinction to the white color of ordinary 
 white cast iron, which has more nearly the appearance of frosted 
 silver. 
 
 Svlphur, — The influence of sulphur upon the formation of 
 graphite is almost the exact opposite of the influence of silicon. 
 That is to say, each increase in the amount of sulphur present 
 increases the amount of combined carbon in the iron. It is 
 
THE CONSTITUTION OF CAST IRON 
 
 325 
 
 usually considered that each 0.01 per cent, of sulphur will neu- 
 tralize fifteen times as much silicon (i.e., QA5 per cent.) in its 
 effect upon the condition of the carbon in the iron. It is also 
 very important -to note that when the sulphur is in the form 
 of MnS, it is not so potent in increasing the combined carbon 
 as when it is in the form of FeS. An interesting example of 
 this is shown in the analysis of the two railroad car wheels given 
 below: 
 
 - 
 
 Fe 
 
 Total 
 Carbon 
 
 Si 
 
 Mn 
 
 P 
 
 S 
 
 Graph- 
 ite 
 
 Good wheel . 
 Poor wheel. . 
 
 94.79 
 95.00 
 
 3.84 
 3.52 
 
 0.69 
 0.65 
 
 0.13 
 0.12 
 
 0.43 
 0.52 
 
 0.12 
 0.19 
 
 3.30 
 2.35 
 
 c.c. 
 
 0.54 
 1.17 
 
 Good wheel required 150 blows of 25-lb. sledge to break it. 
 required 8 blows of 25-lb. sledge to break it. 
 
 Poor wheel 
 
 It will be observed that the poor wheel has more than twice as 
 much combined carbon as the good wheel, although the sulphur 
 in the poor wheel is only about 50 per cent, more than the sulphur 
 in the good wheel, the other impurities being nearly the same. 
 When we come to figure out the amount of MnS and FeS in the 
 two wheels, we find, however, the explanation of the large amount 
 of combined carbon in the poor wheel. We also have an ex- 
 planation of the poor quality of this wheel in the increased amount 
 of FeS present. 
 
 Constituent 
 
 Good wheel 
 per cent. 
 
 Bad wheel 
 per cent. 
 
 MnS 
 
 0.206 
 0.121 
 2.045 
 2.755 
 67.610 
 23.963 
 0.000 
 3.300 
 
 0.195 
 
 PeS 
 
 0.315 
 
 PeSi 
 
 1.923 
 
 Fe,P 
 
 3.335 
 
 Pearlite 
 
 84.492 
 
 Ferrite 
 
 0.000 
 
 Cementite 
 
 7.390 
 
 Graphite 
 
 2.350 
 
 
 
 Totals 
 
 100.000 
 
 100.000 
 
 
 
326 THE METALLURGY OF IRON AND STEEL 
 
 Manganese, — Manganese increases the total carbon in pig 
 iron. Manganese also increases the proportion of the carbon that 
 is in the combined form, but its influence in this respect is far 
 less than that of the sulphur; moreover, the statement requires 
 the following qualification: as much manganese as is combined 
 with sulphur in the form of MnS does not increase the proportion 
 of carbon in the combined form. Indeed, it has really the reverse 
 effect, because it takes the sulphur out of the form of FeS, in 
 which it is most powerful in increasing the combined carbon. In 
 this sense, therefore, the manganese actually decreases the amount 
 of combined carbon. 
 
 The excess manganese over that necessary to form MnS (that 
 is, the manganese in the form of [FeMnJgC) increases the propor- 
 tion of carbon in the combined form, and also increases the amount 
 of total carbon even more potently than does the manganese, 
 which is in the form of MnS. We therefore have a strange con- 
 tradiction, in that when the manganese is high an increase in 
 sulphur will, by decreasing the amount of (FeMn)3C, actually 
 decrease the tendency of manganese to raise the total carbon as 
 well as the combined carbon. To sum up, manganese and sul- 
 phur both tend to increase the total carbon and the combined 
 carbon, and yet they neutralize each other in this respect. 
 
 Phosphorus, — The effect of phosphorus upon the carbon is 
 somewhat self-contradictory; from a chemical standpoint it tends 
 to increase the proportion of combined carbon, and this is es- 
 pecially true when the silicon is low and the phosphorus high 
 (say above 1.25 per cent.). But phosphorus also has the effect of 
 lengthening the period of solidification. That is to say, it makes 
 the iron pass through a somewhat mushy stage of solidification, 
 and this mushy stage lasts for several minutes. This lengthening 
 of the solidification period gives a longer time in which graphite 
 can precipitate. Therefore, when the silicon is relatively high 
 (at least over 1 per cent.), and there is consequently a strong 
 tendency for graphite to precipitate during solidification, this 
 precipitation is actually aided by the phosphorus, and the graphite 
 occurs not only more abundantly, but in larger-sized flakes. 
 When, however, the amount of phosphorus is very large, its 
 chemical effect is great enough to retain the carbon in the com- 
 bined form, in spite of the long period of solidification. We may 
 sum this up by saying that if the chemical conditions are such 
 that graphite is bound to precipitate, then the physical effect of 
 
THE CONSTITUTION OF CAST IRON 327 
 
 the phosphorus makes this precipitation the more easy; but if 
 there is enough phosphorus present to produce a strong chemical 
 effect of its own, or if the other chemical influence is not very 
 powerful (i.e., if the silicon is low), then phosphorus tends to keep 
 the carbon in the combined form. 
 
 The Properties of Cast Iron 
 
 Let us now consider the properties of cast iron, and summar- 
 ize under the head of each the influence of the various elements 
 and conditions upon them. 
 
 Shrinkage. — The shrinkage of cast iron is of more importance 
 than might at flrst appear, because the greater it is the greater 
 will be the strains set up in the cooling of the casting, and con- 
 sequently the liability to check; also, the greater will be the 
 allowance necessary in order that the casting may be true to the 
 size called for by the drawings. Graphite is the most important 
 impurity in this connection, because of the expansion which its 
 separation causes. This separation should take place at the 
 moment of solidification, but is usually not complete, then, and 
 therefore the precipitation continues during the fall of the tem- 
 perature t6 several degrees below the freezing-point. Further- 
 more, when the silicon is high, graphite instead of cementite 
 separates at the lower critical point (i.e., the line PSK in Fig. 253, 
 page 296), As silicon and the rate of cooling are the chief 
 influences which control the separation of graphite, they become 
 the governing factors in the shrinkage of the iron. Indeed, when 
 sulphur is practically normal and no other unusual conditions 
 prevail, there is such a close relation between the size of the 
 castings^ and the percentage of silicon on the one hand, and the 
 amount of shrinkage on the other hand, that any one of the three 
 may be calculated when the other two are known. 
 
 Sulphur is important in this connection, and its effect is con- 
 trary to that of silicon, because of its tendency to retain the 
 carbon in the combined form. Manganese and phosphorus each 
 has a less important influence. Manganese, by increasing the 
 total carbon, tends to increase graphite and therefore decrease 
 shrinkage. So far as it neutralizes sulphur, moreover, its effect 
 is in the same direction. Phosphorus decreases shrinkage, both 
 because it contributes to the fluidity of the metal and therefore 
 
 ^ Which is the chief influential feature in the rate of cooling. 
 
328 
 
 THE METALLURGY OF IRON AND STEEL 
 
 gives a better opportunity for carbon to separate, and also 
 because of the expansion caused when the phosphorus eutectic 
 separates from solution. A hotter casting temperature of the 
 iron has the effect of increasing combined carbon. 
 
 TABLE XXVIL— RELATION OF SHRINKAGE TO SIZE AND 
 PERCENTAGE OF SILICON 
 
 Perpedicular 
 readings show 
 decrease due to 
 increase in sili- 
 con. 
 
 Horizontal read- 
 ings show de- 
 crease of 
 shrinkage due 
 to size. 
 
 Per. 
 cent, of 
 silicon 
 
 i in. 
 square 
 
 lin. 
 square 
 
 2 in.X 
 1 in. 
 
 2 in. 
 
 square 
 
 Sin. 
 square 
 
 4 in. 
 square 
 
 1.00 
 
 ,183 
 
 .158 
 
 .146 
 
 .130 
 
 .113 
 
 .102 
 
 1.50 
 
 .171 
 
 .145 
 
 .133 
 
 .117 
 
 .098 
 
 .087 
 
 2.00 
 
 .159 
 
 .133 
 
 .121 
 
 .104 
 
 .085 
 
 .074 
 
 2.50 
 
 .147 
 
 .121 
 
 .108 
 
 .092 
 
 .073 
 
 .060 
 
 3.00 
 
 .135 
 
 .108 
 
 .095 
 
 .077 
 
 .059 
 
 .045 
 
 3.50 
 
 .123 
 
 .095 
 
 .082 
 
 .065 
 
 .046 
 
 .032 
 
 A table showing the relation between the size of the casting, 
 the amount of silicon, and the shrinkage, is given above and 
 is taken from No. 93. Slight changes must be made in 
 this table by each foundry for the conditions of sulphur, phos- 
 phorus, temperature, etc., obtaining there; but those given 
 herewith will be found sufficiently accurate for all ordinary 
 purposes where conditions are anywhere near normal. 
 
 Shrinkage Tests. — At many foundries it is the custom to make 
 a shrinkage test of the iron from each cupola at least once a day. 
 The simplest way of making these tests is to pour into a mold 
 12 in. long, witTi a sectional area approximately proportionate 
 to the size of the castings made, some of the iron from about the 
 middle of the cupola run. The casting must be poured flat, and 
 the difference between 12 in. and the length of the cold bar is the 
 shrinkage of the metal. This method is somewhat crude and, 
 although it gives valuable results, has been greatly improved by 
 W. J. Keep^ and Prof. T. Turner,^ who have devised simple and 
 inexpensive pieces of apparatus whereby the iron, after it begins 
 its solidification, draws a curve showing first the expansion and 
 later the contraction. It is by means of Professor Turner's appa- 
 ratus that the curves shown in Fig. 275, page 320, were made. 
 
 1 See Chapter XX of No. 93. 
 ^ Reference on page 320. 
 
THE CONSTITUTION OF CAST IRON 
 
 329 
 
 With very little care these curves can be obtained to show with 
 sufficient accuracy for all ordinary purposes the percentage of 
 graphite and also /other conditions being normal, or nearly so), 
 the percentages of silicon and combined carbon, and the strength, 
 hardness, and porosity. Indeed the curves are more useful than 
 many single tests, because they show at a glance the net effect 
 of several varying conditions. 
 
 Density. — The maximum density of gray cast iron occurs with 
 about 1 per cent, of silicon. With l^ss than that, the iron is liable 
 to contain spongy spots, due to high shrinkage on account of low 
 graphite. With more silicon the separation of graphite decreases 
 density. Above 2 per cent, of silicon, the grain of the iron 
 becomes so open as to be actually porous and the density falls off 
 by 12 per cent. 
 
 TABLE OF DENSITES. 
 
 . 
 
 Specific 
 
 gravity 
 
 Weight per 
 cubic foot 
 
 Pure iron 
 
 7.86 
 7.60 
 7.35 
 7.20 
 6.80 
 7.17 
 6.65 
 
 490 
 
 White cast iron 
 
 474' 
 
 Mottled cast iron 
 
 458 
 
 Light gray cast iron 
 
 450 
 
 Dark gray cast iron 
 
 425 
 
 Sample of gray cast iron when cold 
 
 Same, when liquid 
 
 448 
 416 
 
 
 
 To make a close-grained iron for hydraulic work the sulphur 
 should be from 0.03 to 0.055 per cent. If more than this, the iron 
 is liable to be dirty, to contain spongy spots on account of low 
 graphite, to be difficult to machine on account of high combined 
 carbon, and to be weak, because high sulphur, aside from its effect 
 on carbon, reduces the strength. Especially, if the phosphorus is 
 high must the ^sulphur be kept down to these limits, or the iron 
 will be hard, brittle, and weak. 
 
 There should be from 0.4 to 0.6 per cent, of manganese. We 
 do not want more manganese than this, or the casting will be 
 difficult to machine. We do not want less than I have indicated, 
 because manganese assists in counteracting the bad effects of 
 sulphur and phosphorus. 
 
330 THE METALLURGY OF IRON AND STEEL 
 
 Phosphorus has a double-acting influence on the porosity of 
 cast iron: (1) It increases the size of the crystals, decreases 
 shrinkage and causes a large expansion aftej* solidification, as 
 explained in connection with Fig. 275; but '(2) it fills all the 
 crevices between the crystals and in the interior of the iron, which, 
 by decreasing the porosity, counteracts its first influence. When 
 the phosphorus is high, the phosphorous and iron form a eutectic, 
 which remains fluid for a long time and fills the tiniest crevices in 
 the interior of the metal. For this reason iron for hydraulic 
 work may run up to 0.7 per cent, phosphorus, but above that the 
 iron is liable to be weak and "cold-short," especially under 
 impact. In fact, where very strong iron is desired, the phos- 
 phorus should be kept down to 0.4 per cent, at least. 
 
 With the various amounts of impurities mentioned above the 
 combined carbon will be in the neighborhood of 1 per cent, and 
 the graphite about 2.5 per cent., the exact amounts depending 
 upon the thickness of the castings and the rate at which they are 
 cooled. If we desire to keep the combined carbon the same and 
 reduce the graphite it will be necessary to reduce the total carbon. 
 This can be accomplished by mixing in steel scrap and melting 
 fast in the cupola, or by melting in an air-furnace instead of a 
 cupola. This reduction in graphite results in a closing of the 
 grain of the steel, with consequent increase in strength and 
 density. 
 
 Segregation. — A common cause of porosity in castings is 
 segregation, or the collection together of impurities in spots. 
 This segregation is the greatei* the greater the amounts of phos- 
 phorus, sulphur, manganese, and silicon. Phosphorus increases 
 the segregation by making a fluid eutectic, which does not solidify 
 until after the remainder of the casting, but then runs into that 
 part of the metal having the loosest texture. This part is 
 usually in the middle of the larger sections of the casting, and 
 when the silicon is high and there are shrinkage spots the segre- 
 gation will be excessive in the neighborhood of these spots. 
 Manganese and sulphur are also liable to collect in the same way 
 and place. These localities, where the segregation is high, and 
 which are known, when very bad, as "hot spots," are sometimes 
 porous or surrounded by porous parts of the casting. They are 
 sometimes so extremely hard that no tool will cut them. One 
 way of getting rid of them is to use very large risers, or headers, 
 which solidify last and serve as feeders for the remainder of the 
 
THE CONSTITUTION OF CAST IRON 331 
 
 metal. Under these circumstances the segregation occurs in 
 the riser, and is thus temporarily removed. This method is 
 not advisable as a regular practice, however, because these 
 risers xxltimately find their way back into the cupola as scrap 
 and result in increasing the impurities in a subsequent set of 
 castings. 
 
 Headers themselves increase the density of iron castings by 
 feeding the metal and so preventing the porous spots, and also by 
 keeping the metal under a pressure during solidification. This 
 latter is especially serviceable when the phosphorus is high, 
 which tends to make the metal expand during solidification, as I 
 have shown. 
 
 Checking, — The time when a casting usually checks is when 
 it is just above the black heat, when the metal is in a weak and 
 tender condition and, as shown by Fig. 275, page 320, is under 
 strain because it is contracting upon the sand. Sulphur greatly 
 increases weakness at this temperature, because both sulphide 
 of manganese and sulphide of iron are now in a pasty condition, 
 and therefore offer very little resistance to breaking. The sul- 
 phide of iron is much worse, however, because this is spread out 
 in thin plates or membranes which offer much more extended 
 planes of weakness than the sulphide of manganese, which is in 
 small spots or bubbles, resembling blow-holes in its effect. 
 Phosphorus, by decreasing shrinkage, decreases the liability of 
 checking, but phosphorus has another influence, shown in the 
 production of large-sized crystals, and in this respect it increases 
 the liability of the metal to check. 
 
 Manganese, by decreasing the size of crystals, tends to counter- 
 act partially the effect of the phosphorus. The size of crystals 
 can also be decreased to some extent by chilling the weak points 
 and feeding them well under a head of metal. Feeding all 
 localities liable to check has the double advantage of lessening 
 shrinkage and segregation, both of which increases the liability 
 to checking. 
 
 SoftnesSf Workability j and Strength. — It is the combined carbon 
 which is the great hardener of cast iron, the other elements 
 producing hardness chiefly in proportion as they produce 
 combined carbon, except manganese, which not only pro- 
 duces combined carbon, but also produces a compound 
 having the formula (FeMn)3C, which is very hard and difficult 
 to machine. 
 
332 THE METALLUKGY OF IRON AND STEEL 
 
 Silicon, by decreasing combined carbon, decreases hardness. 
 When we get above 3 per cent, silicon, however, there begin to 
 form new compounds with silicon which make the iron hard. 
 Furthermore, silicon above 3 per cent, increases combined carbon, 
 instead of decreasing it. 
 
 The maximum softness of cast iron is obtained with about 
 2.5 to 3 per cent, of silicon, the sulphur being not above 0.1 per 
 cent, and the manganese not above 0.4 per cent. In large or 
 slowly cooled castings the silicon should be near the lower limit, 
 and in small or rapidly cooled castings near the upper limit, in 
 order that the combined carbon may be down below 0.15 per 
 cent, and the graphite more than 3 per cent. Such a cast iron 
 would correspond to a soft steel, mechanically mixed with crys- 
 tals of graphite. This soft steel would machine with great ease, 
 and the graphite would act as a lubricant for the cutting tool. 
 The mixture will have a transverse strength of about 2000 to 2200 
 lb., will be low in density and open in grain. To increase the 
 strength without increasing the hardness, the best way is to cut 
 the sulphur and phosphorus down to a low point, if possible, 
 because sulphur, and next to it phosphorus, are the impurities 
 which weaken iron most (aside from their influence on carbon) . 
 Another way is to decrease the total carbon, and hence the graph- 
 ite, because graphite crystals, especially if large, are great weak- 
 eners of cast iron. 
 
 The strength of steel is more than double the strength of cast 
 iron, the difference being due to almost altogether to the graphite 
 in cast iron, because silicon in itself (aside from its influence on 
 the carbon) is a strengthener of both iron and steel up to at Jeast 
 4 per cent. 
 
 To a slight extent the total carbon may be reduced by 
 melting steel scrap with the iron, or by decreasing the amount 
 of manganese, provided that the manganese left be always at 
 least twice the sulphur, otherwise the iron will be weak and 
 brittle. 
 
 The strength of cast iron may be increased by increasing the 
 combined carbon, but this is done at the expense of softness 
 and workability. Cast iron containing from 1.5 to 2 per cent. 
 of silicon (depending upon the size of the castings and rate of 
 cooling), 0.9 per cent, of combined carbon, 0.5 per cent, of 
 manganese and not more than 0.08 per cent, of sulphur and 0.3 
 per cent, of phosphorus, will work without difiiculty in the 
 
THE CONSTITUTION OF CAST IRON 333 
 
 machine shop and have a tensile strength of over 28,000 lb. per 
 square inch. In many cases foundries are unwilling to go to the 
 expense of such a low sulphur and phosphorus. In this case the 
 strength must be obtained by raising the manganese, which is not 
 advisable, as it decreases the softness more than any other 
 element, causes dull iron and high total carbon. 
 
 An important point in connection with the strength of cast 
 iron is the size of the crystals of graphite — the smaller these 
 crystals are the greater the strength, because the smaller are the 
 planes of easy rupture. A notable example of this is malleable 
 cast iron, which may have a tensile strength of 45,000 lb. per 
 square inch, even when the percentage of graphitic or temper 
 carbon is as high as 3 per cent. The very small size of the flakes 
 of the temper carbon does not reduce the strength as much as the 
 same amount of the larger graphite crystals. 
 
 It is believed by many that smaller graphite crystals are 
 obtained by mixing different brands of iron in the cupola, even 
 though the analysis of the mixture may be the same. This, 
 however, is denied by others, and no reliable data exist upon 
 which we can base a definite statement. It is also believed by 
 many that when the silicon is added to the cast iron immediately 
 before pouring into the molds, the crystals of graphite are smaller 
 than those formed when high-silicon irons are melted in the 
 cupola. The practice of adding a small amount of ferrosilicon 
 to the ladle of cast iron after it is received from the cupola is thus 
 said to be doubly advantageous, because the silicon does not 
 have to go through the cupola, where it suffers some oxida- 
 tion, and it produces the desired softness by precipitating the 
 graphite, but in a form which does not decrease strength so 
 much. 
 
 To obtain high transverse strength, the silicon should be 
 about 0.2 per cent, lower and the combined carbon about 0.2 
 per cent, higher than the figures given for tensile strength. 
 Otherwise the effects are very similar. 
 
 Some estimates of the strength and workability of cast iron 
 are given in the following table,^ and in No. 94; 
 
 Chills. — In the making of cast-iron rolls, railroad car wheels, 
 anvils, etc., at least one surface of the casting is desired to have 
 great hardness, to resist wear, and to be backed by metal which 
 shall be stronger and not so brittle. This is accomplished by 
 
 ^ Abstracted from page 197 of No 120. 
 
334 THE METALLURGY OF IRON AND STEEL 
 
 TABLE OF CAST-IRON STRENGTH AND WORKABILITY 
 
 
 Silicon 
 per 
 cent. 
 
 Sul- 
 phur 
 per 
 
 Phos- 
 phorus 
 per 
 
 Man- 
 ganese 
 per 
 
 Tensile 
 
 strength 
 
 lb. 
 
 Trans- 
 verse 
 strength 
 
 
 cent. 
 
 cent. 
 
 cent. 
 
 lb. 
 
 Soft iron for pul- 
 
 2.20 
 
 Not 
 
 Not 
 
 0.30 
 
 
 
 leys, small cast- ■ 
 
 to 
 
 over 
 
 over 
 
 to 
 
 28,000 
 
 2200 
 
 ings, good tool ng 
 
 2.80 
 
 0.085 
 
 0.70 
 
 0.70 
 
 
 
 Medium iron for 
 
 1.40 
 
 Not 
 
 Not 
 
 0.30 
 
 
 
 engine cylinders, 
 
 to 
 
 over 
 
 over 
 
 to 
 
 30,000 
 
 2500 
 
 gears, etc 
 
 2.00 
 
 0.085 
 
 0.70 
 
 0.70 
 
 
 
 Hard-iron cylin- 
 
 1.20 
 
 
 
 
 
 
 ders for ammo- 
 nia, air-com- 
 pressors, etc 
 
 to^ 
 1.60 
 1.60 
 
 to'* 
 1.90 
 
 Not 
 over 
 0.095 
 
 0.70 
 
 to 
 0.40 
 
 Not 
 over 
 0.60 
 
 25,000 
 
 2800 
 
 chilling the surface which it is desired to harden, and so pro- 
 ducing white cast iron to varying depths, reglated at the will 
 of the foundrymen. The making of this kind of casting is 
 one of the most difficult problems of cast-iron metallurgy. The 
 metal must be very close to the given composition, and the 
 temperature of the mold, of the chill, and of the metal when 
 cast must be regulated with care. Therefore, air-furnaces^ are 
 often employed for melting in this class of work, or else uniform 
 conditions of cupola melting are maintained with great care, and 
 very little, if any, scrap which must necessarily be of somewhat 
 uncertain analysis is used, except the return scrap from the 
 foundry itself, that is, defective castings, sprues, gates, shot-iron 
 spillings, etc., and also scrap castings of like nature, such as 
 worn-out car wheels and broken rolls. The most important 
 factors in regulating the depth of the chill are the silicon and the 
 sulphur, and in the following table is given the depth of clear 
 chill from the surface for several different percentages of silicon 
 and sulphur. The figures here given must only be taken as ap- 
 proximations, as they will vary to an important extent with 
 different conditions in each foundry; but, starting with this as a 
 
 1 If annealed. ^ If cooled fast. » See page 343L 
 
THE CONSTITUTION OF CAST IRON 
 
 335 
 
 FIG. 276.— WHITE PIG IRON. 
 0.75 per cent. Si. 0.120 per cent. S. 
 diameters. Unetched. 
 
 FIG. 277.— GRAY PIG IRON. 
 50 0.75 per cent. Si. 0.012 per cent. S. 
 diameters. Unetched 
 
 50 
 
 FIG. 278.— GRAY PIG IRON. 
 1.75 per cent. Si. 0.025 per cent. S 
 diameters. Unetched 
 
 FIG. 279.— GRAY PIG IRON. 
 
 50 2.5 per cent. Si. 0.012 per cent. S. 50 
 diameters. Unetched. 
 
 FIG. 280.— GRAY PIG IRON. 
 3.5 per cent. Si. 0.025 per cent. S. 
 diameters. Unetched. 
 
 FIG. 281.— NO. 2 CHARCOAL PIG IRON 
 50 VERY SLOWLY COOLED. 
 
 60 diameters, HNOa. 
 
336 
 
 THE METALLURGY OF IRON AND STEEL 
 
 basis, one can quickly prepare a table for himself to suit, the 
 practice in his foundry. Phosphorus has very little effect on the 
 depth of chill, and manganese is also relatively less effective, 
 
 FIG. 282.— METHOD OF MEASURING THE DEPTH OF CLEAR CHILL IN A CAST- 
 IRON ROLL. 
 
 although it increases the hardness -of the chilled portion. The 
 hotter the iron when cast the deeper the chill, 
 
 DEPTH OF CLEAR CHILL FROM SURFACE IN INCHES 
 
 Silicon 
 per cent. 
 
 Sulphur 
 0.2 
 
 per cent. 
 
 Sulphur 
 
 0.15 
 per cent. 
 
 Sulphur 
 
 0.1 
 per cent. 
 
 Sulphur 
 
 0.075 
 
 per cent. 
 
 Sulphur 
 
 0.05 
 per cent. 
 
 1.25 
 
 1.00..... 
 
 0.75 
 
 . 50 .... 
 
 0.625 
 1.000 
 1.500 
 
 0.250 
 0.625 
 1.000 
 1.500 
 
 0.125 
 0.250 
 0.625 
 1.000 
 1.250 
 
 0.000 
 0.125 
 0.250 
 0.625 
 1.000 
 1.500 
 
 0.000 
 0.000 
 0.125 
 0.250 
 fi9.'> 
 
 0.40 
 
 
 0.30 
 
 
 
 1 000 
 
 
 
 
 
 
THE CONSTITUTION OF CAST IRON 337 
 
 Casting Temperature of Iron, — The strength of cast iron is 
 materially effected by the temperature at which it is cast. 
 Longmuir (Reference No. 123) found that the tensile strength of 
 iron, when cast too hot, was 33 per cent, less, and when cast too 
 cold was 26 per cent, less than the same metal cast at the correct 
 temperature. For the cast iron with which he experimented, 
 the correct temperature of casting seemed to be in the neigh- 
 borhood of 1230° C. (2246° F.). 
 
XIII 
 MALLEABLE CAST IRON 
 
 Malleable cast iron has physical properties between gray 
 iron and steel castings. Its tensile strength will vary between 
 40,000 and 60,000 lb. per sq. in. with an elongation of 2 1/2 to 
 51/2per cent, in 2 in. and a reduction" of areaof 21/2 to 8 percent.^ 
 In a transverse test, a 1-ih. square bar on supports 12 in. 
 apart, should bear a load as the center of at least 3000 lb. with 
 a maximum deflection of 1/2 in.; the maximum load may go as 
 high as 5000 lb. and the deflection up to 2 in. or more. In 
 respect to its resistance to repeated stresses and to blows which 
 
 flatten the face of the material without 
 extending their influence very deep, 
 malleable cast iron is especially valu- 
 able. It also has great resiliency and 
 (when very thin) is capable of being 
 flattened out or bent double without 
 cracking. A good test for malleable 
 cast iron is the so-called test lug 
 shown in Fig. 285. These are bent 
 over in the form of a curled-up shav- 
 ing by light blows of a hammer, begin- 
 ning at the knife edge. The distance 
 along the test wedge that the piece 
 will bend before it begins to break, and 
 especially the maximum angle pro- 
 duced, will give a good idea of the 
 softness of the material. 
 Malleable iron castings are used very largely for railroad 
 work. They have long been used for couplers, but are now being 
 replaced to some extent by steel castings for this purpose. On 
 the other hand, malleable cast iron is replacing gray cast iron for 
 other parts, such as journal boxes. Malleable cast iron is 
 also used very largely for parts of argicultural machinery, for 
 
 ^ In the case of iron very carefully melted and annealed in iron oxide, the elongation 
 may go as high as 10 per cent., and the reduction of area as high aa 12 per cent, in rare 
 
 fttttaa 
 
 338 
 
 FIG. 285.— TEST LUG FOR 
 MALLEABLE CAST IRON. 
 
MALLEABLE CAST IRON 339 
 
 pipe fittings— i.e., elbows, unions, valves, etc. — for household and 
 harness hardware and a great variety of work where many cast- 
 ings are to be made of the same pattern, especially castings of 
 small size. Even hammers, hatchets, skates, wood chisels, 
 etc.,' are often made of "black heart" malleable cast iron. 
 
 Process of Manufacture. — The process, which was invented 
 by Reaumur in 1722, but has only been in practical use about 
 100 years, and of importance less than 50, is still followed out 
 in substantially its original form in England and Europe to 
 make what is known as "white heart" malleable castings. 
 In America this process has been modified to make the so-called 
 "black heart" castings, as foUows: The metal is melted and 
 cast into molds of the desired shape and size, and after cooling 
 its fracture will be entirely white, or else white with a small 
 area of mottled iron in the center of pieces 1 in. thick or heavier. 
 In short the composition of the metal is such that it is just on 
 the line between the precipita- 
 tion of grapihte and its retention 
 in the form of cementite and dis- 
 solved carbon. In all cases the 
 majority of the carbon must be 
 dissolved or combined. After 
 carefully cleaning, the castings 
 are reheated to a temperature of 
 625*^ to 875° C. (1250° to 1600° 
 F.), which is roughly 450° C. 
 below their melting-point. They 
 are kept at this temperature for 
 some 60 -hours and under these ^ig. 286.-annealed malleable 
 
 . . CAST IRON, 
 
 conditions occurs the preCipita- Magnified so diametera. Unetched. 
 
 tion of graphite, which normally 
 
 should have occurred during the first cooling. In the majority 
 of cases almost all the combined carbon throughout the body 
 of the casting is changed to graphite. But the graphite does 
 not here form in flakes, as in ordinary gray cast iron, but in a 
 finely comminuted condition, like a powder, to which the name 
 of "temper carbon" or "temper graphite" is given (see Fig. 
 286). In this form it is not nearly so weakening or embrittling 
 to the casting as flakes of graphite would be.^ 
 
 A broken fracture of such an annealed casting will have a 
 
 1 We may liken this to two sainples of putty, in one of which had been embedded a large 
 number of plates of micfa, and in the other the same amount of mica ground to powder. 
 
340 
 
 THE METALLURGY OF IRON AND STEEL 
 
 black center with a rim of white, 1/64 to 1/16 of an inch thick 
 around the outside. The dark center owes its appearance to the 
 precipitated temper carbon, and the white outer rim. consists 
 of iron from which the carbon has not been only precipitated, but 
 actually removed.^ That is, the metal of this outer rim corre- 
 sponds to a soft steel impregnated with a great number of tiny 
 cavities formerly occupied by carbon and which act as a cushion 
 to shocks and battering blows of service. This outer rim is the 
 portion of the casting in which the metallic crystals have formed 
 perpendicularly to the cooling surface as explained on page 248. 
 The crystalline form in this location favors the reaction for the 
 mechanical removal of carbon which will be explained later. 
 
 The Manufacture of RSaumuVj or White Heart Malleable. — In 
 the original, and present European, process of making malleable 
 castings, they are made very thin; they are then annealed at the 
 higher temperatures given above and for a period nearly twice 
 as long, with the result that the conditions described above as 
 prevailing in tHe outer rim, extend here throughout the entire 
 casting. In short, the carbon is mechanically removed, even 
 from the center portion. White heart castings are not as strong 
 as black heart and rarely, if ever, are above 1/2 in. in thickness. 
 
 Manufacture of All-black Castings. — In America pipe fittings 
 and other castings not requiring much strength are sometimes 
 annealed in lime, fireclay, or some other substance which does not 
 decarburize even the skin of the metal. This results in the 
 carbon being deposited as temper carbon, but not being removed. 
 The castings are several thousand pounds per square inch lower 
 in tensile strength, and have substantially no white rim. 
 
 Pig Iron Used. — The analysis of the " hard castings,'' that is, the 
 white castings before the annealing process, should run as follows: 
 
 If melted ia 
 cupola 
 
 If melted in air fur- 
 nace or open hearth 
 
 Silicon, per cent 
 
 Manganese, per cent. . . 
 Phosphorus, per cent^. 
 Sulphur, per cent^ . . . 
 Total carbon, per cent 
 
 0.75 to 1.25 
 0.30 
 
 under 0.225 
 0.04 to 0.32 
 above 2 . 75 
 
 0.45 to 1.00 
 0.30 
 
 under 0.225 
 0.04 to 0.07 
 above 2 . 75 
 
 1 By the material in which they are packed according to the. following reaction: 3C + 
 Fe203=3CO + 2Fe. 
 
 2 In European pTactice the phosphorus and sulphur are often much higher than this. 
 
MALLEABLE CAST IRON 341 
 
 This analysis is obtained by mixing pig iron with varying amounts 
 of cast iron and steel scrap. 
 
 Not more than about 20 per cent, of bought scrap is used on 
 the average in American practice, and a good deal of thie is steel, 
 on account of the desirability of lower total carbon, and because 
 iron scrap is too impure, too variable, and too uncertain in samp- 
 ling and chemical analysis for castings requiring strength, such 
 as those for railroads and machinery. There is, however, always 
 a-large amount of "return scrap" from the foundry, consisting 
 of defective castings, sprues, gates, etc., which, in the case of 
 small castings, may be greater in weight than the castings 
 themselves. This return scrap is low in total carbon and silicon 
 as a result of having already suffered the melting changes. 
 Steel scrap is, of course, a prime factor in reducing the total 
 carbon. The return scrap shotdd all be cleaned from adhering 
 sand to avoid difficulty in melting. The pig iron bought for 
 malleable iron work is known in the trade under the name of 
 '^coke malleable." 
 
 Silicon. — The proportion of silicon will depend upon the size 
 of the casting and the amount of total carbon, because the greater 
 each of these is the less will be the amount of silicon that will 
 cause a precipitation. Furthermore, cupola metal requires 
 more silicon by about 0.25 per cent, than air furnace metal 
 because of the difficulty with which it yields to the annealing 
 operation. It might at first appear that the less silicon the better; 
 but this is not altogether so, because temper carbon will not 
 come out during annealing unless a certain amount of silicon is 
 present; and the more there is the more quickly, easily, and 
 completely will the precipitation occur. For air furnace castings 
 one inch thick the silicon may be as low as 0.35 per cent.; but 
 this is unusual, as three-quarters of an inch thickness is rarely 
 exceeded. For half-inch castings the silicon will be about 0.60 
 per cent., and for very thin and light castings with low total 
 carbon and high sulphur (say 0.2 to 0.3 per cent.), the silicon 
 may be up to 1 per cent. To the" percentage of silicon desired 
 in the castings, we must add the amount which will be burned 
 out in melting. In the cupola this will be about 0.2 to 0.25 per 
 cent., and in the air-furnace from 0.15 to 0.5 per cent., depending 
 on the length of time the metal is kept in the furnace after melt- 
 ing. The hotter we want the iron, the longer this time must be, 
 and therefore the higher the silicon in the original mixture charged. 
 
342 THE METALLURGY OF IRON AND STEEL 
 
 Sulphur. — Sulphur increases the tendency of castings to 
 check, which is especially important in malleable work on account 
 of the contraction of white iron being nearly double that of gray 
 iron. Sulphur also reduces the strength and the ease of annealing. 
 For this reason over 0.06 per cent, should not be permitted in 
 castings requiring strength, but it actually runs up to 0.2 and 0.3 
 per cent, in inferior metal, both in America and England, and 
 especially in small castings, which do not need strength so much, 
 and which, having less length for contraction are not so liable 
 to be checked by cooling strains. 
 
 Manganese. — Low manganese is preferred by many foundries, 
 and one of the highest authorities in America places the limit at 
 0.8 per cent. It should be remembered, however, that the man- 
 ganese should be at least twice the sulphur, and preferably three 
 times, except when the sulphur is as high as 0.3 per cent. Man- 
 ganese of 0.5 per cent, tends to decrease checking. It also 
 protects silicon from oxidation, both during melting and anneal- 
 ing, and on this account hastens and makes more complete the 
 precipitation of temper carbon. It also protects the iron itself 
 from oxidation during annealing and thus prevents the formation 
 of "scaled'* castings. More than 0.6 per cent, manganese makes 
 the iron hard and difficult to machine, which is disadvantageous, 
 especially for pipe-fittings, which must be threaded with great 
 economy in order to meet the trade competition. It also makes 
 the castings difficult to anneal. 
 
 Phosphorus. — ^Phosphorus makes the metal fluid, which is 
 especially desirable where total carbon and silicon are low, or 
 where sulphur and manganese are high. On the other hand, it 
 diminishes two of the most valuable properties of the material: 
 its resiliency and resistance to shocks. It also makes the metal 
 hard, difficult to machine, and liable to check, and amounts over 
 0.225 per cent, should never be permitted by engineers where the 
 castings are subjected to strain. 
 
 Total Carbon. — Tot-al carbon below 2.75 per cent, gives trouble 
 in annealing and therefore makes the castings weak. It also 
 makes the metal more sluggish. It is difficult to get as low as 
 this in cupola melting, although mixing in large percentages of 
 steel scrap and allowing the metal to run out of the cupola as fast 
 as melted will reduce the proportion appreciably. In air-furnace 
 practice the total carbon may be reduced as far as necessary. 
 Annealing in iron oxides also removes carbon from the outer 
 
MALLEABLE CAST IRON 
 
 343 
 
 layers and even to the very center of thin castings. The lower 
 the total carbon in the annealed castings, the better. 
 
 Melting in the Air-furnace. — ^The commonest melting-furnace 
 for malleable cast iron is the air-furnace. After the metal is 
 melted it is retained in this furnace for 15 minutes to an hour 
 longer, and test ingots are poured at intervals, from the fracture 
 of which and the temperature of the iron we determine the 
 correct moment for tapping. The fracture of the test plug 
 should be a clear white throughout, except when the castings 
 are to be of very light section, in which case the metal might be 
 tapped when the test sample shows a few specks of graphite in 
 the center. The practice of judging from test plugs is different 
 
 I 
 
 FIG. 287. 
 
 in each foundry, but there must be some system which insures 
 that the metal shall be of such a composition when tapped that 
 the castings will have not more than a trace of graphite if any 
 at all (say, less than 0.15 per cent, in small castings and a little 
 more in larger ones). If more than this the annealed castings 
 will be very weak, almost "rotten." 
 
 The diameter of the test plug should be as large as any section 
 of casting that is to be poured. The usual size is 1 1/2-inch 
 diameter. 
 
 Air-furnace, — ^The air-furnace is a reverberatory furnace on 
 the hearth of which pig iron in melted by radiation from the flame 
 of a soft-coal fire, or, more rarely, from a gas flame. The furnace 
 is charged by means of a large side door, or by removing the roof 
 in sections bound together with iron, or by taking out the end; and 
 in some cases, to effect an economy in labor, mechanical devices 
 are employed for charging. Several designs of air-furnaces are 
 shown in Figs. 287 to 290 inclusive. 
 
344 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Lining. — The lining of the bottom is made of silica sand of 
 about the same composition as the acid open-hearth furnace, 
 i.e., containing 95 per cent, or more of silica, with just enough 
 
 FIG. 288.— AIR-FURNAGE. 
 
 lime to frit the mass together at the heat of the furnace. A layer 
 about 1 to 2 in. thick is spread all over the hearth and then set 
 on as described in the chapter dealing with the open-hearth 
 
 FIG. 289. 
 
 process. About five layers are put on in this way, and the bot- 
 tom lasts on an average of 6 to 12 heats, although some foundries 
 regularly make up a bottom after the third heat; and in other 
 
MALLEABLE CAST IRON 
 
 345 
 
 cases the bottoms last as many as 30 heats. Longer life will be 
 obtained if the material is charged carefully so as not to break the 
 
 Floor Line 
 
 'Floor lilne 
 
 sand, and if a strongly reducing flame is maintained during the 
 melting period, when there is not a bath of liquid iron to protect 
 the bottom from the corrosive iron oxide. Mixing broken 
 
346 THE METALLURGY OF IRON AND STEEL 
 
 fire-bricks of good quality and refractoriness with the sand seems 
 to give more durable bottoms. 
 
 Charging, — The sprues and small scrap are first charged and 
 spread evenly over the hearth and next the pig iron is piled in 
 carefully. The roof of the furnace not being high, the pig iron 
 must be stacked in order to get .the necessary amount into the 
 available space. Moreover, when the pig iron is stacked, indi- 
 vidual pigs can be pushed down into the liquid bath from time 
 to time so that it will not all melt together. Thus it can be 
 handled more easily and the slag comes up better. The steel 
 scrap is charged after the pig and scrap is melted and the 
 slag has come up. This prevents the steel scrap from being 
 oxidized. 
 
 - Conduct of the Operation. — The purpose of the furnace is to 
 melt the iron and bring it to the proper temperature for casting. 
 Some refining by oxidation is unavoidable, but this is not an 
 intentional feature and the object should be to get the desired 
 analysis by mixing pig iron, scrap, and steel, and not by the 
 oxidation of silicon and carbon. A long time in the furnace after 
 melting is especially to be avoided, as the iron is liable to be 
 oxidized and the silicon reduced below the desired point, making 
 the iron high. After the slag is up it sh*ould be skimmed off, 
 and this may be repeated once or twice. When the test plug 
 shows large, clear, white crystals without any pin holes on the rim, 
 and only a slight amount of mottling, the bath is ready to tap. 
 If there are pin holes in the rim of the test plug it indicates that 
 the metal has been "burned," as it is called. It will usually 
 be dull and there may even be gas in the metal, in which case the 
 iron should be poured into pigs and used over again in small 
 amounts in subsequent heats. If the test plug is mottled, it 
 shows either that the silicon is too high or that the temperature 
 is not high enough, for it will be remembered that the colder the 
 iron the more likely is graphite to separate and vice versa. 
 If the iron is too high, on the other hand, some ferro-silicon 
 should be added and rabbled into the bath. The smaller the 
 castings, the more rapidly will they be chilled and therefore the 
 more silicon should we have in the metal. 
 
 American furnaces usually vary in size from 10 to 45 tons 
 capacity, but they are built as small as 3 tons. It takes from 
 about three and one-half to nine hours to melt down a charge 
 of from 5 to 35 tons respectively, and the waste of metal by 
 
MALLEABLE CAST IRON 347 
 
 oxidation will be about 2 to 5 per cent, of the charge. The 
 fuel used should be a good grade of bituminous coal, and the 
 amount about one-fourth of the weight of the charge. 
 
 Tapping the Charge, — The more rapidly the metal can be 
 gotten out of the furnace and poured, the more uniform will be 
 its composition. Moreover, the metal at the top of the bath is 
 hotter than that at the bottom. For this reason, in many 
 furnaces, there are additional tap holes at a higher level, which 
 are opened first, and also holes on both sides of the furnace which 
 are used simultaneously. 
 
 Cupola Versus Air Furnace Melting. — Cupolas used for melting 
 malleable cast iron are similar to those already described but 
 the amount of fuel is larger than for most gray iron work on 
 account of the higher melting-point of the iron and the very 
 small size of castings into which it is usually poured. The fuel 
 ratio will be about one of coke to four of iron. 
 
 The advantages of the cupola for malleable cast iron work are: 
 (1) That we get metal having practically the same composition 
 at all times of the heat except the slightly higher sulphur at the 
 beginning and end; (2) the temperature is more uniform through- 
 out the heat; (3) it is cheaper to install and operate and requires 
 less skill for melting labor; (4) it can be started and stopped 
 more readily; (5) we can get hotter iron without prolonging the 
 heat with the resultant danger of burning which exists in the air 
 furnace. 
 
 The advantages of the air furnace are: (1) That there is not 
 so much sulphur absorbed; (2) with proper manipulation there 
 is not so much liability for the metal to be burned except just at 
 the thin edge of metal near the fire bridge; (3) the metal anneals 
 more easily and is stronger; (4) it is also easier to get a lower 
 total carbon by proper mixing since there is no contact between 
 coke and iron, and therefore not so much opportunity for absorp- 
 tion of carbon; (5) we can work closer to a desired analysis in the 
 air furnace than we can in the cupola. 
 
 This latter circumstance, as well as the lower sulphur, is the 
 reason why air furnaces are so largely used also for melting what 
 is known as chilled iron castings, such as rolls, etc. The analysis 
 of this chilled iron must be so closely confined within limits that 
 the outside may be white iron where it comes in contact with 
 metallic chills and the interior consist of gray cast iron. 
 
 The explanation of the more difficult annealing experienced 
 
348 THE METALLURGY OF IRON AND STEEL 
 
 with cupola metal over air furnace and open hearth metal has 
 never been made entirely clear. In general, cupola metal is 
 high in sulphur which opposes annealing, but it is believed that 
 even when the metal is of the same analysis in sulphur it requires 
 at least 150° F. higher temperature to anneal it. Another 
 circumstance in cupola metal is that it is usually higher in total 
 carbon. Consequently the carbon must be more strongly held 
 in combination to prevent some of it precipitating as graphite 
 in the hard castings. This condition may cause it to yield less 
 completely to the annealing process. 
 
 Open-hearth Furnaces. — Open-hearth furnaces of small size, 
 but in other respects like the open-hearth steel furnaces, are used 
 for melting iron for malleable castings in a few important 
 foundries in the United States. The great drawback of this 
 furnace is that it must be operated continuously, day and night, 
 which means more flooi: space on which to set molds ready for 
 pouring, because molding cannot well be done during the night, 
 as the artificial light casts shadows that make the work of finish- 
 ing up molds more difficult and confusing. 
 
 The advantage of the regenerative open-hearth furnace over 
 the air-furnace is better control of the operation, and especially 
 of the temperature, shorter heats, and greater fuel economy. 
 The average time of melting in the open-hearth furnace will be 
 2 1/2 hours for a 10-ton heat, and 4 hours for one of 20 tons. 
 The amount of fuel is about 300 to 350 lb. per ton of iron, or 
 twice as much if melting only on the day turn. The lining will 
 certainly last much longer on account of the better control of the 
 character of fiame. Moreover, oil can be used for melting in a 
 regenerative furnace, because it is introduced into a very hot 
 atmosphere, which is not practicable in the air-furnace, since 
 the fuel condenses in the cooler atmosphere of this furnace, 
 especially when the furnace is cold or when there is a cold charge 
 on the hearth. The disadvantages are high cost for installation, 
 more skilled labor and more repairs. Open-hearth linings are 
 best made of dolomite, which enables a basic slag to be produced 
 in the furnace and a reduction in phosphorus and sulphur. 
 Cheaper pig. iron and iron scrap can then be used. This basic 
 lining is not attacked by the iron oxide and slag produced in 
 melting and will last for hundreds of heats if not allowed to cool 
 off too frequently. Basic linings cannot be employed in air- 
 furnaces because dolomite contracts and expands so much on 
 
MALLEABLE CAST IRON 
 
 349 
 
 cooling and heating that the bottoms are soon cracked to pieces, 
 for air-furnaces are not operated continuously. 
 
 Molding and Pouring, — Malleable iron castings are usually 
 made in great quantities from the same pattern, so that this 
 industry offers a special field for molding machines, match 
 plates, etc. It is also a type of molding in which a great many 
 cores are used and especially cores of small size, so that core- 
 making machines are especially advantageous where applicable. 
 
 %' 
 
 FIG. 291.— TUMBLING-BARREL. 
 
 The gates and sprues are usually large in size because the metal 
 has a relatively high melting-point and is consequently more liable 
 to chill. For the same reason the metal is poured into the mold 
 as fast as possible and the great bulk of the work is done with 
 hand-ladles holding 40 or 50 lb. apiece. 
 
 Annealing-boxes. — After the castings are cooled they are 
 carefully cleaned from all adhering sand by tumbling them around 
 in a tumbling-barrel (in which they are mixed with star-shaped 
 pieces of metal something like children's jackstones), or by sand- 
 blast, or by some other suitable method. They are then packed 
 
350 THE METALLURGY OF IRON AND STEEL 
 
 in the cast-iron or steel "saggers" or annealing pots, or boxes, 
 together with the packing. Sometimes, though rarely, the tops 
 of the saggers are closed by means of an iron cover, sometimes 
 by a thick layer of the packing in the upper part, and sometimes 
 by clay or wheel-swarf. These pots last only from 7 to 20 heats 
 before they are largely oxidized away. 
 
 Annealing-ovens. — The boxes are then placed in the annealing- 
 ovens in such a way that the flame may play around them as 
 completely as possible. The general form of ovens is shown in 
 Figs. 292 to 294. The flame usually comes in at the top and goes 
 out at the bottom along the side, and thence through flues under- 
 neath the oven. The fuel used may be coke, coal, oil or gas, the 
 latter being preferable on account of the better control of the 
 temperature, which should be increased at a very gradual and 
 uniform rate during the heating up, and kept as constant as 
 possible during the annealing period. 
 
 Annealing Practice. — It takes about six days for the American 
 annealing operation, including heating up and cooling down. 
 Sometimes this can be shortened a little by decreasing the time 
 at the full annealing heat, and by cooling rapidly or drawing the 
 saggers out of the oven and dumping them while the contents are 
 still at a dull-red heat. This practice is not conducive to a good 
 quality of castings and should never be permitted in important 
 cases. The time at the full heat should never be less than 60 
 hours, and preferably it should be more than that. If less, the 
 temperature of annealing must be higher, and this decreases the 
 strength and ductility of the castings. Annealing should not 
 occupy too long a time, however, because the temper carbon 
 tends to draw together to larger flakes; besides which the metal 
 may become over annealed or else oxidized between the grains, 
 i.e., "burnt." Air-furnace castings should be annealed at 675° 
 to 760° C. (1250° to 1400° F.), and cupola metal at 825° to 
 875° C. (1500° to 1600° F.). The saggers are cooled as slowly as 
 possible in the furnace; when they are at a black heat they are 
 removed, cooled further and then dumped. 
 
 Packing. — As originally planned, the castings were annealed 
 in a packing of iron oxide crushed to a size less than a quarter of 
 an inch in diameter. The packing must surround the castings 
 at every place, both inside and out, and no two castings must 
 touch. Iron ore, mill scale, "bull-dog," and similar forms of 
 iron oxide are used for this purpose. 
 
MALLEABLE CAST IRON" 
 
 351 
 
 Annealing in iron oxide produces a white skin where the casting 
 has been deprived of its carbon, and a black interior, due to the 
 
 B Q [J] B 
 
 ■L> ^yV,>>^/^^//^^^ ^^ 
 
 
 
 Section £^ 
 
 temper carbon; whence the name of "black heart malleable" for 
 this material. Tests have shown that the casting with this white 
 
352 
 
 THE METALLURGY OF IRON AND STEEL 
 
 skin upon it is much stronger than a similar one which has not 
 been decarburized on the surface, and therefore the packing in 
 iron oxide is advantageous, even though not an essential feature 
 of the operation. When the castings are packed in some non- 
 oxidizing material, such as sand, clay or lime, they may receive 
 as perfect an annealing, as far as the production of temper 
 carbon is concerned, but will be without the white skin and of a 
 lower strength. 
 
 Shrinkage. — Hard castings contract in solidifying much like 
 steel, so that, wherever the metal is not "fed," it draws away 
 leaving crocks or cavities. Therefore the castings may be con- 
 sidered as a continuous shell of metal on the outside with an 
 interior full of cracks and planes of separation. In malleable 
 cast iron work this is known as "shrinkage." It is a source of 
 weakness in malleable castings, and is more effective in round 
 than in square bars because the latter have a thicker solid skin 
 on the corners which reduces the area of the cracked portion. 
 
 Contraction, — The contraction of the hard casting is almost as 
 great as that of steel, because almost no graphite forms. The 
 amount of silicon and the sectional area of the castings are 
 still the determining factors in this connection. Indeed, by 
 means of the measurement of the section and percentage of 
 silicon we may estimate the contraction, or by means of the 
 section and the contraction we may estimate the silicon very 
 closely, other conditions and impurities being normal. The 
 following table gives the necessary data for these estimations: 
 
 CONTRACTION IN INCHES PER FOOT OF LENGTH 
 
 Percentage 
 of silicon 
 
 1/4 Inch 
 square 
 
 1/2 Inch 
 square 
 
 3/4 Inch 
 square 
 
 1 Inch 
 square 
 
 0.35 
 0.50 
 0.75 
 1.00 
 
 0.225 
 0.220 
 0.215 
 0.211 
 
 0.200 
 0.195 
 0.190 
 0.183 
 
 0.190 
 0.183 
 0.176 
 0.137 
 
 0-175 
 
 0.170 
 0.162 
 0.102 
 
 Expansion due'to Temper Carbon. — It is a very interesting fact 
 that when the malleable cast iron is annealed and the temper 
 carbon precipitates, the casting expands to an amount approxi- 
 mately equal to that which would have occurred if the graphite 
 
MALLEABLE CAST IRON 
 
 353 
 
 had separated during solidification and gray cast iron had been 
 produced in the first instance. Jn other words, the temper carbon, 
 although in a very finely powdered condition, occupies about the 
 same amount of space as an equal weight of graphite, and causes 
 about the same ultimate difference in size between the original 
 pattern and the annealed casting as when gray cast iron is made. 
 An interesting example of this expansion in annealing is shown 
 in Fig. 295, which is a swivel snap for hitching straps. Casting 
 No. 1 is first poured, cooled and cleaned. It is then embedded 
 in the sand of a mold and casting No. 2. is poured around the 
 shank of it, as shown in No. 3 of Fig. 297. Casting No. 2 shrinks 
 
 FIGS. 295 TO 297. 
 
 upon the shank of No. 1 so as to make a close fit, and no swiveling 
 is possible, but the combined casting is now sent to the annealing- 
 ovens and annealed. This causes the expansion Tef erred to, and 
 as casting No. 2 is larger in diameter, it expands the more, and 
 now turns very easily around the shank of No. 1. 
 
 Miscellaneous Iron Products, — In the form of small castings 
 malleable cast iron and similar products often masquerade under 
 the name of steel, because under that name the producer finds a 
 readier market for them. On account of their fluidity they may 
 be cast very cheaply in small sizes, and therefore the temptation 
 to use them as a material for " cast-steel hammers," ^ " hard-steel " 
 bevel gears, "semi-steel castings," and even automobile "steel" 
 drop-f orgings, is a strong one. Engineers are warned to be on their 
 guard against a deception of this kind, for legal redress has 
 been sought many times in vain. It is usual for the manufacturer 
 when putting material of this kind upon the market to qualify 
 the name "steel" with some other letters or name, such as 
 
 1 The trade would ordinarily understand by this name hammers made of crucible steel, 
 so the use of this name is really a fraud. 
 23 
 
354 THE METALLURGY OF IRON AND STEEL 
 
 "P. Q. steel/' "Smith steel," etc.; but they all differ from true 
 steel in that they were not " cast into an initially malleable mass/' 
 Some are made by melting a large proportion of steel with cast 
 iron, after which the cooled metal may or may not be annealed 
 in iron oxide. Others are made by a long or thorough annealing 
 of ordinary malleable castings in iron oxide, by means of which 
 the metal is decarburized to some depth, and is then carburized 
 again by a cementation process. This makes a very good ma- 
 terial for some purposes, such as small bevel gears not requiring 
 strength or much ductility, but it ought not to be called "steel." 
 If the purpose for which it is to be used does not require any 
 other properties than malleable cast iron possesses, then it 
 should be used under its true name; but if it is to be used under 
 circumstances whereit is liable to strain, calling it "steel" will 
 not enable it to stand up under the work any better. The con- 
 fusion is the more easy because genuine steel is made by the 
 cementation of wrought iron, and wrought iron goes in England 
 under the name of malleable iron. In America we seldom 
 call wrought iron "malleable iron," but we often abbreviate 
 malleable cast iron to "malleable iron," or even to "malleable." 
 Delicate Adjustment of Conditions, — In no metallurgical pro- 
 cess must the conditions of chemical composition, furnace work, 
 heating temperatures and lengths of operation be adjusted more 
 delicately than in the manufacture of high-grade malleable cast 
 iron, as may be seen by the following changes that the metal 
 undergoes during the final, i.e., annealing, process, as revealed 
 by the appearance of the fracture. The metal as originally cast 
 will have the so-called crystalline fracture of white cast iron, with 
 perhaps a little mottle. This fracture should be changed to a 
 black heart after the annealing process, and the metal will have 
 good strength and ductility. If, however, the composition of the 
 metal was wrong in the first instance, it may quite resist the 
 annealing process and precipitate little or none of its combined^ 
 carbon. Or else, if the silicon were too high, the castings may 
 have already precipitated too much graphite before the annealing 
 process begins, and the deposition of temper carbon may take 
 place so readily during annealing that the castings will be mushy 
 and lacking in strength. ^ 
 
 ^ The term combined is used here to represent carbon in solution or any chemical com- 
 bination with the metal. 
 
 ' Iron too high in silicon will precipitate too much graphite during its first cooling, and 
 is technically known as low iron; on the other hand, iron with very little silicon is known 
 as high iron. 
 
MALLEABLE CAST IRON 355 
 
 If, however, the composition of the metal and the melting 
 conditions were correct, a good grade of black heart casting may 
 be produced by a proper anneal, but, should the temperature of 
 annealing be too high, or be continued too long, the black heart 
 fracture may change to a steely fracturej showing an iron in 
 which the precipitated carbon has returned to the combined form. 
 Nor is this all; if the annealed castings are cooled too fast after 
 the anneal, or if the black heart castings, after being successfully 
 made, are reheated and hammered, 'or reheated and quenched in 
 water, the black heart fracture will change to steely. Even the 
 heating to which the finished castings are subjected when they 
 are dipped in spelter for a galvanizing process and then quenched 
 in water, will produce a steely fracture with consequent brittle- 
 ness if the composition of the metal is not within very narrow 
 limits. There is but one remedy for castings having a steely 
 fracture and that is to re-anneal them and produce the so-called 
 Reaumur or white heart castings. In other words, white heart 
 malleable iron is the end of a series of changes through which 
 the carbon passes during a long annealing process. 
 
 Theory of the Black Heart Process. — By referring to the diagram 
 of Fig. 254, suggested by Upton, it will be observed that graphite 
 should normally separate from irons high in carbon during and 
 shortly after solidification. This reaction is avoided in making 
 malleable castings, by cooling the metal rapidly through this 
 range of temperature. The first effect of the annealing process, 
 however, is a slow readjustment of conditions by the occurrence 
 of the reaction which was previously suppressed. That is to say, 
 the carbon is now precipitated within the solid metal. The 
 temperature at which this reaction occurs during the annealing 
 is not as high as the normal temperature for it, but is in the 
 region where cementite (FogC) normally should be produced 
 according to the diagram. If, therefore, the annealing is con- 
 tinued too long, the carbon precipitation is followed by its recom- 
 bination to form the carbide of iron. As the metal is then 
 slowly cooled, this carbide of iron separates and we get the 
 steely fracture, which is characteristic of over-annealed malleable 
 castings. 
 
 If, on the other hand, we have obtained a good black heart 
 casting and this is reheated into the region where cementite 
 normally forms, and suddenly cooled, we produce a metal which 
 consists of cementite, mostly dissolved in an iron matrix. This 
 
356 THE METALLURGY OF IRON AND STEEL 
 
 reasoning fails to explain why tho rapid cooling of castings, 
 after a proper annealing, will sometimes produce the steely 
 fracture, and so far as I know, this occasional result yet lacks 
 an explanation. 
 
 The reason why cast iron too high in silicon precipitates its 
 carbon too readily and gives a mushy iron is readily explained, 
 since it is known that the presence of silicon will promote the 
 separation of graphite, and especially will make the iron open- 
 grained and cause graphite to separate in larger flakes. 
 
 Theory of the RSaumur Process, — In the original R6aumur 
 process the castings are packed in iron oxide during the anneal, 
 and this operation is carried on at a temperature of about 800 to 
 875° C. (1472 to 1607° F.) for a period of four or five days after 
 reaching full temperature. In the first part of this annealing 
 process conditions are produced which tend to result in black 
 heart castings; then the next step results in conditions that 
 would produce the steely fracture; still the anneal -is protracted 
 at the high temperature and, at the end of the 100 hours or so, 
 we have castings in which the white rint of black heart metal 
 extends to the very center — the castings being only 1/2 inch or 
 less in thickness. This white rim, as already explained, is iron 
 from which the carbon has been actually removed bodily, and this 
 decarburization results from a reaction between the iron oxide in 
 contact with the skin of the casting and the carbon in it. Whether 
 the carbon migrates to the surface and is there oxidized by the 
 oxide : 
 
 3C+Fe203 = 3CO + 2Fe, 
 
 or whether oxygen penetrates the metal through its pores in the 
 form of CO gas, or penetrates chemically by forming iron oxide 
 which is acted upon by the carbon: 
 
 C+FeO=00+Fe, 
 is a matter on which authorities cannot agree. The first theory 
 is the one which seems to have the preponderance of evidence in 
 its favor. That carbon will migrate in solution is proven in the 
 cementation processes and, at the annealing temperature, the car- 
 bon is normally all in solution, whether it be as free carbon, or 
 as carbide. Even the white rim of an ordinary black heart 
 casting contains some carbon in solution. 
 
XIV 
 
 " THE HEAT TREATMENT OF STEEL 
 
 We have already discussed the heat treatment of cast iron 
 under other heads, namely: (1) the rapid cooling from the molten 
 state, or "chilling/' and (2) the annealing of malleable cast iron. 
 Heat treatment is of much greater importance in connection with 
 steel, because nearly 99 per cent, of all the steel made is heated 
 either for the purpose of bringing it into the mobile condition in 
 which it can be readily wrought, or for annealing. Indeed, the 
 great majority of steel is heated several times; and some steel is 
 subjected to two or three different kinds of heat treatment. 
 
 Improper Heating of Steel 
 
 Overheating. — If steel be heated to a high temperature, say 
 1100° C. (2010° F.), and then cooled (either slowly or rapidly) 
 without being subjected to strain, it will be "coarse-grained" as 
 it is called, that is, its crystals will be relatively large in size. 
 This can be readily seen by breaking it and examining the frac- 
 ture, which will be bright and sparkling if the crystals are coarse, 
 or dull-looking and fine-grained if they are small (see Fig. 307) . 
 The bright fracture is technically called "crystalline" or "fiery/' 
 while the fine-grained one is called "silky" or "sappy." The 
 size of the crystals may also be learned with great accuracy by 
 means of the microscope (see Figs. 300 to 305 and 307). Now, 
 if the steel which was coarse-grained after heating to 1100° be 
 heated instead to 1200°, the crystals will be still larger in size; if 
 heated to 1300° they will be larger still, and so on. The size of 
 the crystals will depend first upon which of these high tempera- 
 tures it was heated to, and second upon the amount of carbon 
 it contains. Low-carbon steel is normally larger in crystal-size 
 than high-carbon steel. 
 
 Even the best quality of steel, if rendered coarse-grained by 
 "overheating," will suffer in its valuable properties, and may 
 become quite unfit for use. Medium-, and high-carbon steel will 
 lose both strength and ductility; low-carbon steel will lose 
 
 357 
 
358 THE METALLURGY OF IRON AND STEEL 
 
 strength even up to 50 per cent. oLthe original, but does not 
 seem to be materially damaged in ductility unless the overheating 
 is continued for a long 1;ime or at a very high temperature. 
 
 Cure for Overheating, — Let our first example be steel con- 
 taining 0.9 per cent, carbon, that is, steel consisting entirely of 
 pearlite. If this be heated from some point below the line P SK 
 in Fig. 252, page 295, to some point above that line, a new crystal- 
 lization will occur and will largely obliterate previous crystalliza- 
 tion. It seems as if dissolving the ferrite and cementite in each 
 other produces forces which obliterate almost all existing crystal- 
 line forms. So, if this particular steel has been made coarse- 
 grained by overheating, we may make that grain fine again by 
 reheating the steel from below the line P ^ if to just above it. 
 This process is known as "restoring," or, by some writers, "re- 
 fining" the steel. It is an operation which should be thoroughly 
 understood by every metallurgist and engineer. When we reheat 
 the steel we must be careful not to go to a high temperature again, 
 for a new crystal-size is born at the line P S K, and the crystals 
 grow with every increase in temperature. The researches of 
 Professors Howe and Sauveur^ indicate that the size of the crystals 
 is almost directly proportional to the temperature reached above 
 the line P S K. If, therefore, we barely cross the line, we will 
 obtain the smallest grain-size that the steel is capable of (see 
 Fig. 307). 
 
 The cure for coarse crystallization in steel with less than 0.9 
 per cent, carbon is to reheat it from below the line P S Kto above 
 the line G ^, at which the last of the ferrite goes into solution. 
 That is to say, the correct temperature for restoring the grain- 
 size will depend upon the amount of carbon in the steel; low- 
 carbon steel must be heated to nearly 900° C. (1650° F.); 0.4 per 
 cent, carbon steel must be heated to nearly 800° C. (1470° F.); 
 and so on.^ We can never get as small a grain-size in steel with 
 less than 0.9 per cent, carbon as we can in that which is exactly 
 0.9 per cent, carbon^ because a new grain-size begins to grow 
 after we have crossed the line P S K, and yet we cannot entirely 
 eliminate the old grain-size until we cross the line GO S, Where 
 the lines G S and P S K are near together (say, with 0.7 per 
 cent, carbon), the new grain-size does not have much chance to 
 
 1 See page 246 of No. 1, page S. 
 
 ^ It is to be remembered that the changes indicated by the lines in Fig. 252 occur at a 
 higher temperature on heating than on cooling (see page 290) ; so it is well to heat the steel 
 about 25° C. higher than the points on those lines. 
 
THE HEAT TREATMENT OF STEEL 
 
 359 
 
 FIG. 300.~NO. lA. STEEL OF 0.05 
 PER CENT. CARBON ROLLED. 
 
 Magnified 40 diameters. 
 
 FIG. 301.— NO. IB. STEEL OF 0.50 
 
 PER CENT. CARBON ROLLED. 
 
 Magnified 60 diameters. 
 
 FIG. 302.— NO. 2A. SAMEASNO.IA 
 
 OVERHEATED TO 1420° C. (2588° F.) 
 
 Magnified 60 diameters. 
 
 FIG. 303.— NO. 2B. SAME AS NO. IB. 
 
 OVERHEATED TO 1420° C. (2588° F.) 
 
 Magnified 60 diameters. 
 
 FIG. 304.— NO. 3A. SAME AS NO. 2A. FIG. 305.— NO. 3B. SAME AS NO. 2B. 
 REHEATED SLIGHTLY ABOVE ACg. REHEATED SLIGHTLY ABOVE AC2-3. 
 
 Magnified 40 diameters. Magnified 60 diameters. 
 
 Series A by F. C. Wallower in the Metallographic Laboratory of Columbia University, 
 Department of Metallurgy. 
 
 Series B by G. Rocour in the Metallographio Laboratory of Columbia University, De- 
 partment of Metallurgy. 
 
360 THE METALLURGY OF IRON AND STEEL 
 
 grow before the restoration is complete, and therefore we may 
 obtain steel with a pretty small grain; but where they are far 
 apart (as in the low-carbon steels) the restoration can never be 
 very thorough, because we have to go so far above P S K to ob- 
 literate the old grain-size that the new grain-size will have at- 
 tained ample proportions. But the evidence seems to show that 
 the best net result is obtained by going just above the line GO S 
 in all cases. 
 
 In the case of steel with more than 0.9 per cent, carbon a some- 
 what similar condition exists: we must reheat the steel above 
 the line aS a in order to produce complete elimination of the previ- 
 ous grain-size but a new grain-size begins to grow from the cross- 
 ing of the line P S K. But here we disregard the line S a, and 
 restore our steels in every case by reheating them over the line 
 P S Kj just as in the case of pure pearlite. The reason for this is 
 that the lines S a and P S K diverge so rapidly that we have to 
 heat very far above the line P S K before we cross S a,- and there- 
 fore the new grain-size has grown greatly. Furthermore, the 
 only object of heating above the line >S a is to take the excess 
 cementite into solution; for the ferrite and cementite in the 
 pearlite all went into solution as sood as we crossed the line 
 P S Kj but the amount of excess cementite is always small in 
 proportion, and therefore in its influence on restoration. Even 
 with steel containing 2 per cent, of carbon the excess cementite 
 is only 16 per cent. This is different from the low-carbon steels, 
 where the excess ferrite will be usu^ly over 80 per cent. 
 
 Evidence of Overheating. — A piece of steel may be heated 
 many times above the lineP SK and cooled again, but obviously 
 only the latest heating will leave its impression on the structure, 
 because each crossing of the line on the way up removes the effect 
 of previous heat treatment.^ The relation between the size of 
 the crystals and the temperature above P S K is so constant that 
 we may determine what this temperature was from the analysis 
 of the steel and an examination of the grain. To do this it is 
 usually necessary to "get a piece of steel of the same analysis, 
 heat different pieces of it to various temperatures, and compare 
 
 1 This is only true in a qualified sense, in that the previous overheating must not have 
 been very close to the melting-point. We shall discuss this point under the head of ' ' Burn- 
 ing." Indeed, even where burning has not occurred, a skillful microscopist may some- 
 times discern the effect of over-heating after the steel has been restored by reheating; 
 because, although the crystals are all small, they are arranged in groups which show the 
 form of the previous large crystals. 
 
THE HEAT TREATMENT OF STEEL 361 
 
 (see page 367). The analysis must be approximately the same 
 not only in carbon but also in phosphorus, sulphur, silicon, and 
 manganese, as well as in any alloying elements, if present, such as 
 nickel, chromium, tungsten, etc., because all of them have an 
 effect upon the size of grain and also upon the change in size of 
 grain by overheating. Generally, it is not important to know 
 the exact temperature of overheating, but only whether or not 
 overheating in some degree has occurred; and this is not difficult 
 to prove, because almost all who use steel are familiar with the 
 normal fracture of steels of different carbon and can tell at a 
 glance if the grain is large; those who are not so familiar with its 
 appearance may easily become so. The grain being large is proof 
 that overheating was the cause, provided chemical analysis shows 
 everything about normal, especially phosphorus and silicon. 
 
 Steel members of bridges or other structures sometimes break 
 and disclose a crystalline fracture which is often attributed to the 
 effect of vibration. The same thing occurs with points or shanks 
 of rock drills and similar implements. It is the more general 
 opinion among metallurgists that the crystalline fracture in all 
 these cases is due to faulty heat treatment during manufacture, 
 and especially to finishing the forging or rolling while the tempera- 
 ture is still too high. The manufacturers of steel like to maintain 
 the opposite opinion, for obvious reasons, but I do not know 
 of there ever having been any reliable proof offered that vibra- 
 tion had caused, or is capable of causing, large-sized grain in 
 steel. It may be possible, but the more we learn about the sub- 
 ject the more we are inclined to believe that improper manu- 
 facture is the cause, and that the grain was large before the steel 
 was put in service, although its nature was not disclosed until the 
 break occurred. 
 
 Mechanical Cure for Overheating. — When steel is to be rolled 
 or forged it is frequently heated to a temperature of 1100° to 
 * 1350° C. (2010° to 2460° F.), and it might be thought that this 
 treatment would seriously damage it. So it would, but for the 
 circumstance that the subsequent mechanical pressui-e upon the 
 metal breaks down the crystals and reduces them to a small size. 
 The result is that the final size of the crystals is dependent upon 
 the temperature of the material at the finish of the mechanical 
 'operation. In other words, steel finished at 900° C. (1650° F.) 
 has a finer structure than the same steel finished at 1100° C. 
 (2010° F.). We do not feel warranted in stating numerically the 
 
362 THE METALLURGY OF IRON AND STEEL 
 
 exact relation between the finishing temperature and the grain- 
 size, as we have not yet sufficient evidence, but several rules af- 
 fecting the final size of grain seem to be virtually established: (1) 
 It is more advantageous to have the mechanical work applied 
 continuously from the highest temperature employed down to 
 the finishing temperature, rather than to have long waits during 
 which the steel cools; and especially is this true when the amount 
 of work put upon the metal at the lower temperature is small. 
 In other words, if the steel is formed roughly to shape and size 
 at a high heat, is then allowed to cool, and a little work is done 
 upon it at the lower temperature, the grain will not be its best. 
 
 (2) It is best for the metal to be worked by several passes through 
 the rolls, or many blows of the hammer, rather than to effect the 
 same amount of reduction by a lesser number of heavy drafts. 
 
 (3) The greater the amount of reduction the better; that is, to 
 work a large piece down to the desired article gives a better 
 structure. (4) The best temperature at which to finish the work 
 is probably upon, or slightly below, the lines G-O-S or S-K in Fig. 
 252, page 295. 
 
 Action in Rolling. — The exact crystalline action that takes 
 place under mechanical treatment in not definitely known. In 
 the case of rolling Professor Howe has tentatively assumed the 
 conditions graphically shown in Fig. 306,^ in which the line D G 
 represents the size of grain at the different temperatures. At 
 1400° C. (2550° F.) the grain-size is represented by the distance 
 of the line from the axis 0-0. On the first passage through the 
 rolls the grains are crushed to a very small size, but on emerging 
 again they grow very rapidly. Meanwhile, however, the metal 
 has been cooled, and this fact, as well as the inability of the grains 
 to grow instantly, causes the new size of grain to be smaller than 
 before. Therefore, each passage through the rolls renders the 
 crystals smaller in size, the final size depending upon the tempera- 
 ture and the amount of pressure in ,the last pass. The only 
 abnormal assumption in this argument is that the crystals grow 
 rapidly after the crushing, whereas we know that when steel is 
 heated to any of these high temperatures, the growth is relatively 
 slow. This objection is not strong enough alone to refute the 
 theory, but other hypotheses may be advanced for those who 
 require further explanation. -For example, it may be supposed 
 that the steel is so mobile at 'the very high temperatures that it- 
 
 1 Page 263 of No. 1. '' 
 
THE HEAT TREATMENT OF STEEL 
 
 363 
 
 yields to distortion, not altogether by the crushing of the crystals, 
 but by the sliding of the crystals past one another; as the tem- 
 perature becomes lower, however, the mobility of the mass be- 
 comes less, and less sliding is possible, so that more crushing of 
 the crystals takes place, 
 o 
 
 ■A = Diameter of grain at tTus finishing temperature 
 
 33°F 12 3 4 5 
 
 Dia'meter of G-rain 
 FIG. 306.— From Howe, "Iron, Steel and other Alloys." 
 
 Finishing Temperatures. — William Campbell has studied the 
 finishing temperature ol steel containing 0.5 per cent, carbon 
 and finds that the very best' qualities are produced in the steel 
 if mechanical work is ended just at the time when ferrite begins 
 to separate from solid solution, that is to say, just when the steel 
 is below the line G-O-S in Fig. 252, page 295. Work. below 
 that temperature greatly increases the brittleness of the material. 
 
364 THE METALLURGY OF IRON AND STEEL 
 
 while finishing the work at a higher heat results in lower strength. 
 Upon the evidence at hand, we may tentatively assume like con- 
 ditions for steels of any carbon, and expect the best results if 
 mechanical work is ended when the steel is at a temperature 
 which brings it exactly upon the line G-O-S or S-K, but reserv- 
 ing, perhaps, the right to change this statement" slightly when 
 more data are obtained. 
 
 Welding. — This brings us to the subject of welding, or the 
 joining of two pieces of wrought iron or steel by pressing or ham- 
 mering them together while at a very high temperature. In this 
 way a joint may be made which cannot be seen by the eye unless 
 the steel is polished and etched with acid, which usually develops 
 the junction line very clearly. The exact temperature of welding 
 is not known, but probably it is very near the melting-point, 
 where the steel is in a soft and almost pasty condition. Low- 
 carbon steel welds most easily; moreover, all impurities, especially 
 silicon and sulphur, reduce weldability. The procedure in weld- 
 ing is very simple, and consists in heating the two pieces that are 
 to be welded to a high temperature, dissolving off the iron oxide, 
 and then pressing the two pieces together forcibly. The dis- 
 solving off of the oxide is usually accomplished by rubbing the 
 metal in some flux, such as borax. At the present time various 
 patented " welding plates" are sold. These consist of thin plates 
 of flux which are put between the two pieces to be welded and 
 so get rid of the oxide, the pieces being hammered together 
 with the plate between them. 
 
 In the actual manipulation for welding -the two pieces that are 
 to be joined together are usually "upset," or in some way en- 
 larged in size, so that after the junction the part of the bar right 
 at the weld is larger in size than the remainder. This' part is 
 then hammered continuously until the metal is at a red heat, 
 the object being to break up the coarse crystals produced by the 
 high temperature, and, by having a low "finishing temperature," 
 to obtain a small grain-size. With proper welding this object 
 will be attained so far as the metal immediately adjacent to the 
 weld is concerned, but there is always a spot within six inches 
 or so of the weld whigh must necessarily have been overheated 
 without subsequently receiving mechanical treatment, i.e., " ham- 
 mer refining," down to the proper finishing temperature. Thus 
 it is that most welded pieces break at a point not far from the 
 junction and under a strain much less than the original strength 
 
THE HEAT TREATMENT OF STEEL 365 
 
 of the bar. Blacksmiths and experienced welders are wont to 
 declare that if a welded bar does not break in the weld itself, then 
 it must be as strong as the original metal. However, this is by 
 no means true. In a welding test carried on with great care in 
 this country by skillful and experienced welders who were placed 
 upon their mettle,^ the strength and elastic limit of the welded 
 bar was almost never as great as the original bar, and in some 
 cases was less than half. In ductility even worse results were 
 obtained. In a similar test carried on at the Royal Prussian 
 Testing Institute the average strength of welded bars of medium 
 steel was only 58 per cent, of the original, that of softer steel only 
 71 per cent., and of puddled iron only 81 per cent., while the poor- 
 est results were only 23, 33, and 62 per cent, respectively. It 
 was seen that bad crystallization adjacent to the weld was the 
 cause of the damage. 
 
 Welded steel and iron bars should therefore always be re- 
 heated after cooling to a temperature just above the line G-O-S, 
 in order to restore by heating the grain-size of all parts. 
 
 Burning, — In the vernacular of the trade, all overheated 
 steel is termed "burnt," but this is not correct usage, because 
 true burning takes place only when the overheating is most abu- 
 sive, and, indeed, when the metal is heated almost to its melting- 
 point. It is probable that steel is burnt only when it is heated 
 above the line A-a in Fig. 252, page 295. Alfred Stansfield has 
 studied this question very ably,^ and distinguishes three stages 
 of burning. The first stage is reached when the steel barely 
 crosses the line A-a, that is, when the first drops of melted metal 
 begin to form in the interior of the mass. They segregate to the 
 joints between the crystals and cause weakness. Stansfield 
 thinks that steel burned only to this stage may be restored by 
 reheating it first to a high temperature, cooling, and then heating 
 again to a temperaturie j ust above the lines G-O-S-K. The second 
 stage in burning is reached when these liquid drops segregate as 
 'far as the exterior and leave behind a cavity filled with gas. 
 Stansfield thinks that steel burned to this stage might be re- 
 stored by combined reheating and forging. As a matter of 
 safety, however, I believe it would be well to remelt all such ma- 
 terial; in other words, send it to the scrap pile. The third and 
 last stage of burning is reached when gas collects in the interior 
 
 » See pages 401 to 406 of No. 2. 
 ^ See No. 143, 
 
366 
 
 THE METALLURGY OF IRON AND STEEL 
 
 of the metal under sufficient pressure to break througk the skin 
 and project -liquid steel, which produces the well-known scintil- 
 lating effect at this temperature. Into the openings formed by 
 these minute explosions air enters and oxidizes the interior. 
 There can be no remedying of steel which has been burned to this 
 extent. 
 
 Metcalf Test. — A very interesting experiment is the "Met- 
 calf test/' originated by William Metcalf.^ It is best performed 
 upon a bar of high-carbon steel, because this material shows the 
 
 FIG. 307-308.— METCALF TEST. FRACTURES OF STEEL CONTAINING ONE PER 
 CENT. OF CARBON. 
 
 differences in structure so readily to the eye. A bar of steel about 
 12 in. long is notched with a hacksaw or chisel at intervals of 
 an inch. One end is then placed in a fire and heated to a tem- 
 perature at which it scintillates, while the other end is at a black 
 heat. Then it is removed and cooled. It is immaterial whether 
 the cooling be rapid or slow, but time may be saved by plunging 
 it into water. It is then broken at every notch, and an exam- 
 ination of the fractures will show a very large size of crystals at 
 the end which is burnt, gradually decreasing until a fine and 
 silky appearance is presented where the metal was exactly at 
 
 1 See pp. d05, 406 (especially 406) of No. 116. 
 
THE HEAT TREATMENT OF STEEL 367 
 
 the temperature of restoration, while beyond that point the 
 fracture will b^ the same as that of the original bar. In case 
 this test is made upon low-carbon steel it should not be notched 
 before treatment, but afterwards it should be cut apart with a 
 hacksaw at intervals of an inch, and then polished and. examined 
 under the microscope, because soft steel will not break without 
 bending, and this bending destroys the indications of the frac- 
 ture. This Metcalf test is very serviceable in case we desire to 
 compare one steel, suspected of being overheated, with another 
 of like analysis, to determine the degree of overheating. 
 
 Castings do not Burn.— It might be thought that every steel 
 casting would suffer the injuries due to burning because it is 
 cooled through the space between the lines A-B and A-a. Such 
 injury, however, does not ordinarily take place, and this fortu- 
 nate circumstance is explained in part by each of three differ- 
 ences existing between the heating and cooling of steel: (1) When 
 steel is heated into the area where burning takes place, it is sub- 
 jected longer to the burning temperature, because it generally 
 takes longer to heat steel than to cool it. (2) When steel is 
 being heated, the heat is traveling inward from the outside, and 
 therefo^fe all parts are expanding, and there is some opportunity 
 for the crystals, to draw apart and form cavities. On the other 
 ' hand, when it is cooling from the molten state, the outside layers 
 are the cooler, and tend to contract upon the interior and hold the 
 crystals more firmly together, as well as preventing drops oozing 
 out. (3) When steel is cooling from a molten state, it is con- 
 stantly giving off from solution hydrogen and other deoxidizing 
 gases, which are soluble in it while liquid, and these gases prevent 
 the oxidation of the crystal faces by the percolation of air into 
 the interior. 
 
 Ingotism. — I have already discussed ingotism and said that the 
 crystals in cast steel are larger than those of rolled steel, due to 
 growth while the metal is at a high temperature, and I have 
 stated that sometimes these crystals are very large, because the 
 conditions of casting cause the steel to occupy a longer time in 
 cooling from the liquid state down to a black heat. It is prob- 
 able that ingots and castings do not show the effects of ingotism 
 to any marked extent unless they are a long time above 1100° C. 
 (2010° F.). In case these coarse crystals do form, they may be 
 restored to some extent by reheating the casting to a point just 
 above the line G-O-S. 
 
368 THE METALLURGY OF IRON AND STEEL 
 
 Stead's Brittleness. — In addition to the damage caused by 
 overheating, steel very low in carbon (say under 0.15 per cent.) 
 is subject to another and peculiar danger, for if this soft steel 
 be held for a very long time at temperatures between 500° and 
 750° C. (930° and 1380° F.); the crystals become enormous and 
 the steel loses a large part of its strength and ductility. Fortu- 
 nately it takes a very long time, in fact days, to produce this 
 effect to any alarming degree, so that it is not liable to occur, 
 even through carelessness, during manufacture or mechanical 
 treatment. But steel is sometimes placed in positions where it 
 may suffer this injury, for example, in the case of the tie-rods of 
 furnaces, supports for boilers, etc., so that the danger should be 
 borne in mind by all engineers and users of steel. I recall an 
 instance where the breaking of a piece of chain that supported 
 one side of a 50- ton open-hearth ladle caused a loss of life under 
 the most horrifying conditions, due to the fact that the wrought- 
 iron chain had been heated up many times to a temperature above 
 500° C. (930° F.), and had finally reached a condition of coarse 
 crystallization, so that it was unable to bear the strain upon it 
 when the ladle was full of metal. 
 
 This phenomenon of coarse crystallization in low-carbon steel 
 is known as "Stead's brittleness," after J. E. Stead, who has 
 explained its cause. The effect seems to begin at a temperature 
 of about 500° C, and proceeds more and more rapidly with an 
 increase in temperature until we reach 750° C, above which no 
 growth seems to take place. The damage may be repaired 
 completely by heating the steel just above the line G-O. In other 
 words, the remedy for coarse crystallization in this case is the 
 same as that for coarse crystallization due to overheating, and 
 all steel which is placed in positions where it is liable to reach 
 these temperatures frequently, should be restored at intervals 
 of a week or a month, or as often as may be necessary. 
 
 Hardening of Steel 
 
 If steel be raised to a bright-red heat and then rapidly cooled, 
 as, for example, by plunging it into water, it becojnes very much 
 harder and at the same time stronger and more brittle. Gne cir- 
 cumstance is absolutely necessary to produce the increase in 
 hardness, namely, that the temperature from which rapid cooling 
 takes place shall be above the critical temperature of the steel. 
 
THE HEAT TREATMENT OF STEEL 369 
 
 Take, for example, steel containing 0.9 per cent, carbon; we may 
 heat this ever so little below the point S in Fig. 252, page 295, 
 and no increase in hardness will t^ke place, even though we cool 
 with extreme rapidity. On the other hand, if we cool the same 
 steel rapidly from ever so little above the point ,aS, it will be hard 
 enough to scratch glass and brittle enough to fly into pieces under 
 a blow of the hammer. This is the maximum practical hardness 
 which can be obtained in this steel for if we quench it at a higher 
 temperature, the only result of importance is to damage it by 
 increase in grain-size. In case we have less than 0.9 per cent, 
 carbon in our steel, the best temperature for hardening is just 
 above the line.G-0~Sj because that gives the maximum hardness 
 and also the best grain-size.^ The best temperature from which 
 to harden steel with more than 0.9 per cent, carbon is just above 
 the line S-K, because that gives the best grain structure, although 
 it is true that greater hardness is obtained if we cool from above 
 the line S-a. 
 
 Carbon and Hardness. — The hardness of steel increases with 
 every increase of carbon. This applies to the hardness of steel 
 in its natural state, and still more appropriately to its harness 
 after heat treatment. Although iron free from carbon is hardened 
 by rapidly cooling from above the point Ac^ (760°C. = 1400° 
 F.), and a little more so when rapidly cooled from above Ac^ 
 (900° C. = 1650° F.), yet this degree of hardness is so slight as to 
 be perceptible only by means of delicate laboratory tests. With 
 0.25 per cent, carbon the. hardness begins to be perceptible by 
 crude tests, but it is only when we get above 0.75 per cent, carbon 
 that ordinary steel acquires sufficient hardness for the process 
 to be used commercially — ^for example, for springs, saws, etc. 
 Metal-cutting tools are usually made of steel containing 1 per 
 cent, or so of carbon, while very hard implements, such as files, 
 etc., wiU'contain 1.5 per cent, or slightly more. 
 
 Rate of Cooling and Hardness. — The degree of hardness of 
 steel also varies with the speed of cooling from above the critical 
 range of temperature. When the cooling is very slow, as, for 
 example, when it takes several days to cool, the steel will be as 
 soft as it is possible to make it. When it is cooled by being taken 
 out of the furnace and suspended in the air, or thrown upon a sand 
 floor, it will still be relatively soft. ' When cooling is still more 
 rapid, as, for example, when it is taken out of the furnace at a 
 bright-red heat and plunged into a heavy oil with a low conduct- 
 
 24 
 
370 
 
 THE METALLURGY OF IRON AND STEEL 
 
 ing power for heat, it becomes quite hard and springy, provided 
 its carbon is in the neighborhood of 0.8 per cent, or above. 
 Quenching in a thin oil from the same temperature makes it still 
 harder. Quenching in water makes it harder stiJI; and so on, the 
 degree of hardness increasing as we quench in liquids which take 
 the heat away from it faster and faster, such as ice-water, ice- 
 brine, ice-sodium chloride solution, and mercury near its freezing- 
 point (-39°C.=-38°F.). 
 
 THE EFFECT OF CARBON ON THE MECHANICAL PROPERTIES 
 OF STEEL AFTER HARDENING IS SHOWN BELOW: ^ 
 
 
 Tenaile s 
 
 trength in soft state 
 
 Tensile strength when hardened. 
 
 Carbon 
 
 
 
 
 
 
 
 per cent. 
 
 Ultimate 
 lbs. per 
 sq.in. 
 
 Elastic 
 limit lbs. 
 per sq. in. 
 
 Elonga- 
 tion in 
 2"-per 
 cent. 
 
 Ultimate 
 lbs. per 
 sq. in. 
 
 Elastic 
 limit lbs. per 
 sq. in. 
 
 Elongation 
 in 2" -per 
 cent. 
 
 0.10 
 
 60.300 
 
 36,300 
 
 29 
 
 66,4002 
 
 40,300 
 
 24 
 
 0.14 
 
 61.500 
 
 35,200 
 
 27 
 
 73,1002 
 
 39,600 
 
 22 
 
 0.23 
 
 66,600 
 
 41.200 
 
 26 
 
 99,4002 
 
 54,000 
 
 14 
 
 0.52 
 
 97,800 
 
 52,600 
 
 20 
 
 132,100 
 
 81,400 
 
 9 
 
 0.60 
 
 116,400 
 
 66,500 
 
 14 
 
 153.400 
 
 102,100 
 
 4 
 
 0.72 
 
 130,700 
 
 75,800 
 
 9 
 
 180.100 
 
 105.500 
 
 
 
 Theories of Hardening. — Gne essential feature of hardening is 
 that the steel must be heated to a temperature above the line 
 P-S-Kj that is to say, to a point where at least some of the solid 
 solution of iron and carbon is formed. There are several different 
 theories to explain the hardness produced by rapid cooling from 
 this point, the two most important being the "carbon theory" 
 and the " allotropic theory." Both of these theories depend upon 
 the following line of reasoning: At temperatures above the critical 
 range the molecules of steel are in a hard condition.^ As they 
 cool from this point and cross the critical range of temperature,the 
 molecules change from the hard state to a soft state, but this 
 
 1 See No. 149. 
 
 2 The increase in strength in these soft steels is due in large part to the finier size of grain 
 produced by rapid cooling through the critical range. 
 
 ^ It is difficult for some to understand how molecules of steel can be in a hard condition 
 at a temperature at which we know that the mass as a whole is soft and mobile; but this 
 can be explained by the following comparison: A ball of wet sand and clay is soft and mo- 
 bile as a whole, because the particles move by each other readily and the mass changes 
 its shape under pressure; yet the individual particles composing this ma^s are many of them 
 hard enough to scratch glass with ease. 
 
THE HEAT TREATMENT OF STEEL 371 
 
 change is not rapid and requires time for its accomplishment; 
 rapid cooling does not afford the necessary time, and so perpetu- 
 ates the hard state of the molecules. This line of reasoning 
 leaves only one point in doubt, namely, what causes the molecules 
 to be hard when the steel is above the critical temperature, i.e., 
 when iron and carbon are dissolved in each other? 
 
 The Carbon Theory. — The carbon theory assumes that the 
 hardness of steel is due altogether to the carbon dissolved in it, 
 and in evidence its advocates point to the extreme hardness of 
 one form of carbon — the diamond. This theory has the ad- 
 vantage of simplicity, and has in its favor the fact that Jhe hard- 
 ness varies almost directly with the amount of carbon. Against 
 the theory, it is urged that the amount of carbon is really too 
 small to produce such a great degree of hardness in the whole mass 
 of metal. Furthermore, although carbon is found in many metals 
 it does not confer hardness on any of them except iron. 
 
 The Allotropic Theory. — We have already learned that in the 
 solid solution the iron is present in the gamma allotropic form, 
 and the majority of metallurgists attribute the hardness of steel 
 to the allotropic form, and deny that carbon is the sole factor. 
 It has been shown, by very delicate laboratory tests on iron prac- 
 tically free from carbon, that the gamma and beta allotropic 
 modifications are harder than the alpha modification. These did 
 not show how great was the increase of hardness of one form over 
 another, because it was not possible to cool the iron fast enough to 
 prev^ent it changing back in part to the alpha form. In the 
 absence of carbon the change from gamma to beta and from beta 
 to alpha is very rapid, so that cooling has to be almost instantan- 
 eous in order to prevent it, and this, of course, is impossible. The 
 " allotropists " explain the greater hardness of high-carbon steel 
 as compared with low-carbon steel upon the basis that the pres- 
 ence of carbon makes the change from gamma to beta and to 
 alpha iron slower, and therefore enables more of the iron to be 
 retained in the gamma and beta forms by the rapid cooling. 
 
 An additional argument in favor of the allotropic theory is 
 that when steel cools slowly through the critical range, it loses 
 its hardness slightly before the carbon comes out of solution. 
 This would indicate that the allotropic change took place before 
 the carbon change, and that the allotropic change was the cause 
 of hardness. What is true of the loss of hardness is also true to 
 the other physical changes which take place at the same time. 
 
372 THE METALLURGY OF IRON AND STEEL 
 
 That is to say, the steel regains its magnetism, decreases in electric 
 resistance, and increases in thermo-electric power in a large part 
 before much carbon is separated from the solution. 
 
 The Compromise Theory. — Several theories have been ad- 
 vanced which are a compromise between that of " the carbonists " 
 and that of ''the allotropists." The simplest of these, and the 
 theory now most generally accepted, is that the hardness of the 
 molecules of steel above the critical range is due partly to the 
 allotropic form in whic^ the iron exists, and partly to the f aict that 
 we have a solid solution of iron and carbon.^ In other words, 
 the hardness is due to the fact that we have a solid solution of 
 carbon in an allotropic form of iron. 
 
 Tempering. — Hardened steel is too brittle to be used without 
 some degree of tempering, except for a small variety of purposes, 
 such as the points of armor-piercing projectiles, the face of armor 
 plate, etc. In order to understand just what tempering does, 
 let us consider the exact condition of hardened steel: it is in a 
 hard and brittle condition which is not natural to it at atmos- 
 pheric temperatures, but which has been brought down with it 
 from a higher temperature by means of rapid cooling. Theo- 
 retically, when the temperature fell below 690° C. (1272° F.), 
 the molecules of steel should have changed over to the soft form. 
 Their hard condition is not in equilibrium at the lower tempera- 
 ture, in the same sense that ice is not in equilibrium in hot weather- 
 Why, then, does not the steel change back into the soft form? 
 Ice, if given time enough, will all change into water when the 
 temperature is above 0° C. (32° F.). The reason the change does 
 not take place, in the steel after we have cooled it to the atmos- 
 pheric temperature is that the mass as a whole becomes too rigid 
 and immobile at the lower temperature to permit any alteration 
 in its molecules to take place. 
 
 However, it is only necessary to decrease this rigidity in order 
 to permit a slight change. For example, if a piece of hardened 
 steel be kept in boiling water for some days it will lose a part of its 
 hardness; if it be heated a little more, it will lose more hardness 
 and lose it much more quickly. Each loss in hardness is accom- 
 
 ^ Where I speak of a solution of carbon and iron. I intend to include under this also the 
 solution of iron and a carbide of iron. That is to say, we have two substances, iron and 
 carbon, and they are dissolved in each other. It may be that the carbon is united with 
 part of the iron to form a carbide, and then that this carbide is dissolved in the rest of the 
 iron; but I use the term "solid solution of iron and carbon" to cover either this condition 
 or that of elemental carbon dissolved in iron. 
 
THE HEAT TREATMENT OF STEEL 373 
 
 panied by a loss in brittleness as well. If it be heated to about 
 200° C. (392*='' F.), quite a little of the brittleness will be lost and 
 a part of the hardness. ^ It is now in condition to be used for steel 
 engraving tools, lathe tools, and other implements to cut metals. 
 If we heat to 250° C. (480° F.), we temper still further, and so on. 
 
 It is interesting to note tha*t when hardened steel is tempered, 
 the physical changes produced by the tempering — the decrease 
 in hardness and brittleness, increase in electric conductivity, etc. 
 — precede the separation of carbon from the solid solution. By 
 tempering we may lose 70 per cent, of the hardness, 93 per cent, 
 of the electric resistance, and nearly 100 per cent, of the thermo- 
 electric power produced in the steel by the hardening operation, 
 when only 13 per cent, of the carbon has been changed from the 
 dissolved form.^ 
 
 Hardening^ Tempering, and Annealing. — -Only quenching in 
 water, or in some other medium which takes the heat away as fast 
 or faster, goes under the name of hardening. Quenching in 
 heavy oil, melted lead, etc., cools the steel less rapidly, and makes 
 it less hard and less brittle than quenching in water, so to this 
 operation the name of "tempering" is given. Cooling in the air, 
 in sand, in the furnace, or by any other slow method, is called 
 "annealing." 
 
 Magnetism and the Lines G-O-S-K.— It will be remembered 
 that iron is present in the solid solution in the gamma allotropic 
 form, and therefore the solid solution is non-magnetic. Therefore 
 all the steels to the right of the point in Fig. 262, page 295, lose 
 the last of their magnetism at the same time as they cross the 
 lines O'S-K. These steels comprise all containing 0.4 per cent, 
 of carbon and more. To harden, anneal, or restore such steels we 
 may guide our work of heating by means of an ordinary horse- 
 shoe magnet, which makes a most accurate and simple tool. 
 Let the magnet hang outside of the furnace and take a sample of 
 steel out at intervals to test it. When it no longer attracts the 
 magnet, begin to cool it. For steels with less than 0.4 per cent, 
 carbon we can use the magnet to tell us when the temperature 
 corresponding to the line M-0 is reached, for all iron loses its 
 magnetism at that point, and then it is a comparatively simple 
 matter to judge by eye the relatively short temperature intervals 
 
 ^ After the heating it is immaterial whether cooling is fast or slow, as the same result 
 will be produced. 
 
 ^ See footnote, p&ge 302. 
 
374 THE METALLURGY OF IRON AND STEEL 
 
 above that point which it is necessary for the steel to traverse 
 before it crosses the line G-O. If this method i« followed I think 
 it will be found in many works that annealing temperatures have 
 been much too high, and that better steel will be obtained in 
 future. We do not get the steel any softer by annealing it hotter, 
 but only by slower cooling from the correct temperature. 
 
 The Constituents of Hardened and Tempered Steels 
 
 It is now pretty generally admitted by metallurgists that aus- 
 tenite is the solid solution of gamma iron and carbon. When 
 this cools slowly through the critical range it decomposes into 
 ferrite and cementite. It is also very generally admitted that the 
 decomposition does not take place "spasmodically, but progresses 
 by stages, and many believe that the substances identified under 
 the microscope as martensite, troostite, and sorbite are the prod- 
 ucts of these stages. That is to say, martensite is the first stage 
 of decomposition of austenite, troostite is the second, sorbite is 
 the third, and pearlite is the consummation. If this is so, then 
 austenite will be found in the steels cooled with the greatest ra- 
 pidity, martensite will be found in those cooled with the next de- 
 gree of rapidity, troostite will be found in the next intermediate 
 steels, sorbite in the next, and pearlite in those slowly cooled. 
 To this extent, indeed, the facts agree with the argument, but 
 hardly any dare predicate further too positively from this evi- 
 dence alone. Our knowledge upon this whole subject is still very 
 new, and though a small army of workers is busy collecting evi- 
 dence and interpreting it as best they can, all our present state- 
 ments must necessarily be made tentatively, with the idea of pre- 
 senting the facts so that they may be of some practical benefit, 
 even though later information may oblige us to change slightly 
 the scientific basis upon which we found them. 
 
 Austenite. — Austenite- can be obtained at atmospheric tem- 
 peratures in ordinary carbon steels only when three conditions 
 are collectively present: (1) The brake action of carbon must be 
 very strong, that is, there' must be above 1.1 per cent, of carbon 
 present; (2) the steel must be cooled with the greatest rapidity, 
 as by quenching it in iced solutions at, or a few degrees below, 
 zero C; and (3) it must be cooled from above 1000° C. (1830° F.). 
 Even then austenite cannot be preserved throughout the whole 
 mass of the steel, but at least a part of it will be decomposed to 
 
THE HEAT TREATMENT OF STEEL 375 
 
 the stage represented by the martensite structure (see Fig. 312) 
 while cementite will separate also if the carbon is very high. Un- 
 der the microscope austenite may be differentiated from marten- 
 site by its white color after etching with a 10-per-cent. solution of 
 hydrochloric acid. Better results are obtained if the etching is 
 aided by electrolysis, the steel being made the anode, or positive 
 pole, and a piece of platinum being made the cathode. Aus- 
 tenite does not occur often in cold commercial carbon steels, but 
 is found frequently in alloy steels. 
 
 Martensite: — Martensite is the chief constituent of ordinary 
 hardened steels, that is, of steels quenched from above the criti- 
 cal range in water or in an iced solution. Its structure is shown in 
 Fig. 312. It is even harder than austenite, for a steel needle 
 drawn across the surface of a polished piece of steel will scratch 
 the austenite plainly without making a mark upon the martensite. 
 Its structure is developed by the "polish attack" method' of F. 
 Osmond (see p. 447). Martensite is also found very largely in 
 the tempered steels, to which they doubtless owe their quality of 
 hardness. It was long thought to be the solid solution itself, and 
 is still spoken of as such in several books upon the subject; but 
 the microscope shows positively that it is not free from some 
 decomposition, 
 
 Hardenite, — Hardenite is the name sometimes given to "sat- 
 urated martensite," that is, martensite containing 0.9 per cent, 
 carbon. It is often found in German and French, and occasion- 
 ally in English and American books. 
 
 Troostite. — Troostite is obtained in either of three ways: (1) 
 By quenching steel in water when it has cooled just to the line 
 P'S-K in Fig. 252, page 295; (2) by quenching steel from a higher 
 temperature in boiling water, oil, or some other of the tempering 
 mediums; and (3) by hardening in the usual way and then tem- 
 pering by reheating. In other words, troostite is a product of the 
 u^fet tempering operations, and is found abundantly in tempered 
 steels. Under the microscope it may be distinguished by the 
 "polish attack" (page 447) or by etching with tincture of iodine. 
 It appears as yellow, brown, blue, ^or black colorations on the 
 borders of the martensite, and often between the martensite and 
 sorbite. The division between it and the martensite is very sharp, 
 but it sh.ades very gradually into the sorbite (Figs. 314, 315). 
 
 Sorbite. — Sorbite is very close to pearlite, and differs from it 
 chiefly in that the crystals of ferrite and cementite are not quitq 
 
376 
 
 THE METALLURGY OF IRON 7VND STEEL 
 
 FIG. 309.— NO. lA. STEEL OF 0.34 
 PER CENT. CARBON OVERHEATED 
 TO ABOUT 1300° C. (2372° F.) 
 Magnified 250 diameters. 
 
 FIG. 310.— NO. 2A. SAME AS NO. lA. 
 
 REHEATED SLIGHTLY ABOVE AC^. 
 
 Magnified 250 diameters. 
 
 FIG. 
 
 311.— STEEL OF 1 PER CENT. 
 CARBON BURNT. 
 Magnified 265 diameters. 
 
 FIG. 312.— MARTENSITE. 
 Magnified 250 diameters. 
 
 FIG. 313.— MARTENSITE (WHITE) AND 
 
 TROOSTITE (DARK). 
 Magnified 500 diameters. Etched lightly 
 with tincture of iodine. (H. C. Boynton.) 
 
 FIG. 314.— AUSTENITE (WHITE) AND 
 
 TROOSTITE (DARK). 
 Magnified 60 diameters. Steel of 1.41 per 
 
 cent, carbon. Etched with picrio acid. 
 
 (William Campbell.) 
 
THE HEAT TREATMENT OF STEEL 
 
 377 
 
 perfectly developed and segregated from one another (see Fig. 
 315). It might at first seem almost like splitting hairs to differ- 
 entiate between the two, but this is not so, because of the im- 
 portance of sorbite, due to its having greater strength than 
 pearlite. Pearlite has a finer and more intimately entangled 
 structure than any other slowly cooled steel, and to this we 
 attribute the fact that pure pearlite steel (0.9 per cent, carbon) 
 is stronger than any of them- 
 But the structure of sorbite is 
 even finer than that of pear- 
 lite, and sorbitic steels are 
 correspondingly stronger than 
 pearlitic steels. 
 
 Sorbite may be obtained by 
 quenching the steel immedi 
 ately below, or just at the end, 
 of cooling through the critical 
 range, or. by cooling the steel 
 pretty fast through the critical 
 range without actually quench- 
 ing, or by .rapid cooling and 
 then reheating .to about 600° 
 C. (1110° R). 
 
 Osmondite, — Only very recently what is apparently another 
 constituent of hardened and tempered steels has been discovered, 
 and to this the name of "osmondite" has been given. It would 
 seem that osmondite is a solid solution of carbon (or of a carbide 
 of iron) in the alpha allotropic modification of iron. It is there- 
 fore an anomaly, and can only exist in equilibrium momentarily, 
 so to speak, because the normal solid solution contains the iron in 
 the gamma allotropic form. It would seem, however, that when 
 this gamma solid solution decomposes, it passes through a phase 
 wherein the iron has changed from the gamma to the alpha form, 
 before the prepipitation of the ferrite occurs. In other words, 
 then, osmondite is simply another stage in the decomposition of 
 austenite into pearlite. It differs from -the other stages by having 
 a definite constitution and nature, whereas martensite, troostite, 
 and sorbite are more or less indefinite and uncertain in compo- 
 sition, being in fact probably a mixture of two or more constitu- 
 ents rather than definite and individual components. 
 
 It is evident that osmondite can never compose the whole of 
 
 FIG. 315.— FERRITE AND SORBITE IN 
 
 RAIL STEEL. 
 
 Magnified 250 diameters. Etched with 
 
 picric acid. 
 
378 
 
 THE METALLURGY OF IRON AND STEEL 
 
 any piece of steel, because tlie earlier stages of the austenite de- 
 composition will grade into it on the one side and the later stages 
 on the other. That is to say, we would not expect that the entire 
 piece of steel would change simultaneously and instantaneously 
 from austenite into osmondite, and then likewise from osmondite 
 to pearlite. Osmondite may be obtained by tempering hardened 
 steel at 400° C. (750° F.), but a method of distinguishing it under 
 the microscope is not known at present. Our chief means of 
 recognizing it are by the fact that it causes the steel to dissolve 
 more rapidly in dilute sulphuric acid and to be colored more deeply 
 by alcoholic hydrochloric acid, and by the fact that steel contain- 
 ing a good deal of osmondite has lost a major part of all the qual- 
 ities given to it by hardening, although almost all the carbon is 
 still in the solid solution. 
 
 It is to be noted that the evidence proving the existence of 
 osmondite is still open to the possibility of' doubt, although the 
 probabilities are greatly in favor of it. 
 
 Summary, — We may summarize the constituents of hard- 
 ened and tempered steels in the following table, which is adapted 
 from one designed by Howe: 
 
 STAGES OF THE TRANSFORMATION FROM AUSTENITE INTO 
 
 PEARLITE 
 
 Name 
 
 Constitution 
 
 Austenite. 
 
 Martensite, a transition substance . 
 
 Troostite, a transition substance . 
 
 Osmondite. 
 
 Sorbite, a transition substance . . 
 
 Pearlite. 
 
 Solid solution of an iron carbide in 
 gamma iron. 
 
 The next step. Some of the gamma 
 iron is changed to beta and to alpha 
 iron. It is harder than austenite. 
 
 The third step. The quantity of 
 gamma and beta iron constantly de- 
 creasing, that of alpha iron con- 
 stantly increasing. 
 
 Solid solution of iron carbide in alpha 
 iron. 
 
 A mixure of a constantly decreasing 
 quantity of osmondite with a con- 
 stantly increasing quantity of pearl- 
 ite, too fine to be resolved by the 
 microscope. 
 
 A conglomerate, or a mechanical mix- 
 ture of free alpha iron (alpha f errite) 
 with the iron carbide, FejC, cemen- 
 tite. 
 
THE HEAT TREATMENT OF STEEL 379 
 
 In a very illuminating discussion of the constitution of iron 
 carbon alloys,^ Albert Sauveur predicates that when steel cools 
 slowly through the critical range and there takes place the de- 
 composition of austenite^the solid solution of carbon and gamma 
 iron — it first changes into a solid solution of carbon and beta iron, 
 thence into a solid solution of carbon and alpha iron, and thence 
 into cementite and alpha iron. In other words, Sauveur con- 
 siders that, instead of the solid solution decomposing by the 
 precipitation of its gamma ferrite, which' immediately changes 
 to beta and then to alpha ferrite, the ferrite in the solid solution 
 is changed from gamma to beta and then to alpha, after which the 
 precipitation occurs. By the light of recent researches this 
 theory of Sauveur' s is greatly strengthened. 
 
 Practice in Annealing, Hardening, and Tempering. 
 
 Furnaces. — There is not space here to discuss the many different 
 types of furnaces used for heating for these purposes, but a good 
 descriptive chapter on the subject will be found in Reference No. 
 149. The requisites of such a furnace are: (1) that it shall be of 
 a uniform temperature in all parts of the heating chamber; (2) 
 that it shall heat the pieces uniformly from all sides; (3) that its 
 temperature shall be readily under the control of the operator. 
 It is very desirable also that the temperature should be indicated 
 by means of some scientific instrument instead of the unaided eye 
 of the operator. Skillful as these operators are, it is now gen- 
 erally admitted that a good pyrometer will give more reliable 
 results, and they have the further advantage of giving an auto- 
 graphic record of the temperature which may be kept for future 
 reference. (4) That the metal shall not come in contact with 
 solid fuel, because the carbon or injurious impurities in this fuel, 
 such as sulphur and phosphorus, will be absorbed by the metal 
 and vitiate it. The usual means of heating these furnaces is the 
 flame from combustion of solid fuel, gas or oil, or heat converted 
 from electric energy. Electric heat, although expensive, gives 
 the most effective control of the temperature and also enables us 
 to get with ease, either the highest or lowest temperatures. 
 (5) That ready access to -the pieces being heated shall be had so 
 that they may be removed quicldy and easily; (6) that the pieces 
 shall be so supported as not to bend imder their own weight 
 
 ^Joum. Iron and Steel Inst., No. 4, 1906. 
 
380 THE METALLURGY OF IRON AND STEEL 
 
 at the heat to which they are subjected. (7) That means shall 
 be provided, where possible, to prevent oxidation of the iron. 
 This is of major importance in the case of small pieces, or of pieces 
 which are to be treated after machining, or cold rolling, so that 
 they shall be left with a bright surface and good corners. The 
 means for accomplishing this object are usually to heat the pieces 
 invade of muffles around the outside of ^hich the combustion 
 takes place, or else to pack them in air-tight boxes, and sometimes 
 inside these boxes or muffles a stream of reducing gas is led to 
 prevent oxidation, or else the metal is packed in lime or other 
 neutral material, and sometimes a few pieces of. charcoal are put 
 inside the muffle or mixed with the lime to unite with any oxygen 
 which finds its way inib the interior. Another method of heating 
 to prevent oxidation, which method has the further advantage of 
 obtaining uniform temperature, heating the piece uniformly on 
 all sides and giving an excellent control of the temperature, is 
 to have the pieces immersed in baths of oil, lead, barium chloride, 
 or other substances in a molten condition. These baths are then 
 uniformly heated and transmit their temperature to the metal. 
 
 Heating for Annealing^ etc. — Annealing has for its object the 
 treble purpose of (1) relieving any strains put upon the metal 
 during its cooling, or by mechanical treatment or otherwise; (2) 
 to restore the grain of the metal to that minute size which gives 
 it the best possible qualities, or (3) to soften it after hardening. 
 We may accomplish these three objects, or any of them, in one 
 annealing operation. The usual temperatures of annealing lie 
 between 200° and 1000° C. (390° and 1830° F.). Temperatures 
 from 200° to 500° C. (930° F.) are sufficient to relieve strains in 
 the metal and to soften it because of mechanical treatment or 
 hardening. It is only when we desire to produce a new grain size 
 that we heat above the critical temperature, and we have already 
 explained that for this latter purpose, it is best not to exceed 
 the critical temperature any more than is necessary, with, how- 
 ever, the qualifications noted on pages 358 and 365, in the case of 
 very bad overheating or ingotism. Where we have obtained by 
 means of cold rolling, wire drawing, or other mechanical work 
 in the cold, a grain size already small, it is bad practice to form 
 a new grain size,, and therefore the steel should then be heated 
 only to that low temperature which will soften it. 
 
 Hardening Practice. — When a piece of steel is rapidly cooled, 
 tremendous strains are developed in it as it passes through the 
 
THE HEAT TREATMENT OF STEEL 381 
 
 critical temperature. We have already studied the phenomenon 
 of recalescence, or liberation of heat within the metal by the 
 chemical reaction of the separation of ferrite and cementite. 
 
 The recalescing steel, which has been rapidly contracting due 
 to the falling temperature, now tends to expand because of this 
 liberation of heat. But, as cooling takes place from the outside 
 inward, the outer shell of steel passes through the recalescence 
 period first, and is already beyond that point and contracting 
 further when the interior begins to liberate heat and tend to 
 expand. This brings a double pressure to bear upon the steel, 
 and the strains are so large as to frequently cause what are known 
 as "water cracks" in the hardened steel. Sometimes the outer 
 shell of steel actually flies to pieces under the strain, or " bursts " 
 as it is called. Again a long cavity will be opened down the 
 center of the piece. In the case of armor plate, for example, 
 the appearance of cracks after the hardening of the steel is often 
 welcomed, as it shows that the strains produced in the metal 
 have been relieved to some extent, and a curved turret will 
 actually be stronger under impact after cracking than before, 
 because the latent strains are relieved. The operator must. take 
 these matters seriously into consideration, remember that the 
 greatest strains are produced by uneven cooling which may occur 
 through: (1) the juxta position of thick and thin sections; (2) 
 uneven temperature of the piece which is heated up, which may 
 best be avoided by the slow heating of large pieces; (3) uneven 
 temperature of thei bath in which the steel is quenched, or (4) 
 careless immersion in the bath. An uneven temperature of the 
 bath may be avoided by having small-sized baths with abundant 
 circulation of the quenching medium therein, or else by quench- 
 ing the objects in a voluminous spray of liquid, such as is prac- 
 tised in the hardening of armor plate, for example, the operator 
 should move the piece around in the bath as much as is necessary 
 to prevent vapor accumulating next to the steel and thus form- 
 ing a place where the cooling is not so rapid, and the piece must 
 be held in tongs and not allowed to drop on to the bottom of the 
 quenching tank. 
 
 Hardened pieces occasionally warp as a result of the strains 
 set up. This may be prevented to some extent by having the 
 thickest parts of the piece enter the bath first, and also by an- 
 nealing the specimen at a relatively low temperature before 
 heatip.g for hardening, in order to previously remove all strains. 
 
382 THE METALLURGY OF IRON AND STEEL 
 
 It may be necessary to clamp pieces of very intricate shape in 
 order to mechanically prevent their warping. 
 
 Quenching Baths. — The object of the quenching bath is to take 
 the heat away from the metal at the rate of speed best suited for 
 the hardness or other qualities desired, to be given to it. Noth- 
 ing from the bath enters the steel, and it has no virtues other 
 than to remove the heat. It may be, however, that certain baths 
 will cool steel rapidly through the critical range and then slowly 
 thereafter, or vice versa, and there may sometimes be virtue in 
 this property. Saturated solutions of certain salts might de- 
 posit their dissolved substance uniformly on the surface of the 
 steel and this incrustation may have the effect of cooling the steel 
 more slowly at a lower temperature. Further than this char- 
 acteristic, the speed of cooling will depend upon the following 
 factors: (1) temperature of the bath; (2) heat conductivity of 
 bath; (3) viscosity of the bath. (The importance of viscosity 
 results from the circumstance that liquids carry heat away 
 largely by convection) ; (4) specific heat of the bath, because the 
 less the specific heat of substances, the more heat will be absorbed 
 in a given time by the layers adjacent to the quenched body. 
 Furthermore, the less the specific heat, the greater amount in 
 heat will the bath absorb without rising in temperature; (5) the 
 volatility of the liquid, because the more readily it forms vapor 
 when heated, the more liable is there to be a cushion of vapor 
 between the quenched body and the body of the liquid itself. 
 Mercury, because of its low melting-point, can be refrigerated to 
 a low point and this makes an effective hardening medium. 
 Solutions of calcium chloride will act in the same way. On the 
 other hand, liquid air, in spite of its very low temperature, is a 
 poor quenching medium because of its ready volatility. 
 
 Combined Hardening and Tempering, — If steel be cooled from 
 above the critical range by quenching in some slow heat-con- 
 ducting liquid in the first instance, as cylinder oil, or melted 
 solder at 200° C. (392° F.), the same intermediate hardening and 
 embrittling effect will be produced upon it as if it were first 
 hardened in water and then tempered a certain amount. There- 
 fore the quenching in oil and similar mediums has come to be 
 called "tempering." Another method of combining hardening 
 and tempering after only one heating is used in the tempering of 
 the cutting edges of chisels and similar tools: the end of the tool 
 is first heated just above the critical range, and then the extreme 
 
THE HEAT TREATMENT OF STEEL 383 
 
 point only is quenched in water until it is black, after which it is 
 withdrawn and rubbed bright upon a piece of sandpaper, or upon 
 a brick. This is done merely to give a bright surface upon which 
 to observe the play of temper colors. The heat from the shank 
 now begins to creep down into the point, which takes the various 
 temper colors in order, beginning with the lemon. When the 
 desired degree of tempering is reached — say, the pigeon-wing 
 color — the whole tool is put into water. This is merely to "put 
 out the fire" and stop more heat coming down into the 
 tempered-point; it has nothing to do with the tempering opera- 
 tion itself. 
 
 Reheating for Tempering. — A more accurate method of tem- 
 pering, because it enables pyrometers to be used and because 
 the operation can be conducted more slowly, is to allow the 
 hardened piece to cool and then reheat it to a desired temperature 
 at which its temper is to be drawn. By this means also we can 
 reheat the piece in baths of lead-tin alloys, melting at low heats, 
 which insures accurate control and that all parts of the piece are 
 uniformly heated. 
 
 Temper Colors. — Nature has provided a ready means of de- 
 termining the temperature of steel between 200° and 300° C. 
 (390° and 570° F.) without the aid of thermometers or other in- 
 struments, and, since this is the range of temperatures in which 
 practically all of the tempering of hardened steel takes place, this 
 provision is a most fortunate one. It comes about through the 
 oxidation of the metal at those different points. At 200° C. 
 (390° F.) a thin film of oxide forms upon the steel, but is not 
 sufficient to entirely hide the white color underneath, so that the 
 combination produces a light lemon color. As the temperature 
 rises the film of oxide becomes thicker and the yellow color darker 
 until, at about 225° C. (437° F.), it has changed to faint straw. 
 
 In table XXVII are the temperatures and other phenomena 
 relating to the treatment of steel and the temperatures for- 
 tempering various tools. 
 
384 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXVII 
 
 > 
 
 o p 
 
 ° C. 
 
 
 ScraDers for brass 
 
 437 
 to 
 455 
 
 225 
 
 to 
 235 
 
 Steel engraving tools. 
 
 Light turning tools 
 
 Hammer faces. 
 
 Planers for stfeel 
 
 Planers for iron. 
 
 Ivory cutting tools 
 
 Paper cutters. 
 
 Wood engraving tools. 
 
 Bone cutting tools 
 
 DriUs. 
 
 Milling cutters. 
 
 
 
 Wire drawing plates 
 
 456 
 to 
 
 482 
 
 236 
 to 
 250 
 
 Boring cutters. 
 
 Leather cutting dies 
 
 Taps 
 
 Screw cutting dies. 
 Chasers. 
 
 Rock drills 
 
 Mill chisels and picks. 
 
 Penknives 
 
 Punches and dies. 
 
 Reamers 
 
 Shear blades. 
 
 Half-round bits 
 
 Planing and molding 
 
 Gouges 
 
 cutters. 
 Stone cutting tools. 
 
 Plane Irons 
 
 
 
 
 Twist drills 
 
 483 
 to 
 527 
 
 251 
 to 
 275 
 
 Flat drills. 
 
 Wood borers. . , 
 
 Pressing cutters. 
 <!]oopers' tools- 
 
 Cup tools 
 
 Edging cutters 
 
 Wood bits and augers 
 
 Dental and surgical instruments 
 Hack saws. 
 
 528 
 
 to 
 
 572 
 
 276 
 to 
 300 
 
 Cold chisels for steel. 
 
 Axes and adzes. 
 
 Cold chisels for cast iron. 
 
 Saws^for bone and ivory 
 
 Needles, gimlets 
 
 Chisels for wood. 
 Framing chisels. 
 Screw drivers. 
 Springs. 
 
 Circular saws for metal 
 
 Saws for wood 
 
 
 Case-hardening, Carbonizing, or Cementation 
 
 Steel low in carbon is soft and tough but will not withstand 
 wear or abrasion; hard steel, on the other hand, is brittle except, 
 of course, such a steel as manganese alloy steel, whose use is 
 limited by price. Where a steel has to resist a combination of 
 stresses consisting of penetration or abrasion and violent shocks, 
 such a combination of stresses, in other words, as is brought to 
 
THE HEAT TREATMENT OF STEEL 385 
 
 bear upon armor plate, plow shares, certain gears and pinions, 
 crank shafts, pivots in moving machinery, etc., or a combination 
 of cutting and shocks, such as that which jail-bars and burglar- 
 proof safes might have to resist, it is customary to have a com- 
 bination of metals as explained on page 181. Or else, we may 
 produce a high-carbon surface on a piece of soft steel by the 
 process known as case-hardening or carbonizing. This case- 
 hardened metal is then heat-treated by a hardening or tempering 
 process which makes the outside very hard but leaves the interior 
 tough and ductile. The high carbon of the outside is obtained 
 by virtue of the so-called cementation process. 
 
 Theory of Cementation, — The principle upon which case- 
 hardening depends is the slow absorption of carbon by iron or 
 steel at a bright red heat as already partially explained on page 
 69. If a piece of soft steel be immersed in melted potassium 
 cyanide at a temperature of 900° C. (1650° F.), the carbon con- 
 tained in the cyanide will slowly penetrate the metal. It enters 
 the steel in the form of a solid solution and the amount of carbon 
 and the depth ^ to which it penetrates will depend upon the 
 following conditions in the order of their importance : 
 
 (1) The temperature of cementation. 
 
 (2) Time during which it proceeds. 
 
 (3) The kind of cement material used; for example, potassium 
 cyanide gives a quicker penetration than pure charcoal. 
 
 (4) The kind of steel used; that is to say, manganese, chromium, 
 tungsten, molybdenum, etc., in the steel increase the rate of 
 cementation, while silicon, aluminum, nickel, titanium, etc., 
 decrease it. 
 
 Temperature of Cementation. — The cementation process takes 
 place with extreme slowness below the critical temperature of 
 steel, and increases rapidly with the temperature above that 
 point. It is not very advisable, however, to carry it on at a 
 very high temperature because the long over-heating gives the 
 core of the steel a coarse structure and makes it weak and 
 brittle. Lake (No. 149) gives the following table for the speed 
 of cementation with different substances at varying temperatures. 
 
 25 
 
386 THE METALLURGY OF IRON AISTB STEEL 
 
 TABLE XXVIII 
 
 
 Materials use 
 
 d and rate of penetration in inches in 8 hours 
 
 
 Temperature 
 
 
 
 
 
 Temperature 
 
 in degrees 
 
 Charcoal 60 
 
 Ferro-cyanide 
 
 
 
 in degrees 
 
 Fahrenheit 
 
 per cent. +40 
 
 66 per cent. + 
 
 Potassium 
 
 Powdered 
 
 Centigrade 
 
 
 per cent, of 
 
 34 per cent. 
 
 ferro-cyanide 
 
 wood char- 
 
 
 
 carbonate of 
 
 of bichromate 
 
 alone 
 
 coal alone 
 
 
 
 barium 
 
 
 
 
 
 1290 
 
 
 
 - 
 
 
 700 
 
 1475 
 
 0.020 
 
 0.033 
 
 0.020 
 
 0.020 
 
 800 
 
 1650 
 
 0.088 
 
 0.069 
 
 0.079 
 
 0.048 
 
 900 
 
 1830 
 
 0.128 
 
 0.128 
 
 0.128 
 
 0.098 
 
 1000 
 
 2010 
 
 0.177 
 
 0.177 
 
 0.198 
 
 0.138 
 
 1100 
 
 A very usual temperature for cementation is 900° C. (1650° F.). 
 
 Influence of Time on Cementation. — The following two tables 
 from Lake also show the influence of the time on the depth of 
 penetration of the carbon for different materials. 
 
 Length of 
 time in 
 hours 
 
 Materials used and rate of penetration in inches at 100° C. 
 
 Carbon 60 per 
 cent. +40 per 
 cent, of car- 
 bonate 
 
 Ferro-cyanide 
 
 66 per cent. + 
 
 34 per cent, of 
 
 bichromate 
 
 Powered 
 
 wood char- 
 coal alone 
 
 Charcoal and 
 
 carbonate of 
 
 potassium 
 
 Unwashed 
 animal 
 black 
 
 1 
 
 0.031 
 0.039 
 0.047 
 0.078 
 0.118 
 
 0.033 
 0.037 
 0.049 
 0.074 
 
 0.128 
 
 0.028 
 0.053 
 0.063 
 0.072 
 0.098 
 
 0.059 
 0.078 
 0.094 
 0.011 
 0.138 
 
 035 
 
 2 
 
 4 
 
 6 
 
 8 . . 
 
 0.059 
 0.088 
 0.106 
 0.128 
 
 
 Length of 
 
 Materials used and rate of penetration in inches at 900° C. 
 
 time in 
 hours 
 
 Charred 
 leather 
 
 Ground wood 
 charcoal 
 
 Barium carbonate and 
 ■wood charcoal 
 
 2 
 
 4 
 
 8 
 
 12 
 
 0.045 
 0'062 
 0.080 
 0.110 
 
 0.028 
 0.042 
 0.062 
 0.070 
 
 0.055 
 0.087 
 0.111 
 0.125 
 
THE HEAT TREATMENT OF STEEL 387 
 
 Carbonizing Material. — Iron has been carbonized by means 
 of a diamond as early as 1815 without melting the metal, which 
 shows that pure carbon alone may accomplish the result. How- 
 ever, more impure forms of carbon give much more rapid carbon- 
 ization, and for commercial purposes the materials used are 
 generally charcoal from wood, with which is mixed other sub- 
 stances; powdered bone; charred leather or sugar, this latter being 
 valuable because of the absence of injurious impurities; horn, 
 animal black, lamp black, anthracite, graphite, etc. In the 
 lamp black process it is common to deposit the soot on the sur- 
 face to be carbonized by means of a smoky gas or oil flame. 
 Gases containing carbon may also be used for cementation, of 
 which carbon monoxide is one of the best; acetylene is also 
 effective. We may also use liquids containing carbon, such as 
 melted potassium cyanide, ferro-cyanide of potash (Prussian 
 blue or ferro-prussiate -of potash). It would appear that the 
 presence of nitrogen assists in the absorption of carbon by the 
 steel and for this reason the various animal and vegetable 
 products mentioned above are preferred to charcoal from the 
 purer materials. It is also common in some cases to introduce 
 gases containing nitrogen, such as ammonia, etc., into the recep- 
 tacles where the cementation is being carried on. 
 
 A very important consideration in case-hardening is that the 
 carbonized material shall penetrate the steel at all points alike 
 and this is assisted by the more intimate contact between the 
 two, so that carbonizing with gas is often preferred to packing 
 the steel in solid materials. Carbonization in liquid cyanides 
 has the same advantage, and doubtless the nitrogen in these 
 cyanides is an important element in their efficiency. The 
 potassium cyanide bath is a very common means of cementation 
 on account of the convenience of its use and readiness of control. 
 One objection to it, however, is the very bad and poisonous 
 fumes coming from this melted liquid. 
 
 Steel Carbonized. — We ordinarily carbonize a steel running 
 from 0.10 to 0.22 per cent, of carbon in order to have a very 
 tough core and also to avoid the ill effects of over-heating of the 
 core which increases with the amount of carbon. The manga- 
 nese in the steel should be low, because this not only increases the 
 over-heating by lowering the critical temperature, but tends to 
 make the outside case brittle. The amount of manganese should 
 generally be under 0.35 per cent, and preferably less than 0.25 
 
388 THE METALLURGY OF IRON AND STEEL 
 
 per cent. Chromium makes the outer surface harder, and 
 refines the grain of the steel besides raising the temperature at 
 which over-heating occurs. It increases the strength of the 
 material, but makes it harder to machine. Many pieces for 
 case-hardening contain from 0.50 per cent, to 1.50 per cent, of 
 chromium. Vanadium and titanium are said to counteract to 
 some extent the effect of chromium in making steel difficult to 
 machine, and to increase its resistance to shock. Nickel is very- 
 valuable in case-hardened steel because it increases its strength, 
 decreases the brittleness and the tendency of cracks to spread. 
 From 2 to 3.50 per cent, of nickel is commonly used for high grade 
 material which can stand the extra price. The amount of silicon 
 is usually kept as low as possible on account of its decreasing 
 the depth and speed of cementation, and phosphorus should 
 be under 0.035 per cent, as this makes the metal brittle under 
 shock. 
 
 Percentage of Carbon Added. — On the very surface of the case- 
 hardened material, the amount of carbon will usually be nearly 
 1 per cent, or more, and will run as high as 2.50 per cent, in the 
 case of a very long cementation as in the manufacture of armor 
 plate, for example. Lake cites the case of a round bar that was 
 carbonized to the depth of 3/16 in. and showed the following 
 percentages of carbon in each 1/16 in. turned off the outside; 
 1.24, 0.85, 0.24, and 0.13 per cent. 
 
 Heat Treatment for Carbonizing. — In some cases it is customary 
 in order to save time and expense, to quench out the pieces as 
 they come from the cementation furnace, but it is much better 
 practice to cool them down to a black heat and then reheat and 
 harden in a separate operation, because this causes a somewhat 
 better diffusion of the carbon at the point where the case-harden- 
 ing joins the core. This junction line is always a line of weakness, 
 and when the pieces are quenched out from the quenching fur- 
 nace, there is a distinct line of demarkation, which is notably 
 shaded when the metal is reheated for hardening. Moreover, 
 the reheating restores the grain of the steel which has been 
 somewhat damaged by overheating during the cementation 
 process. 
 
XV 
 ALLOY STEELS 
 
 We have already described ordinary steel (which to distinguish 
 it frorti the so-called alloy steels is often known as ''carbon steel'') 
 as an alloy of iron and carbon. But there is another class of 
 materials to which the specific name of "alloy steels'* is applied. 
 This comprises steels to which a controlling amount of some alloy- 
 ing element in addition to the carbon is added. 
 
 Definitions. — The International Committee upon the nomen- 
 clature of iron and steel defines alloy steels as "those which owe 
 their properties chiefly to the presence of an element (or elements) 
 other than carbon." The distinction between an element added 
 merely to produce a slight benefit to ordinary carbon steel and the 
 same element added to produce an alloy steel is sometimes a very 
 delicate one. For example, manganese is added in amounts usu- 
 ally lesS^ than 1.50 per cent, to all Bessemer and open-hearth steels 
 for the sake of getting rid of oxygen and neutralizing the effect of 
 sulphur. Likewise silicon is sometimes added in amounts of 0.1 
 to 0.2 per cent, to get rid of blow-holes. But neither of these 
 additions produce what is known as an alloy steel. When we 
 make "manganese steel" containing 10 to 20 per cent, manganese, 
 the material has new properties quite different from the same 
 steel without manganese and we therefore get an alloy steel. 
 Similarly "silicon steel" containing 2 or 3 per cent, of silicon will 
 have an entirely new set of properties due to the silicon, and will 
 therefore become an alloy steel. 
 
 Ternary Alloys. — A ternary alloy, or three-part alloy, is an 
 alloy composed of iron, carbon, and one other influential element. 
 This class includes the alloy steels which are used most abundantly 
 by man and the most important of which are nickel steel, man- 
 ganese steel, chrome steel, tungsten steel, molybdenum steel, 
 silicon steel, etc. There are several other ternary steels which 
 have been investigated and used to a small extent, such as boron 
 steels, cobalt steels, etc. The field of useful ternary steels has 
 not yet been investigated except in the most meager degree and 
 a wide scope is left for future inventors. There are many ele- 
 
 389 
 
390 THE METALLURGY OF IRON AND STEEL 
 
 ments whose influence on steel has not yet been studied, and 
 even among those which are commonly used, there are some of 
 which only limited proportions have been employed. 
 
 Quarternary Steels. — Quarternary, or four part, steels consist 
 of iron, carbon, and two other alloying elements. The common- 
 est and most important of these are nickel-chromium steels, tung- 
 sten-manganese and tungsten-chromium steels, nickel-manga- 
 nese, manganese-silicon, tungsten-molybdenum, tungsten-nickel, 
 nickel-vanadium steels, etc., etc. The result produced by adding 
 an alloying element to ordinary carbon steel is astonishing and 
 incapable of being predicated, and that obtained by a combina- 
 tion of two alloying elements is far more so. New products re- 
 sult with properties entirely different, and in some cases almost 
 the opposite, of those of its constituents, so that almost any com- 
 bination at random may lead to a surprise, even when the effect 
 of different combinations of the same components is known. 
 Therefore the possibilities of quarternary steels seem to be very 
 great and the field has, as yet, hardly been touched. 
 
 Manufacture of Alloy Steels, — The manufacture of alloy steels 
 is usually very simple and calls for no special comment here. As 
 a general thing the alloying element is added like the recai^urizer. 
 For example, in the manufacture of manganese steel the requisite 
 amount of ferro-manganese will be added at the end of the pro- 
 cess; in the manufacture of nickel steel we may add ferro-nickel in 
 the same way, but it is more common here to add shotted-nickel 
 during the process and allow it to dissolve in the steel bath and re- 
 main there until the metal is tapped; tungsten steels and tung- 
 sten-chrome steels are often made by the crucible process and the 
 requisite amount of ferro-chrome and ferro-tungsten, or of me- 
 tallic chromium and metallic tungsten, is placed on top of the 
 charge when the crucible is filled. 
 
 The treatment of some of the alloy steels is not so simple: 
 Nickel steel may be heated, rolled and forged without any great 
 precaution, but these operations are performed upon manganese, 
 tungsten, and some of the other alloy steels only after great diffi- 
 culty and experience. Under the head of these different steels, I 
 shall describe the proper method of treatment. 
 
 Nickel Steels 
 
 Nickel steels are the most important of all the alloy steels 
 and are the most abundantly used. In the ordinary commercial 
 
ALLOY STEELS 391 
 
 alloys the nickel ranges from 1.50 to 4.50 per cent, and usually 
 from 2.00 to 3.75 per cent., while the carbon is usually from 0.20 
 to 0.50 per cent. Not counting armor-plate, which is really 
 a quarternary steel, containing both nickel and chromium, the 
 most important uses of nickel steel are for structural work in 
 bridges, railroad rails, especially on curves, steel castings, ord- 
 nance, engine forgings, shafting, especially marine shafting, 
 frame and engine parts for automobiles, wire cables, axles, es- 
 pecially for automobiles and railroad cars, etc., etc. We can 
 best learn the reasons for these particular uses by discussing 
 the distinctive properties conferred by the nickel and their use- 
 fulness. 
 
 Tensile Properties. — The chief distinction between nickel- 
 steel and carbons-teel is the higher elastic limit of the former, and 
 especially the fact that this higher elastic limit is obtained with 
 only a slight decrease in ductility. About 3.50 per cent, of nickel 
 added to carbon-steel will increase the elastic limit nearly 50 
 per cent., while reducing the ductility only about 15 to 20 per 
 cent. It is this increase in elastic limit which is probably the 
 chief reason for the increased resistance of nickel-steel to what 
 is known as "fatigue," that is to say, its resistance to repeated 
 stresses and alternating stresses^ under which all steel will ulti- 
 mately break down, even though the load is far less than that 
 it can bear indefinitely if constantly applied. It is probable 
 that the minute structure of nickel-steel is also advantageous 
 in this same connection. About 3.50 per cent, of nickel will 
 give steel approximately six times the life in resistance to fatigue. 
 The records of shipping show that the great majority of acci- 
 dents to vessels at sea come from the breaking of the propeller 
 shafts which is doubtless due to alternate stresses because these 
 long shafts are put out of alignment by each passing wave; so 
 now practically all large vessels use hollow nickel-steel shafts. 
 It is the higher elastic limit that is responsible also for the use of 
 nickel-steel in bridges, ordnance, automobile parts, and wire cables, 
 for we may obtain equal strength with less weight or greater 
 strength with the same weight. Besides the elastic limit the ulti- 
 mate tensile strength of nickel-steel is increased also by the addi- 
 tion of nickel. The increase is not so great in this particular, and 
 
 1 Repeated stresses- are stresses put upon a body at intervals and relieved meanwhile, 
 while alternate stresses are stresses first in compression and then in tension, such, for 
 instance, as the stresses in a wire that is bent backward and forward, or in a rotating 
 shaft that is not absolutely in alignment. 
 
392 THE METALLURGY OF IRON AND STEEL 
 
 consequently the elastic ratio, i.e., the ratio of the elastic limit 
 to the tensile strength, is increased still more greatly. The elastic 
 limit of ordinary rolled carbon-steel should be at least one-half 
 of the tensile strength, while that of 3.50 per cent, nickel-steel 
 should be at least 60 per cent. 
 
 Crystalline Structure, — ^The crystalline structure of nickel-steel 
 is more minute than ordinary carbon-steel, and this is prob- 
 ably one of the chief causes for the toughness of nickel-steel, and 
 also for the fact that cracks develop in it relatively slowly; in the 
 yielding of steel to fatigue the damage starts by the opening of a 
 crack of microscopic proportions through the cleavage planes of 
 the crystals, and this crack grows and spreads from crystal to 
 crystal until it is visible to the unaided eye, after which it proceeds 
 with still greater rapidity. As already stated, this development 
 is much slower in nickel-steel than in carbon-steel. Furthermore 
 if an armor plate is struck by a projectile, it does not crack so 
 easily, and the cracks do not extend so far if the plate is made of 
 nickel-steel. This fact and the greater strength of nickel-steel are 
 the chief reasons for the 3.25 per cent, of nickel in all modern 
 armor plate. 
 
 Modulus of Elasticity. — The modulus of elasticity of nickel- 
 steel containing not more than 4,or 5 per cent, nickel is about the 
 same as that of carbon-steel, namely, about 29,500,000 pounds 
 per square inch. With higher contents of nickel, however, and 
 especially with more than 20 per cent, of nickel, the modulus 
 of elasticity is lower. This results in the steel being much more 
 resilient or springy, and this is one of the important reasons why 
 not more than 4 per cent, of nickel is put into structural steels, 
 for a bridge built of steel which was resilient, even though strong, 
 would vibrate so much with the motion of a passing load as to 
 be unpleasant, and even unsafe on account of the repeated 
 stresses set up. The price of nickel, of course, enters into the 
 limitation of the amount used in structural material as well, and 
 it is found that 3.50 per cent, can be added without great expense 
 and with beneficial results. 
 
 Hardness. — Nickel-steel is harder than carbon-steel, though 
 not so much so but that it can be machined without difficulty. 
 This is taken advantage of in the use of nickel-steel railroad rails 
 for curves and other locations where the steel soon wears out. 
 The additional strength of the nickel-steel is also an advantage in 
 this connection and nickel-steel rails have been tried with success 
 
ALLOY STEELS 393 
 
 upon the famous horseshoe curve of the Pennsylvania Railroad 
 and other places. The hardness of nickel-steel is also accom- 
 panied by a lower coefficient of friction, and those properties, 
 together with the additional strength, are taken advantage of in 
 the use of nickel-steel in axles for automobiles, locomotives and 
 railroad cars. Equal strength can be obtained in an axle of 
 smaller size which has, of course, less bearing surface, and there- 
 fore still further reduced friction. 
 
 Soundness. — Nickel-steel castings are relatively free from 
 blow-holes and this together with the strength is a reason for the 
 use of this material for castings. They also have a lower melting- 
 point and run more easily in the molds. 
 
 Expansibility. — The coefficient of expansion of nickel-steel is 
 one of its most astonishing and unusual characteristics, for in 
 different samples it varies all the way from practically zero up to 
 the ordinary figure for carbon-steels. 
 
 Colby^ gives the following figures for the average coefficient 
 of expansion for each 1° C. temperature: ^ 
 
 Carbon-steel 0.00001036 (Guillaume) 
 
 Carbon-steel (0 . 25 per cent. C.) . 00001150 (Charpy) 
 
 Soft Carbon-steel 0,00001078 (Browne) 
 
 5.00 per cent. Nickel-steel 0.00001053 (Guillaume) 
 
 Invar. — But the coefficient of expansion with the ordinary 
 atmospheric changes of temperature becomes less and less as the 
 percentage of nickel increases until, when we reach 36 per cent, 
 of nickel, it is less than any metal or alloy known and amounts 
 practically to zero. This alloy is patented and sold under the 
 name of "Invar" and is used for scientific instruments, pendu- 
 lums of clocks, steel tape-measures for accurate survey work, etc. 
 In a paper read before the American Association for the Advance- 
 ment of Science in December, 1906, it was stated that tapes made 
 of invar used experimentally for United States Government sur- 
 vey work showed a very great increase in accuracy over ordinary 
 steel, tapes, and also in rapidity of use. These tapes varied an 
 infinitesimal amount during the first few months, after which 
 they became practically constant in length. The cause of this 
 peculiar effect of nickel upon the dilation of steel with an increase 
 in temperature is a result of the effect of nickel upon the critical 
 ranges of the steel, which we shall describe later. 
 
 iPage79of No. 152. 
 
394 THE METALLURGY OF IRON AND STEEL 
 
 Platinite. — As the amount of nickel increases beyond 36 per 
 cent., there is a slight increase in the coefficient of expansion 
 so that when we reach about 42 per cent, of nickel, the steel has 
 the same coefficient of expansion and contraction with the at- 
 mospheric temperature as has glass. It can therefore be used 
 for the manufacture of "armored glass," i.e., a plate of glass 
 into which a network of steel wire has been rolled and which 
 is used for fire-proofing, etc., because even though the glass should 
 break, it is held together by the steel network. It can also be 
 used for the electric connections passing through the glass plugs 
 in the base of incandescent electric lights. Platinum has been 
 used altogether for this latter purpose, in spite of its very high 
 cost, because it was the only metal hitherto known that had 
 the correct coefficient of expansion and contraction, and there- 
 fore the name of ''platinite" has been given to this patented 
 nickel-steel alloy. 
 
 Corrodihility. — Nickel-steel corrodes less than carbon-steel, 
 both in the presence of the atmosphere, fresh and salt water, the 
 ordinary acids, the smoke of locomotives, etc. Moreover, the de- 
 gree of corrodibility decreases with each increase in the amount of 
 nickel present. For this reason 30 per cent, nickel boiler tubes 
 have been used, especially in marine boilers. The great expense 
 of this material is, however, an obstacle to its common adoption. 
 
 Other Properties. — Ordinary nickel-steel containing about 3.50 
 per cent, of nickel has several other properties which distinguish 
 it from carbon-steel, among which we may mention its higher com- 
 pressive strength and greater toughness under impact. This 
 latter makes nickel-steel especially resistant to shocks, for it not 
 only takes a greater blow to bend it but it will bend through more 
 of an angle before cracking. Nickel-steel also has a greater shear- 
 ing strength which makes it advantageous for rivets, because 
 smaller rivets may be used and this means smaller holes in the 
 structural members that are being joined, and consequently a 
 greater area of these members left to support the strains upon 
 them. In this connection, however, it should be remembered 
 that nickel-steel does not weld as well as carbon-steel, and there- 
 fore greater care is required in upsetting the rivets during the pro- 
 cesses of construction. Nickel segregates very little in iron and it 
 also has the advantageous property of hindering the other ele- 
 ments from segregation, so that nickel-steel is less liable to these 
 irregularities than carbon-steel. In steel over 0.50 per cent, car- 
 
ALLOY STEELS 395 
 
 bon, nickel has a tendency to make the carbon come out as 
 graphite. 
 
 Critical Changes: — Nickel has a very important effect upon the 
 critical changes of iron and steel. This fact will readily be be- 
 lieved because it is known that many of the elements added to 
 steel produce important changes in the critical points. G. B. 
 Waterhouse/ while investigating the effect of 3.80 per cent, 
 of nickel upon iron, showed that this amount of nickel did not 
 make any appreciable difference in the mode of occurrence of the 
 critical points on cooling, but it did reduce the temperature at 
 which these critical points came by about 75° C. (167° F.). 
 
 As the amount of nickel in the alloys increases the tempera- 
 tures at which the critical ranges occur become lower and lower 
 until we reach 25 per cent, of nickel, when the critical ranges occur 
 below the atmospheric temperature. That is to say, the steel 
 does not ordinarily cool to the point at which the solid solution is 
 decomposed and the beta and alpha allotropic modifications are 
 assumed. 
 
 Irreversible Transformations, — The great peculiarity of the 
 critical changes of the nickel-steel alloys with less than 25 per 
 cent, of nickel is that they are irreversible. By this we mean that 
 the change which takes place at one temperature on cooling is 
 not reversed on heating at the same temperature, or anywhere 
 near that temperature. In other words when we cool a nickel- 
 steel containing 20 per cent, nickel the solid solution is not de- 
 composed and the alpha allotropic modification is not assumed 
 until we get below 100° C. (212° F.). But having cooled the steel 
 to that point and decomposed the solution, we can now heat it 
 nearly to 600° C. (1112° F.) before the reverse change takes place 
 and we again form the solid solution and the gamma allotropic 
 modification. In other words, it is possible to have a sample of 
 nickel-steel between 100° and 600° C. which shall be in either 
 condition we like. With 20 per cent, of nickel, nearly 1 per cent, 
 of^carbon and 1.40 per cent, of manganese, the transformation 
 point on cooling is 188° below zero C. (306° below zero F.), while 
 the transformation point on heating is well above the atmospheric 
 temperature. Therefore at atmospheric temperature we may 
 have such a piece of steel in either condition we like, and a very 
 interesting experiment is formed by having a bar of this steel 
 one end of which has been cooled more than 188° below zero C, 
 
 iNo 150. 
 
396 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 316.— 1.54 PER CENT. 
 ROLLED. 
 Magnified 225 diameters 
 EtcSied ■vrith picric acid, 
 
 CARBON. FIG. 317.— 1.24 PER CENT. CARBON. 
 ROLLED. 
 
 Magnified 225 diameters. 
 Etched with picric acid. 
 
 FIG. 318.-1.24 PER CENT. CARBON. 
 
 SLOWLY COOLED STEEL. 
 
 Magnified 225 diameters. 
 
 Etched with picric acid. 
 
 FIG. 319.-1.24 PER CENT. CARBON 
 
 COOLED EXTREMELY SLOWLY. 
 
 Magnified 265 diameters. 
 
 Etimed with picric acid. 
 
 FIG. 320.-1.24 PER CENT. CARBON. FIG. 321.— 0.41 PER CENT. CARBON. 
 
 COOLED EXTREMELY SLOWLY. COOLED EXTREMELY SLOWLY. 
 
 Magnified 265 diameters. Magnified 265 diameters. 
 
 Etdied with picric acid. Etched with picric acid. 
 
 Nickel steels containing about 3.80 per cent, nickel, 0.12 per cent, dlicon, 0.05 per cent. 
 
 manganese, 0.014 per cent, sulphur, 0,008 per cent, phosphorus, 0.01 per cent, aluminum 
 
 (G. B. Waterhouse in the Author's Laboratory.) 
 
ALLOY STEELS 397 
 
 while the other end has not. The end that has been cooled will 
 be magnetic and the other end non-magnetic. 
 
 When we have more than 40 per cent, of nickel in our steels, the 
 critical transformations are reversible like ordinary steels. That 
 is to say, they occur at nearly the same temperature on heating 
 as the reverse change does on cooling. It is an interesting fact 
 that the steels in which the irreversibility of the transformation is 
 most marked — that is to say, the steels from 12 to 25 per cent, of 
 nickel — have the highest strength and elastic limit; at about 25 to 
 30 per cent, of nickel, where the irreversible transformation is 
 most erratic, and beyond that point, the strength is much lower. 
 The whole interesting question of reversible and irreversible trans- 
 formation is discussed very fully in Dumas's paper No. 153. 
 
 Occurrence of Nickel, — Waterhouse tested his steel contain- 
 ing 3.80 per cent, of nickel and found that a part of the nickel 
 was dissolved in the cementite which had the formula (FeNi)3C. 
 The amount of nickel in the cementite was not, however, as great 
 as that in the ferrite. That is to say, the steel, as a whole, con- 
 tained 3.80 per cent, nickel, while the cementite contained only 
 1.86 per cent., showing that the nickel dissolves more easily in the 
 ferrite than it does in the cementite. 
 
 Micro-structure. — In Figs. 316 to 321 1 show some photomicro- 
 graphs of nickel-steels taken by Waterhouse, and in reference 
 No. 1522 will be found L. Guillet's researches upon this subject. 
 
 Manganese Steel 
 
 We owe the discovery of manganese steel to the untiring in- 
 genuity of Robert A. Hadfield, of Sheffield, England, and its 
 story will be aii inspiration to every inventor, for it resulted in a 
 material whose properties are not only the opposite of what we 
 might reasonably have expected on logical grounds, but whose 
 combination of great hardness and great ductility was hitherto 
 unknown and might readily have been believed to be impossible. 
 Constant study and perseverance must have been the qualities 
 that led to this revolutionary invention, and it has established 
 beyond question the principle that because a given amount of any 
 element produces a given effect upon steels, it does not follow that 
 a different amount will give the same effect in a different degree. 
 Indeed a different amount may give an entirely different, and per- 
 haps an exactly contrary, effect, as is the case of the effect of man- 
 ganese upon steel. 
 
398 THE METALLURGY OF IRON AND STEEL 
 
 When the manganese in steel is over 1 per cent, the metal 
 becomes hard and somewhat brittle, and these qualities increase 
 in intensity with every increase of manganese until, when we have 
 4 to 5.50 per cent, the steel can be powdered under the hammer. 
 But as the manganese is increased from this point, these properties 
 do not increase and when we reach 7 per cent., an entirely new 
 set of properties begin to appear. These are well marked at 10 
 per cent, of manganese, and reach a maximum at 12 to 15 per cent. 
 
 Composition. — Manganese steel usually contains about 12 to 13 
 per cent, of manganese and 1.25 to 2 per cent, of carbon.^ With 
 this amount of manganese the strength and ductility of the ma- 
 terial reaches its maximum. This high carbon has been neces- 
 sary hitherto, because ferro-manganese contains much carbon, 
 which therefore unavoidably finds its way into the steel. In 
 recent years, however, manganese metal, relatively free from 
 carbon, has been made by the Goldschmidt thermit process and 
 otherwise, and this enables manganese steel low in carbon to be 
 made, which is now in process of development and is givingevi- 
 dence of having new and useful properties of its own and of being 
 more easily treated and worked. 
 
 Treatment — After manganese steel has been cast into an ingot 
 or casting, and slowly cooled, it is almost as brittle as glass. But 
 it is then reheated to a temperature of more than 1000° C. (1832° 
 F.) and rapidly cooled by plunging it into- water. The tempera- 
 ture from which it is necessary to quench it can readily be deter- 
 mined, for it must be so high that when the steel is quenched little 
 blue flames of hydrogen will appear on the surface of the water. 
 These are due to the decomposition of water into hydrogen and 
 oxygen by the intense heat of the steel at the moment of touching 
 it. The steel which was very brittle before this treatment is after- 
 ward as ductile as soft carbon steel or wrought iron, while its 
 tensile strength is about three times as great. Thus the sudden 
 cooling, which produces brittleness in ordinary steel, produces 
 ductility in manganese steel. The hardness of the manganese 
 Bteel is about the same in the slowly cooled and in the quenched 
 condition, and is so great that it is not commercially practicable 
 to machine it and there is no method known of making it softer. 
 
 Manganese steel must be heated very slowly and uniformly 
 lest it crack. It is also very difficult to forge it, and this can only 
 be accomplished within a narrow range of temperature above a 
 
 1 Manganese steel that ia to be forged or rolled is made lower in carbon than thia. 
 
ALLOY STEELS 399 
 
 red heat and by beginning with very light taps of the hammer. 
 After a little working it becomes so tough that it can be rolled, 
 although somewhat gingerly. 
 
 Uses. — Manganese steel is used chiefly for the jaws and wear- 
 ing parts of rock-crushing machinery and similar apparatus, for 
 railroad frogs and crossings, for railroad rails on curves, mine car 
 wheels, and burglar-proof safes. Its life in these classes of ser- 
 vice is very many times that of all other kinds of steel, because it 
 is not only extremely hard^but is without brittleness. There is a 
 famous curve on the Boston elevated railroad where carbon-steel 
 rails were worn out in a very short time and the use of manganese 
 steel rails has proved very advantageous and economical. The 
 use of the steel for burglar-proof safes is also very advantageous, 
 because there is no known method of making the steel soft enough 
 to be penetrated by a drill. The uses of manganese steel are 
 limited chiefly because the metal must ordinarily be formed by 
 casting, since machining and cutting to shape is practically out of 
 the question, and forging is difficult. For structural work the 
 advantages of its high combination of strength and ductility are 
 somewhat offset by its low elastic limit, which is only about 35 
 per cent, of its ultimate tensile strength. One peculiarity of man- 
 ganese steel is that when it yields to tensile stresses it is elongated 
 more uniformly over its 'W'hole length than carbon steel, which 
 suffers its greatest elongation near the poitit of final rupture where 
 a certain amount of "necking" takes place. It will foe remem- 
 bered that wrought iron stretches more uniformly over its whole 
 length than steel; manganese steel has this property in a still 
 more marked degree even than wrought iron. 
 
 Critical Changes, — ^The hardness of manganese steel is due, 
 in part, to the hardness of manganese, but still more potently to 
 the fact that the steel is, in the austenitic condition. That is to 
 say, the manganese has reduced the temperature at which the 
 critical changes occur below that of the atmosphere, and there- 
 fore manganese steel consists entirely of austenite. It is, of course, 
 non-magnetic. 
 
 Chrome Steel 
 
 Chromium has the effect of making the critical changes of steel 
 take place at a higher temperature and much more slowly. For 
 the latter reason chromium steels are capable of greater hardness, 
 because rapid cooling is able more completely to prevent the 
 
400 THE METALLURGY OF IRON AND STEEL 
 
 decomposition of austenite . They contain usually 1 to 2 per cent, 
 of chromium and from 0.80 to 2 per cent, of carbon and are used in 
 the hardened state. They are particularly adapted for making 
 armor-piercing projectiles, on account of their hardness and also 
 their very high elastic limit. They are also used for armor plate 
 for the same reason, for parts of crushing machinery, and for 
 very hard steel plate. This latter is not ordinarily used by 
 itself, but is made into 3-ply and 5-ply plate for plows and 
 burglar-proof safes, as described on page 181, for if the hardened 
 chrome steel were used alone, its brittleness would cause it to be 
 shattered. 
 
 Armor Plate. — Krupp armor plate contains about 3.25 pei' 
 cent, of nickel, 1.50 per cent, of chromium and 0.25 per cent, of 
 carbon. 
 
 Automohile Steels. — Chromium from about 0.50 to 2 per cent, is 
 also used for automobile steels where hardness is required, as for 
 instance in gears and other parts requiring great hardness or 
 great strength. For the latter purpose it is more common to use 
 a nickel-chrome steel, and this is often subjected to a double heat 
 treatment or a simple oil tempering. This treatment has the 
 effect of greatly increasing its strength and elastic limit, so that 
 steels of this character will have properties similar to those shown 
 in Table XXX. There cannot be said to be any uniform com- 
 position or method of treatment for automobile steels, and my 
 inquiries among American manufacturers seem to indicate that 
 there is not very much of the high-priced alloy steels used in 
 American cars, except for the frames which, as already stated, 
 are frequently made of 3.50 per cent, nickel-steel. The next 
 most important use is probably for gears of chrome steel of nickel- 
 chrome steel, first case-hardened by cementing with carbon to a 
 depth of an eighth of an inch or so, and .then heat-treated. The 
 composition and treatment of alloy steels used in French auto- 
 mobiles is shown more fully in Quillet's article. No. 1514. 
 
 Self-hardening and High-speed Tool Steels 
 
 Self-hardening Steel. — Self-hardening steel is steel which is 
 hard without being subjected to any heat treatment or other 
 process for making it so. It is steel which cannot be made soft, or 
 annealed, by any process known at present. It is often called 
 "air-hardening steel" because when it cools in the air from a red 
 
ALLOY STEELS 401 
 
 heat or above, it is not soft like ordinary steel, but is hard and 
 capable of cutting other metals. Manganese steel is a typical 
 self-hardening steel and so obviously is any steel which is in the 
 austenitic condition at atmospheric temperatures — that is to 
 say, whose critical temperature is below the atmospheric tempera- 
 ture. All the self-hardening steels are therefore non-magnetic, 
 
 Musket Steel, — The name "self-hardening" steel was first 
 applied to an alloy steel invented by Robert Mushet and which 
 owed its self -hardening properties to the simultaneous presence 
 of both tungsten and manganese. The analyses varied greatly, 
 but were probably limited to between 4 and 12 per cent, of tung- 
 sten with 2 to 4 per cent, of manganese and 1.50 to 2.50 per cent, 
 of carbon. A typical sample and one having excellent qualities 
 contained about 9 per cent, of tungsten, 2.50 -per cent, of man- 
 ganese, and 1.85 per cent, of carbon. This steel is incapable of 
 being made soft by any known process and is non-magnetic. It 
 is one of those curious phenomena met with in the metallurgy of 
 steel, where a combination of two elements will produ"Ge a result 
 entirely -different from anything that might be predicated: Tung- 
 sten does not lower the temperature of the critical change in steel 
 and 2.50 per cent, of manganese has but a slight effect in that 
 direction. Nevertheless the combination of these two reduces 
 the critical point below the atmospheric temperature. 
 
 Mushet steel has been, for many years, a famous tool steel be- 
 cause of its capacity for performing a large amount of heavy 
 cutting work. It is very hard and durable and will retain its 
 cutting edge for a long time and under very severe service. It, or 
 its equivalent, is used very largely at the present time for very 
 heavy, or deep, cuts and especially for cutting extra hard metal, 
 such as the roughing cuts on armor plate and other hard alloys, 
 The cutting speed of which it is capable is not much, if any, 
 greater than ordinary carbon tool steel, but the economy of its 
 use is due to the fact that it will take such deep cuts and last so 
 long without regrinding. 
 
 Other Self-hardening Tool Steels. — The 2.50 per cent, of man- 
 ganese in Mushet steel can be replaced by 1 or 2 per cent, of 
 chromium and again produce a self-hardening tool steel which 
 has the advantageous properties of Mushet steel. This result is 
 even more astonishing than the self-hardening properties of 
 Mushet steel, because chromium has a tendency to raise the 
 temperature at which the critical change comes, and yet the 
 
 26 
 
402 THE METALLURGY OF IRON AND STEEL 
 
 addition of 1 or 2 per cent, of chromium to a tungsten steel, which 
 was not previously self-hardening and whose critical temperature 
 was about 600° C. (1112° F.), reduces the critical temperature to 
 below the atmosphere. We may also replace the 9 per cent, of 
 tungsten in Mushet steel with 4 to 6 per cent, of molybdenum, 
 and it is stated that this latter change produces a self-hardening 
 tool steel which is a little tougher than Mushet steel. 
 
 Taylor and White. — ^Frederick W, Taylor and Maunsel White 
 of the Bethlehem Steel Works experimented for a long period of 
 time with the self-hardening steels existing in 1899 and previ- 
 ously, for the purpose of improving them by heat treatment. 
 The full record of these and other researches were presented by 
 Taylor in his presidential address to the American Society of 
 Mechanical Engineers in 1906, and form one of the most interest- 
 ing records of the kind ever presented to the world. The result 
 of these experiments was to produce a wholly new kind of steel 
 which has fairly revolutionized the machine shop industry of the 
 world. Taylor and White found that by applying a new method 
 of heat treatment to the self-hardening tool steels, they gave 
 them much greater toughness at a red heat, so that they could do 
 their cutting work at a speed so fast that the point of the tool 
 would become red hot with the heat of friction and the great 
 chips of steel, which were thick and heavy on account of the 
 depth of cut which could be made, were raised to a temperature 
 of nearly 300° C. (572° F.). In other words, the steel tool never 
 lost its temper nor its toughness at a red heat. The heat treat- 
 ment which Taylor and White employed consisted in raising the 
 steel almost to the melting-point and then plunging it in a bath of 
 molten lead at a temperature between 700° and 850° C. (1300° 
 and 1550° F.), where it was kept until it was of the same temper- 
 ature as the bath, and then removed and cooled by plunging into 
 oil. They usually followed this cooling by reheating the steel to 
 a temperature between 370° and 670° C. (700° and 1240° F.). 
 
 The first public exhibition of the Taylor and White steel was 
 made at the Paris Exposition in 1900 and created first incredulity 
 and then astonishment. The amount of work performed by a 
 tool was unheard of, as also was the speed at which the tool was 
 made to travel through the metal it was cutting, and the length of 
 time that elapsed before it was necessary to regrind it. It was 
 realized that a new epoch in the tool-steel industry had been 
 inaugurated. The fact that the method of heat treatment used 
 
ALLOY STEELS 403 
 
 by Taylor and White was subsequently shown to be unnecessary 
 and that therefore the manufacture of high-speed steel tools, hav- 
 ing qualities like theirs, was begun by everybody, in no way 
 lessens the credit due them for teaching the world how to produce 
 a new kind of metal and effecting a tremendous decrease in the 
 price of machine work. 
 
 High-speed Steels, — The name ''high-speed steels," was not 
 given by Taylor and White to their product, but has subsequently 
 been adopted for all steels capable of these rapid-cutting speeds 
 which theirs had. Soon after they had shown the world what 
 could be done, it was found that the only heat treatment neces- 
 sary to give the steel its peculiar hardness and toughness at a low 
 red heat was to raise it to a temperature very near its melting- 
 point and then cool it with moderate rapidity, as for example, by 
 holding it in a blast of cold air until it was below a red heat. The 
 essential feature seems to be that the steel shall attain a high 
 temperature which, in many cases, is so great that melted oxide 
 drops from it, and it is almost ready to scintillate — that is to say, 
 it has almost crossed the line Aa in Fig. 252, page 295. After this 
 heating it sometimes suffices to merely allow the steel to cool in 
 air, but in this case its hardness is not as great, and cooling in a 
 stream of air is more usual. 
 
 Composition. — It was also soon found that the composition of 
 the self-hardening steels was not the best one for high-speed 
 steels. Tungsten was the element which gave the steel the prop- 
 erties of hardness and toughness at a red heat. After the peculiar 
 heat treatment had been learned and the presence of manganese 
 or chromium in addition to the tungsten was shown to be unneces- 
 sary, it was found that more durable qualities could be obtained 
 by increasing the percentage of tungsten, and steels have been put 
 upon the market with even as high as 24 per cent, of this element. 
 At the same time the carbon was greatly reduced and at the pres- 
 ent usually varies from 0.40 to 0.80 per cent, in the best high- 
 speed steels.^ 
 
 1 It 13 commonly stated in the trade that tungsten 'will take the place of carbon in producing 
 hardness, but this is not true. It is far more correct to gay that tungsten will assist car- 
 bon in producing hardness and therefore with high tungsten steels we may have lower carbon. 
 This distinction may appear merely academic, but it is well worth recognition by those who 
 expect to make a study of these steels. No amount of tungsten or any other element will 
 make steel hard in the absence of carbon, or even when the carbon is low. The tungsten 
 produces hardness by its effect upon the condition of the carbon — that is by helping to 
 retain the carbon in its solid solution — and not by any effect of its own. It is for this 
 reason that a lesser amount of carbon will produce hardness in the presence of tungsten or 
 some similar agent. 
 
404 THE METALLURGY OF IRON AND STEEL 
 
 It was also found that molybdenum could replace tungsten 
 as far as producing high-speed qualities was concerned, and many 
 believe that the molybdenum steels are more tough and durable 
 than the tungsten steels. Some difficulty was met with at first in 
 working the molybdenum steels as they proved to be seamy and 
 liable to cracks, but this was overcome with experience. The 
 molybdenum steels do not require so high a temperature for heat- 
 ing previous to cooling down in the air blast as the tungsten steels. 
 According to the researches of Carpenter,^ the molybdenum 
 steels should be heated between 1000° and 1100° C. (1832° and 
 2012° F.), while the tungsten steels'must be heated in the neigh- 
 borhood of 1200° C. (2192° F.). Furthermore it only takes about 
 one-half as much molybdenum as tungsten to produce the desired 
 result, which means that there is more iron in the molybdenum 
 steels than in the tungsten steels. In other respects the analysis 
 of the molybdenum and the tungsten steels is about the same, 
 containing usually 0.60 to 0.80 per cent, of carbon and anywhere 
 from zero chromium up to sometimes as much as 4 per cent. 
 Indeed chromium is sometimes recommended as high as 6 per 
 cent, and over, because it gives hardness, but it also reduces 
 toughness. The more durable qualities of the molybdenum 
 steels than the tungsten steels are believed to be due to the 
 larger amount of iron in them and the lower temperature neces- 
 sary for tempering them. Vanadium is also added with beneficial 
 results. 
 
 In some cases molybdenum and tungsten have been used to- 
 gether in high-speed steels. In fact there are at the present time 
 scores of brands and analyses of high-speed steels on the market, 
 made both in America, England, and Europe and the art of 
 manufacturing them is constantly advancing so that no very 
 general results can be quoted. In America the most advantage- 
 ous percentages of molybdenum, 6 to 15 per cent., are patented 
 and although at one time a great many tons of this kind of high- 
 speed steel were manufactured and gave very good results (con- 
 taining usually about 10 per cent, of molybdenum) it can now be 
 made only by one company, so that tungsten steels are more 
 common. 
 
 Forging. — High-speed steels can only be forged at tempera- 
 tures above a bright red heat, that is to say from 1050° to 1 150° C. 
 (1922° to 2100° F.) and higher. 
 
 1 See page 460 of No. 1517. 
 
ALLOY STEELS 405 
 
 Annealing. — ^The heating and annealing of high-speed steels 
 requires a great deal of care. They must be heated up to the 
 annealing temperature (say about 800° C. = 1472° F.) with 
 extreme slowness, and cooled down in lime or ashes or in- the 
 furnace. They are then soft enough to be machined easily, but 
 not as 'Soft as carbon steel. A shorter method of annealing is 
 obtained by a double heating r^ Heat slowly to 700° C. (1300° F.) 
 and cool slowly to 400° C. (750° F.) or lower; then reheat to 
 700° C, hold there for 30 minutes and cool in air. 
 
 Tempering. — The reason tungsten and molybdenum produce 
 in steel the high-speed quality of not losing a temper at a red heat 
 is because of their effect upon the critical temperatures. Their 
 effect seems to be to prolong the critical range of temperatures of 
 the steel on slow cooling; that is to say, instead of the critical 
 range coming in the neighborhood of 690° C, (1285° F.), as with 
 the carbon steels, it begins, when the cooling is slow, at about 
 700° C. and spreads out all the way down to 300° or 400° C. (572° 
 or 752° F.), or even lower. Molybdenum is more active than 
 tungsten in causing this prolongation of the critical range. But 
 if the steels are first heated to a very high temperature (1000° to 
 1100° C. for molybdenum steer and 1200° C. for tungsten = 
 1832° to 2012° and 2192° F, respectively) and then cooled moder- 
 ately fast, this treatment suffices to prevent the critical changes 
 altogether and preserves the steel in the austenitic condition. We 
 know that this austenitic condition is one of hardness and tough- 
 ness and its peculiarity under these circumstances is that it is 
 not transformed into the pearlitic condition until the steels are 
 heated to 650° C. (1202° F.) or thereabouts. 
 
 Magnetic Steels. — Strictly speaking, the steels used for per- 
 manent magnets are not high-speed steels, because, of course, 
 they are never used for cutting work, but as their composition is so 
 similar it seems well to introduce them here. A permanent 
 magnet is made by putting a piece of hardened steel in a magnetic 
 field for a few moments, as, for instance, by winding an insulated 
 wire around it and passing an electric current through. The 
 magnetic force which it obtains in this way will remain in it for a 
 very long period of time. It is found that a steel containing 
 about 4 to 5 per cent, of tungsten and 0.50 to 0.70 per cent, of 
 carbon, if heated to a red heat (say 800° C. =1472° F.) and 
 quenched in water, will retain its magnetism better than ordinary 
 
 * See page 460 of No. 1517. 
 
406 THE METALLURGY OF IRON AND STEEL 
 
 hardened carbon steel. Sometimes about 0.50 per cent, of 
 chromium is added to the alloy ajso, which increases the perma- 
 nency but decreases the magnetic force, the reason being no doubt 
 that there are more gamma and less alpha molecules. (See p. 
 312). 
 
 Silicon Steels 
 
 The genius of Hadfield has also given us a silicon steel alloy 
 of importance and usefulness. In 1888^ Hadfield investigated 
 many alloys of iron and silicon and although these showed some 
 remarkable properties, especially in the matter of tempering and 
 cutting qualities, they did not lead to any alloy steels that were 
 produced in abundance. At a later period, however,^ Hadfield 
 developed a silicon steel which, after a double-heat treatment, 
 showed some truly remarkable magnetic qualities. It had 
 always been believed that pure iron had the highest magnetic 
 force and permeability of any known substance or of any combi- 
 nation that could be produced, but Hadfield's new silicon steel, 
 whose composition and treatment is shown below, had not only a 
 greater magnetic permeability than the purest iron but al^o, 
 characteristic of silicon steels, it had a high electrical resistance. 
 Its hysteresis is, of course, low, this property always accompanying 
 a high permeability. It is therefore a very valuable material for 
 use in electromagnets, and in electrical generating machinery is 
 the most efficient material known. Its high magnetic permeabil- 
 ity gives high motor efficiency, and in addition its high electrical 
 resistance reduces the "eddy currents" which are a source of 
 waste^. 
 
 Composition and Treatment — Hadfield's patent covers silicon 
 from 1 to 5 per cent., but the alloy which he recommends con- 
 tains 2.75 per cent, of silicon and the smallest possible amounts of 
 carbon, manganese, and other impurities. Before the steel is 
 ready for use, it is subjected to a double-heat treatment by first 
 heating it to between 900° and 1100° C. (1652° and 2012° F.) and 
 cooling quickly, and then reheating to between 700° and 850° C. 
 (1292° and 1562° F.) and allowing to cool very slowly. In some 
 cases his second cooling has been extended over several days. 
 He finds the best results by heating first to 1070° C. (1958° F.) and 
 cooling quickly to atmospheric temperature and then heating to 
 
 1 See No. 167. 
 
 ^ U. S. Patent 12,691. September 3, 1907. 
 
ALLdY STEELS 407 
 
 750° C. (1382° F.) and cooling slowly, after which he sometimes 
 again reheats to 800° C. (1472° F.) and then cools slowly. 
 
 Vanadium Steels 
 
 Vanadium steels are still in their infancy, for although the 
 element has been used in steel metallurgy for many years, it is 
 only recently that any important development work has been 
 carried on. The results, however, are so very encouraging that 
 we may expect great extension of their employment and impor- 
 tant progress in their metallurgy. With the single exception of 
 carbon no element has such a powerful effect upon steel as vana- 
 dium, for it is only necessary to use from 0.10 to 0.15 per cent, in 
 order to obtain very powerful results, while 0.30 per cent, should 
 probably not be exceeded so far as present knowledge indicates. 
 In addition to acting as a very great strengthener of steel, es- 
 pecially against dynamic strains, vanadium also serves as a 
 scavenger in getting rid of oxygen and possibly nitrogen. It is 
 also said to decrease segregation, which we may readily believe, 
 as most of the elements which quiet the steel have this effect. 
 Vanadium steel also has the advantage of welding readily. 
 
 The effect of vanadium is shown very well by Tables XXX and 
 XXXII, ^ and it will be seen how efficiently this material resists al- 
 ternating stresses and the other forces producing fatigue. It is to 
 be observed that vanadium isespecially advantageous when added 
 to nickel and to chromium steels, greatly increasing their strength, 
 toughness, and temper. In this connection it is important to 
 note that the nickel-vanadium steels have better quality when 
 the carbon is low, especially if they are to be subjected to heat 
 treatment. 
 
 It would seem that vanadium should have especial advantage 
 in high-speed tool steels, but, strange to say, the results have not 
 always been favorable. Nevertheless experiments in this direc- 
 tion continue. 
 
 Manufacture, — Pure vanadium has a very high meHing-point 
 and the element is therefore added to steel in the form of ferro- 
 vanadium. This alloy should be added to the bath about two or 
 three minutes after the manganese used for recarburizing, because 
 if the vanadium is added before the manganese, it is wasted by 
 oxidation. It therefore should be added always under reducing 
 
 ^For which I am indebted to the American Vanadium Co. 
 
408 
 
 THE METALLURGY OF IRON AND STEEL 
 
 
 
 
 ^ 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 s 
 
 
 
 
 
 1 
 
 rn 
 
 1 
 1 
 
 1 
 
 S 
 
 
 g 
 
 
 1 i 
 
 . 
 
 i-q 
 
 
 
 M 
 
 
 
 < 
 
 
 
 . . . . . 
 
 to ■ ■ • o • 
 
 o 
 
 
 8S»SIili» 
 
 »0 CO rH 
 
 
 ■ <M 
 
 
 ■ 
 
 
 
 
 
 
 
 iO 
 
 
 
 • to 
 
 
 
 
 TOi^aojL 
 
 CO <N CO 
 
 
 • ■* 
 
 
 : 
 
 
 
 
 
 
 
 •^ 
 
 
 
 
 
 
 9l3aB^q3TH 
 
 00 <M C4 
 tH r-i rH 
 
 
 
 
 
 o 
 
 I-H 
 
 
 
 
 
 
 00 
 
 I-H 
 
 
 
 
 
 
 1 
 
 S8ai!),e JO 
 
 8S8 
 
 
 
 lO 
 
 g 
 
 
 
 
 
 
 
 s 
 
 
 
 ■^ 
 
 
 
 
 HOT^BTUajIV 
 
 ^ rH rH 
 
 
 
 
 ^H 
 
 
 
 
 
 
 
 I-H 
 
 
 
 i.^ 
 
 
 ?1 
 
 /3 
 
 B^oiq 
 %0'edvai 
 
 
 »0 lO 
 <N CO 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 § 
 
 
 
 
 
 
 >^ 
 
 spnnod 
 
 
 
 
 "3 
 
 
 
 
 
 
 
 
 
 
 
 P 
 
 
 ■•^j 'B^oBdrai 
 
 U3 CO -* 
 »H .-1 
 
 
 
 CO 
 
 w 
 
 
 
 
 
 
 
 CD ■* 
 
 
 t> o 
 
 rH .-1 
 
 t— 1 
 
 r^ 
 
 tnn|npuaj 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P 
 
 ^ 
 
 
 
 
 
 
 
 
 
 <1 
 
 H 
 
 •(ja j3d 
 
 
 
 
 ^ 
 
 X- 
 
 '■eajB JO 
 
 O b- 00 N 
 
 i-t Oi Ui <o 
 
 oONuit^^osco-^eoososo 
 
 ^ 
 
 o ■* »o © 
 
 <0 Ui to Ui 
 
 *OCDCD"3»OCO»0'*COCOCOCO 
 
 < 
 > 
 
 
 iioi!joBj<^aoo 
 
 
 
 
 >> 
 
 
 lO 
 
 
 
 Q 
 ^ 
 
 S 
 
 ^0 jod '^^g ui 
 
 O N ■* ^ 
 
 CO OS N W 
 
 OI>-00l^«^-*C0OO»OM 
 
 noi!ji3auoia 
 
 »0 CO CO CO 
 
 CO <N cq « 
 
 (Mr-4i-Hi-li-l<-IC0i-IC0<-l'<^(N 
 
 <! 
 
 *q,p jad 
 
 0> ■* CD 
 
 N ^ eo CO 
 
 W U3 -H -^h 
 (O CO t^ N 
 
 OS 00 l> U5 00 O 00 
 CDiOC<l»OTl<COCONiHeOOt^ 
 
 h-^ 
 
 
 oh^H 
 
 O U3 U3 I> 
 
 ® t> t- 00 
 
 ooosososososocot-osoob- 
 
 H 
 
 
 
 
 
 
 
 
 
 
 W 
 
 
 
 o oT bC i>r 
 
 o o o o 
 
 oooooooooooo 
 
 
 •tn 'bs J3d 
 
 00-** CO o 
 
 OOOOOiOCOOOOOaO 
 
 H 
 
 
 'sqx 'q^3u3j^s 
 
 O «3 (N O 
 CD l>-" ci" lO 
 
 00WOOOI>-C0tJ400O0100 
 I< O" 00 t-T (N N l> (N ■* »0 Ui o" 
 
 ;< 
 
 
 ^U^nax 
 
 lO <© 00 t* 
 
 OS 00 OS <N 
 
 <NC0C0-*T-(C0Q0c0tHC0U3O 
 
 
 
 
 T-i 
 
 i-lrHrHrHCqtN rtr-l« i-t 
 
 >^, 
 
 
 o o o o 
 
 o o o o 
 
 oooooooooooo 
 
 O 
 PQ 
 
 
 •m -bs 
 
 lO 00 b- o 
 
 I> O rt o 
 
 OOOIOOOCOOCOOOSOS 
 
 
 jad •eqt 
 
 CO O M CO 
 
 .-T oo' oj" iC 
 
 »C ■<1< W o 
 
 CO ■*" r-T CO 
 
 .-HO»Ob-OOOOSOOt-CO 
 o" ■<* l> O" O' -dT !< r^ N ^" -* 00 
 r-tCi)(NTtlO(N'*O00N'<*it* 
 
 i-tiHi-HrHCqCq rH (N 
 
 
 
 '*!™u oj^BBia 
 
 CO CO -^ »o 
 
 CO CO t> O 
 
 i-H 
 
 
 
 ■ ■ o» 
 
 
 
 O 
 
 1 
 
 pnoyN 
 
 
 
 CO 
 
 
 
 
 
 * 
 
 
 
 • 
 
 
 
 
 ^/^ 
 
 
 
 
 
 I> 
 
 
 
 lO 00 
 
 
 (M 
 
 
 
 Uinip'BU'B^ 
 
 
 
 
 
 
 -S «; ^-1 
 
 
 ^H 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 -g £ 
 
 
 
 
 
 raniniojqQ 
 
 
 
 
 o 
 
 
 
 o >o 
 
 
 o 
 
 CO 
 
 
 fe 
 
 asaa'BSn'Bpg; 
 
 eo Ui Ui Ui 
 
 
 
 •o o 
 
 CO OS 
 
 
 ^ 
 
 
 ^ 
 S 
 
 1 
 
 
 
 
 
 ll 
 
 !> 
 
 
 
 
 
 aaqj-BQ 
 
 
 
 
 s^ 
 
 
 Ui 
 
 l-l 
 
 
 ^ 
 
 
 
 N IN 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 m ^ w 
 
 O 
 
 
 'C 
 
 - 
 
 
 +- 
 
 1 
 
 
 
 
 
 "« 
 
 3 
 
 
 V 
 
 
 1 
 
 M 
 
 
 "i < 
 
 
 
 og 
 
 
 ■a rt-S 
 
 1 
 
 
 g 
 
 2 
 
 ^ 0) 
 
 
 
 ?=-- 
 
 
 s a 
 
 ><1 
 
 
 "S 
 
 f- 
 
 •E ^ 
 
 
 
 ^lo 
 
 
 
 M 
 
 s 
 
 m 
 ^ 
 
 
 1 
 
 1 
 
 
 1 
 
 - 1 
 A E 
 
 ll 
 
 '1 
 
 
 
 1 
 
 ■1 
 
 1 
 
 
 1 
 
 
 1 
 
 i 
 
 
 
 
 
 13 
 
 £ 
 
 
 
 
 
 
 
 1 
 
 1. a 
 
 c 
 
 a 
 c 
 
 
 
 CO 
 
 bD 
 
 tH ^ 
 
 Q ^ 
 
 •a 
 
 & 
 
 fl 
 
 
 
 -»3 
 
 1 
 
 . 
 
 5 
 
 ■43 
 
 
 rt 
 
 ■u 
 
 «S 
 
 
 — 
 
 01 
 
 ffl 
 
 
 u 
 
 to 
 
 EO 
 
 
 in 
 
 fl 
 
 •s 
 
 
 1 
 
 5 
 
 1 
 
 
 n 
 
 "C 
 
 ■s 
 
 
 1 
 
 H 
 
 9. 
 
 
 W 
 
 . 
 
 ff> 
 
 
 M 
 
 1 
 
 rS 
 
 
 
 1 
 
 ,.M 
 
 
 ^' 
 
 o 
 
 0) 
 
 
 
 T3 
 (I) 
 
 01 
 
 
 2 
 
 1 
 
 1 
 
 
 a 
 
 13 
 9. 
 
 
 
 
 ji 
 
 ^ 
 
 +3 
 
 fl 
 
 
 Tl 
 
 }0 
 
 s 
 
 
 fe 
 
 J 
 
 
 
 OS 
 
 T3 
 
 In 
 
 
 
 § 
 
 :S 
 
 
 ^ 
 
 '-' 
 
 ■£ 
 
 
 2 
 
 5 
 
 iH 
 
 
 
 "3 
 
 TS 
 
 fl 
 
 cT 
 
 
 c 
 
 03 
 
 -* 
 
 
 tE] 
 
 rr 
 
 u 
 
 
 n 
 
 Ul 
 
 13 
 
 
 
 
 
 
 
 00 
 
 
 h* 
 
 4S 
 
 .?^ 
 
 u 
 
 
 1 
 
 <e 
 
 :^5 
 
 
 
 
 
 fl 
 
 ^' 
 
 1 
 
 a 
 
 T? 
 
 W 
 
 <D 
 
 ;fl 
 
 1 
 
 
 
 
 D 
 
 C3 
 
 CD 
 
 5d" 
 
 SS2 
 
 o 
 
 J3 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 U 
 
 1 
 
 -13 
 
 ■Ri 
 
 m 
 
 s 
 
 43 
 
 3 
 
 -5 
 
 n 
 
 ■1 
 
 fl 
 
 T3 
 
 bo 
 
 ^ 
 
 T^ 
 
 M 
 
 ta ja 
 
 s 
 
 S 
 
 k 
 
 1 
 
 bD 
 
 fl 
 
 IS 
 
 
 a 
 
 2 
 
 
 
 m 
 
 O 
 
 p 
 
 
 a 
 
 O 
 
 « 
 
 -3 
 
 a 
 
 «? 
 
 £ 
 
 ^ 5 
 
 ■§ 
 
 
 
 
 n 
 
 5; 
 
 s 
 
 s 
 
 13 
 
 m 
 
 
 <! 
 
 m 
 
 -»3 
 
 
 
 
 
 
 
 
 o 
 
 ttH 
 
 
 o 
 
 a 
 
 S 
 
 ^ 
 3 
 
 •3 
 
 o 
 
 
 1 
 
 fl 
 
 5 
 
 2 
 
 •§ 
 
 cq 
 
 1 
 
 
 •9 
 
 
 Ui 
 
 s-^ 
 
 o 
 
 
 PS 
 
ALLOY STEELS 
 
 409 
 
 H 
 
 P 
 
 l-H 
 
 P 
 > 
 
 o 
 
 
 
 
 .. "3 
 
 1 
 
 
 
 43 <- 
 
 
 
 S r 
 
 
 
 » N O O (N 
 
 
 H 
 
 j^ ^ (M CO rH 
 
 
 ft 
 
 lUl 
 
 
 ^ 
 
 
 
 
 
 
 . ic O 
 
 
 
 -t^ Ui ■ 
 
 
 M 
 
 d - rH U3 
 
 S 1 1 N 
 
 
 O 
 
 O U3 O - 00 
 
 
 t^ -^Ji 00 rH ,H 
 
 
 •■ 
 
 fS . « d d 
 
 
 Q 
 
 ■ s § 1 
 
 
 ^ 
 
 Carbon 
 Mangan 
 Chromii 
 Vanadii 
 
 
 
 , . "3 O 
 
 
 
 -*i io ■ 
 
 
 rH 
 
 d • rH U3 
 
 
 
 "l^^"^" 
 
 
 i 
 
 j^ -* 00 rH rH 
 
 
 Q 
 
 P 
 
 Carbon 
 
 Manganese 
 Chromium. 
 Vanadium. 
 
 
 
 
 
 * . O O O <£> 
 
 
 
 4i (N -* 00 rH 
 
 
 
 s • • ■ ■ 
 
 
 o 
 
 01 
 
 Per c 
 arbon 
 
 anganese. 
 iromium.. 
 anadium.. 
 
 !2: 
 o 
 
 $ 
 
 M 
 
 
 O SOP> 
 
 
 . o o 
 
 03 
 
 
 +j (M -* 
 
 o 
 
 
 S * i" 
 
 Ph 
 
 
 s 
 
 O O N 
 
 1^ 
 
 p:) 
 
 tl 
 
 CO lO rH 
 
 o 
 
 ft 
 
 
 JM 
 
 
 >> 
 
 
 
 H 
 
 Carbon 
 
 Mangar 
 Chromi 
 Vanadi 
 
 
 
 . O O 00 
 
 
 
 4^ M »0 rH 
 
 
 « 
 
 S i' i" o r 
 
 
 ^ 
 
 a >o o -CD 
 
 
 ^^ M ■* W rH 
 to ' ' ' ■, ' 
 
 
 ^" 
 
 p 
 
 Carbon .... 
 Manganese 
 Chromium. 
 Vanadium. 
 
 
 
 . O O 00 
 
 
 
 +3 CO "3 r-t 
 
 
 d 
 
 O iC O O (O 
 
 
 SJ '^. ^- -I "• 
 
 
 
 p 
 
 Carbon 
 
 Manganese 
 Chromium 
 
 Vanadium. 
 
 
 ft 
 
 
 
 ^ O O 00" 
 
 
 
 ^ CO lO rH 
 
 
 iH 
 
 l-ii^'i 
 
 
 o 
 
 ^ C^ '^ . ^ 
 
 
 iz; 
 
 f^ ' ' . ' 
 
 
 < 
 
 rbon . . . 
 nganese 
 
 romium . 
 nadium . 
 
 
 
 6 
 
 ie> 1 
 
 
 bO <o . 
 
 
 S d ® 
 
 
 ^11 
 
 
 rt " « 
 
 
 ^^a 
 
 
 
 
 " <a g t3 "S 
 
 
 OS -S^ 1^ S 
 
 
 « « « ft 
 
 
 d «^ 43 
 
 
 5 -C S 
 
 
 ^- « i . 
 
 
 - o 
 
 
 prings 
 mobile 
 and 1 
 work. 
 
 
 02 
 
 
 fc d 'd 
 
 
 •Sal 
 
 
 
 
 ^ oT - 
 
 
 V <n EQ 
 
 
 S S r- 
 
 
 ^^s 
 
 
 lid 
 
 lilw 
 a r 
 ins. 
 
 
 O S ^ ft 
 
 
 03 
 
 
 ^ ^ bO 
 
 
 ediate 
 ver 
 for ca 
 
 holdin 
 
 etc. 
 
 03 
 
 Iz; 
 
 o 
 
 [nterm 
 steel 
 Useful 
 axles, 
 
 bolts. 
 
 l-H 
 
 
 H 
 
 
 g 
 
 and 
 n is 
 Lent. 
 
 h-1 
 
 -S .2 a 
 
 . o e 2 
 ,M 2 Q a — - 
 
 
 P 03 „ « J ■ 
 
 
 i3 a.S ^^ 
 
 
 .^&^°& 
 
 
 4^ 
 
 
 d >> 
 
 
 _g f— 1 
 
 
 on d 
 
 
 O T3 
 
 
 " d . 
 
 
 .a p'^ 
 
 
 ^-s.i 
 
 
 
 ^1^^ 
 
 
 d 
 
 
 o 
 
 
 • iH 
 
 
 m 
 
 
 5.2 . 
 
 
 
 
 i a- .1 
 
 
 m m . ft 
 
 
 r^ fl a -i<J 
 
 
 d 03 »4 d 
 
 
 Ea fc, =3 c3 
 
 
 5h ^c5 
 
 
 -S 
 
 
 les. 
 rods, 
 ain ro 
 
 es. 
 
 
 Mill 
 
 43 o d Sf !- 
 
 
 -i £ =3 .9 S 
 
 
 ^ i 1 1 1 
 
 
 
 : UiSQft 
 
 
 I 
 
 1 s 
 
 
 Quench in oil from 
 900° C. and draw 
 back at 400° C. 
 to 450° C. in lead 
 bath preferred. 
 Cool in air. 
 
 
 Anneal at 800° C. 
 for one hour. 
 Cool slowly, tak- 
 ing great care not 
 to chill or to 
 pass from 800° C 
 to 600° C. too 
 quick. 
 
 iz; 
 
 g 
 i 
 
 
 
 
 Quench from 950° 
 C. in lard oil and 
 let down at 360° 
 C. for i to i hour 
 preferably i n 
 lead bath, cool 
 in air. 
 
 
 Quench from 900° 
 C. in lard or fish 
 oil and anneal at 
 550° C. for i to 2 
 hours, according 
 to size in air. 
 
 
 Anneal at 800° C. for 
 one or two hours. 
 Cool in air or ashes 
 according to nature 
 of piece. 
 
 a 
 
 ^ 
 
 Ui z ~ 
 
 SS ^ '=* 
 
 " 3 
 
 5 = 
 
 pa 
 
 ffl E " 
 
 
 d" d 
 
 
 ja d ■- 
 
 [fi 
 
 »•'"&? 
 
 1^^ 
 
 % 
 
 03 o O 
 
 H 
 
 g-33 
 
 P^ 
 
 S ^fe? 
 
 
 ^ = Ui 
 
 
 f^Q® 
 
 
 - 
 
 
 ^ - T3 
 
 
 ft a g 
 
 
 ifs 
 
 
 XI <]3 " 
 
 
 ft Jl T3 
 
 
 
 
 d >i d 
 
 
 tn e3 
 
 
 a s "S 
 
 
 £ d d 
 
 
 
 
 
 
 ^ sa ua 
 
 
 •^ 'S % 
 
 
 ^ iu 
 
 
 S -a -d 
 
 
 p, -*^ 43 
 
 - 
 
 " « CO 
 
 
 
 fl) 
 
 
 <^^ 
 
410 
 
 THE METALLURGY OF IRON AND STEEL 
 
 rn 
 
 
 yA 
 
 
 H 
 
 
 |x) 
 
 
 H 
 
 
 m 
 
 
 fA 
 
 © 
 
 H 
 
 S 
 
 W 
 
 "Irt 
 
 H 
 
 pj 
 
 O 
 
 -^ 
 
 
 a 
 
 fl 
 
 Tr 
 
 ^ 
 
 < 
 
 
 >| 
 
 2 
 
 1— I 
 
 a 
 
 P 
 
 03 
 
 < 
 
 :S 
 
 ^ 
 
 ««-• 
 
 < 
 
 o 
 
 > 
 
 uo 
 
 
 
 ^ 
 
 '53 
 
 < 
 
 ^ 
 
 O 
 
 (i> 
 
 
 i 
 
 H 
 
 *3 
 
 ^ 
 
 CO 
 
 O 
 
 ^ 
 
 
 
 H 
 
 •^ 
 
 H 
 
 cr 
 
 
 Cl> 
 
 hJ 
 
 fH 
 
 ^ 
 
 s 
 
 hH 
 
 bf) 
 
 ;?; 
 
 a 
 
 < 
 
 ^ 
 
 w 
 
 o 
 
 o 
 
 a> 
 
 S 
 
 
 p^ 
 
 +3 
 
 o 
 
 iii 
 
 
 sS 
 
 CO 
 
 03 
 
 H 
 
 U 
 
 2 
 
 S 
 
 t:? 
 
 P, 
 
 CO 
 
 ^" 
 
 [±j 
 
 a. 
 
 p^ 
 
 •m 
 
 
 
 
 "e* 
 
 1 
 
 rCl 
 
 
 
 W 
 
 O 
 
 d 
 
 3 
 
 H 
 
 
 Kl 
 
 
 CQ 
 
 
 <J 
 
 
 H 
 
 
 
 § '^ CO 
 
 "^ S ■-§ ^ ^ 
 
 ^1 H 
 
 
 ^ '^ ft 
 > J^ & 
 
 - H 
 
 o 
 
 oil -^ 
 
 o ■ 
 
 X 
 
 I 
 
 p 
 
 o * " 
 
 t^ CD ^H 05 i-H 
 
 : OS w CO 
 
 00 b- o 00 eq 
 
 1^-^ 00 <N W3 
 
 O CD CO tH Tt* 
 
 CO CD -5j< 00 CO 
 j^ lO CO lO 
 00 
 
 ■* o 
 
 w o 
 
 I-H O 
 
 CO o 
 
 » 
 
 o 
 
 00 C^ (N .-H N 
 
 CO O 
 
 - CO 
 
 T3 
 O 
 
 eg 
 ft 
 
 13- 
 
 Pi 
 a 
 
 o 
 
 bC 
 
 ca 
 
 bO 
 0> 
 
 CT* :-3 
 
 ft -*^ 
 
 (-1 
 
 '3 S 
 
 >^3 
 
 .*-• flj 
 
 CO ««-i 
 
 H ^ 
 
 o Xi 
 
 ft a 
 g « 
 
 c3 
 
 =1 
 
 o 
 
 §1 
 
 o 
 
 «3 «, - 
 
 bc £; .2 
 
 S -g § d fc 
 
 T3 
 
 a 
 
 hfl 
 
 o3 
 
 ft 
 J 
 
 a 
 
 T3 
 
 o 
 d 
 
 d 
 o 
 
 -a 
 
 s ^ 
 
 bO or >a 
 
ALLOY STEELS 411 
 
 conditions. Beyond this there is no special difficulty in manu- 
 facture; as the amount of alloy used is so small and as it dis- 
 tributes itself readily through the metal and does not segregate. 
 
 Treatment — Vanadium steels must be heated gradually but 
 are forged without difficulty, although they must be worked a 
 little tenderly at first. Like all steels, they must not be forged at 
 too high a temperature, and like all alloy steels, they become 
 even more brittle when forged below a black heat than carbon 
 steel does. 
 
 Uses, — It- is advantageous to use vanadium steel for prac- 
 tically every purpose^that will stand the additional price, which is 
 about the same as that of 3.50 per cent, nickel-steel. It is, of 
 course, especially useful for all purposes where strength and light- 
 ness are desired, such as springs, axles, frames and other parts of 
 railroaij rolling stock and automobiles. It would also seem to 
 have special advantages for shafts and other ratating parts. It 
 also does very well for case-hardened gears, on account of the 
 good combination of hardness and toughness and -has been used 
 to some extent for steel castings where resistance to vibration 
 has been demanded. 
 
 Vanadium has also been added to cast-iron on account of its 
 ability to remove oxygen and possibly nitrogen (if any is ever 
 present), and because it makes the metal more fluid and tougher. 
 It is said also to make chilled cast-iron rolls more durable and 
 more ductile. 
 
XVI 
 
 THE CORROSION OF IRON AND STEEL 
 
 Iron offers so little resistance to rusting or corrosion that there 
 are almost no circumstances of service in which it can safely be 
 placed without some means of protection from the elements. 
 Certain parts of machinery, in situations where rusting is not 
 very rapid, and where the metal will receive constant cleaning and 
 oversight, are used without any protective coating, but structural 
 work, in and outside of buildings, tin roofs, ^ wire fences, pipes, 
 and other metallic structures, all require to be protected by 
 some coating, such as paint, galvanizing,^ tinning, nickel plating, 
 oxidizing, etc. Boiler tubes and the inside of boilers, tanks, and 
 pipes it is usually impossible to protect by paint, galvanizing, 
 etc., and there is an annual loss of doubtless many thousands of 
 tons of iron and steel from the decay of these classes of articles 
 alone. 
 
 The Cause and Operation of Corrosion 
 
 The brown powder with which we are all far too familiar under 
 the name of rust is a hydroxide of iron — ^ferric hydro^dde, 
 FeOgllg, It is formed wherever iron is exposed to the action of 
 water and air. Neither dry air nor water free from oxygen has 
 any effect upon it alone, but as air is always moist and ordinary 
 water contains some oxygen in solution, the conditions for 
 corrosion practically always prevail against the iron in service. 
 The alternate attack of oxygen and water within a brief period is 
 far more destructive than the attack of either one. For example, 
 heavy rains, the splashing of water intermittently upon piers 
 and columns, the rise and fall of the tide, etc., corrode the metal 
 much faster than exposure all the time to either damp air or 
 oxygenated water alone. Acids, while not essential to corrosion, 
 greatly hasten its action, so that the presence of carbonic acid in 
 the air, sulphurous, sulphuric, and hydrochloric acids in the 
 
 1 Tinplate consists of a thin sheet of steel or wrought iron covered with metallio tin. 
 ^ I.e., coating with metallic zinc. 
 
 412 
 
THE CORROSION OF IRON AND STEEL 413 
 
 smoke from locomotives and other fires, all greatly increase the 
 speed of rusting. It seems evident also that at least some weak 
 electrolysis is essential for any corrosion to occur. I shall show 
 later that this electrolytic action will always take place where 
 iron and water are in contact with each other, but it is very 
 small in amount under these conditions. Where a greater 
 electric force exists, as for example where pipe lines run close to 
 trolley tracks and receive a leak of electricity so as to carry a part 
 of the current, the electrolytic action is increased and corrosion 
 much hastened. As the use of electricity to-day is far more 
 general than ever before and the amount of coal burned and con- 
 sequently the amount of corrosive gases in our atmosphere is 
 much larger, it follows that the question of corrosion is of con- 
 stantly increasing importance. Affairs have indeed reached such 
 a state that the subject is now occupying a great deal of attention 
 from metallurgists and engineers in the hope of getting better 
 means of protection. Pipe lines embedded in the earth, which 
 are alternately wet and dry, supports in tunnels, subways, and 
 other damp places, wire fences and tin roofs are all suffering 
 severely from decay. 
 
 Theories of Corrosion. — In an able research,* AUerton S. Cush- 
 man has studied the current theories that have been advanced 
 to account for the corrosion of steel. He shows that the theory 
 that carbonic acid, or some other acid, is necessary is wrong, and 
 that corrosion will take place even in weak alkaline solutions. 
 He also shows that hydrogen peroxide is not the agency pro- 
 ducing the rust. By means of a series of extremely careful and 
 accurate experiments he demonstrated the probability, if not, 
 indeed, the certainty, that the two factors without which the 
 corrosion of iron is impossible are electrolysis and the presence 
 of hydrogen in the electrolyzed or "ionic" condition. In brief, 
 it is the ions of hydrogen which first cause the metal to pass in 
 solution: 
 
 Ions of hydrogen are electropositive to iron, and when the reac- 
 tion takes place, they transfer their electric charge to the metal. 
 This is, of course, an electrolytic action and the hydrogen which 
 takes part in it is converted from the electrolyzed (or ionic) 
 condition to the atomic (or gaseous) condition. Therefore it is 
 
 1 See No. 160. 
 
 ^ 4H+20=2H2O in the electrolyzed condition. 
 
414 THE METALLURGY OF IRON AND STEEL 
 
 evident that the action must take place in the presence of oxygen, 
 or some other oxidizing agent that will complete the electrolysis, 
 else the formation of ferrous , hydroxide will soon cease. This 
 explains why the presence of oxygen so greatly increases the 
 corrosion of iron, although it is not the oxygen itself which, after 
 all, is the cause of the primary attack. 
 
 Unfortunately traces of ionic hydrogen are always present even 
 in the purest water, and larger amounts in ordinary water. 
 Substances which increased the hydrogen ions, such as oxides, 
 promote the rusting, and the same may be said of anything which 
 increases the electrolytic action, while substances which restrain 
 the formation of hydrogen ions will decrease corrosion. Indeed, 
 dipping a piece of bright iron into a solution of potassium bichro- 
 mate and then wiping if off will put it into an apparently oxidized 
 condition in which it will resist corrosion for days or even weeks. 
 Cushman, therefore, suggests dissolving a small amount of chro- 
 mic acid or potassium bichromate in boiler waters in order to 
 restrain corrosion of the metal. 
 
 RtisL — ^Ferrous hydroxide, Fe02H2, is soluble in water and its 
 formation and solution is the first step in the production of rust. 
 Because of its solubility, however, it does not ordinarily make 
 its presence known until a further reaction occurs: 
 
 2 FeOJI^+B.^O + = 2 FeOgHg. 
 
 The rust, FeOgHg, so formed precipitates from solution. 
 
 Segregation. — Evidently anything that increases the electro- 
 lytic activity will increase the attack by hydrogen and therefore 
 the formation of rust. Unfortunately even the purest piece of 
 iron will show differences of electric potential at different parts 
 and therefore produce an electrolytic effect. When the metal is 
 impure or is badly segregated, these differences in potential will be 
 quite large, and when several pieces of steel are joined together, 
 as in a bridge or other structure, the difference in potential be- 
 tween the different parts may be great. It is probable that each 
 of the different microscopic constitutents of iron and steel has a 
 different electric potential and therefore either assists or retards 
 the progress of rusting. Also scale, the slag in wrought iron, etc. 
 Self-protection from Corrosion. — It is generally believed that 
 certain constituents in iron and steel assist in protecting underly- 
 ing layers of the metal from attack. For example, the graphite 
 which forms such a large proportion (say 10 per cent, or more) 
 
THE CORROSION OF IRON AND STEEL 
 
 415 
 
 of "the volume of cast iron, slag which forms nearly 4 per cent, of 
 the volume of wrought iron, and cementite in steel, all corrode 
 less rapidly than the pure metal, and it is probable that they are 
 beneficial in protecting it. It must not be forgotten, however, 
 that they likewise cause a difference in potential and to that ex- 
 tent probably tend to hasten corrosion. Their net effect can be 
 
 FIG. 325.— CORRODED STEEL PLATE. 
 
 learned only by experiment. Scale or foreign substances on the 
 surface of the metal would also produce large differences in 
 potential. 
 
 Relative Corrosion of Iron and Steel. — It is generally believed 
 that cast iron corrodes less rapidly than either wrought iron or 
 steel and, for. this reason, cast-iron pipe is greatly preferred for 
 city water mains and like uses where great strength is not re- 
 quired. The belief is a reasonable one, since it might well be 
 expected that the presence of graphite would be a protection, but 
 
416 THE METALLURGY OF IRON AND STEEL 
 
 it is only right to remark that the theory rests upon no experi- 
 mental evidence and that there are other circumstances connected 
 with cast-iron pipe which may not have given rise to the belief 
 but which are not inherent in the material itself. These condi- 
 tions are (1) the common practice of dipping cast-iron pipe in 
 asphaltum, paint, or some similar protecting material, before 
 it is sold and thus putting it in a condition to be in service for a 
 long time before the metal itself is subjected to the corrosive 
 influences. (2) When the iron is cast in the sand there seems to 
 take place some union between the metal and the inside surface of 
 the mold, whereby a very resisting silicibus coating or "skin" 
 is formed on the casting. Some believe that after this skin is 
 worn away, cast iron corrodes as rapidly as wrought iron or 
 steel. (3) Cast-iron pipe is thicker than the same diameter of 
 wrought iron and steel pipe, because of the difficulty of making 
 castings with thin sections of metal; tlierefore such pipe would 
 remain in service much longer than steel or wrought iron, even 
 though the rate of corrosion were the same. I have cited the above 
 facts not to argue against the belief in the slow corrosion of cast 
 iron, but merely to explain the situation as it exists, for we have 
 as yet no scientific data upon which to form an opinion either way. 
 
 Wrought Iron vs. Steel. — Th^re is also a very prevalent and 
 widespread conviction that steel corrodes much more rapidly 
 than wrought iron. This opinion too rests upon no very exact 
 experimental evidence, although there are not incidents lacking 
 which make in favor ef it, while others make in the opposite 
 direction, but it seems to be based principally upon the fact that 
 corrosion is very much more rapid to-day, when steel is the 
 world's great metal, than it was some years ago when wrought 
 iron was chiefly used! But I have already shown that conditions 
 to-day are more conducive to rapid corrosion than they were at 
 any previous time.^ 
 
 Opposed to this popular belief is the result of a great many 
 scientific tests which have shown, in almost every case, that the 
 
 1 Indeed I hardly think that this argument ought to have anything like the weight that is 
 popularly given to it. It is not at all uncommon to hear the statement that wrought 
 iron made 20 years ago is still in service in places and conditions alongside of steel which 
 has been replaced several times. But I am informed by a testing engineer of much ex- 
 perience and reputation that he has a steel fence upon his property which has been in service 
 many years and has outlasted several other steel fences under similar conditions, but 
 formed of steel made recently. From this he argues that the steel made to-day is more 
 subject to corrosion than the steel made many years ago. If this is so, we are permitted 
 to ask: Is the wrought iron made to-day more subject to corrosion than the wrought iron 
 made several years ago? 
 
THE CORROSION OF IRON AND STEEL 
 
 417 
 
 difference in the speed of corrosion between wrought iron and steel 
 is very small, although favorable to wrought iron in the case of 
 sea water and alkaline water and to steel in the case of acids and 
 acidulated water. As has been pointed out, however, these 
 scientific tests are not altogether reliable as a basis for commer- 
 cial comparison, because they have not usually been carried to 
 the point where either one material or the other becomes unfit 
 for service, but merely allow corrosion to proceed for several 
 
 FIG. 326.— CORRODED WROUGHT-IRON PLATE. 
 
 months and then show the relative loss in weight. Neither have 
 these experiments taken into account sufficiently the localized, 
 i.e., "pitting," corrosion to which badly made material is espe- 
 cially subject. It is immaterial whether or not the metal has lost 
 but little weight, provided it has pitted in any one spot sufii- 
 ciently to fail, or to have become dangerously thin. This 
 
 27 
 
418 THE METALLURGY OF IRON AND STEEL 
 
 pitting is believed to be due chiefly to blowholes and possibly 
 to segregation resulting in a local increase in electric potential. 
 
 Manganese and Corrosion, — It has been suggested that the 
 presence of manganese in steel causes an increase in the rate of 
 corrosion, but this assertion is based upon no reliable evidence 
 so far as I am aware. It was brought forward to explain the 
 supposed rapid corrosion of steel as compared with wrought iron, 
 but if it is a true influence in this direction, then steel should 
 corrode in acids faster than wrought iron, which it apparently 
 does not if the steel has been made with due care and is free 
 from blowholes and much segregation. 
 
 Badly Made Material. — There can be no doubt that badly 
 made steel is much more liable, to corrosion and to pitting than 
 well-made steel, and it may be from this cause that the bad name 
 which steel is popularly given comes. There can also be no 
 doubt that badly made wrought iron is extraordinarily subject to 
 rusting, and of this kind of material we are to-day getting a good 
 deal. As noted on page 47, probably more than half of the 
 wrought iron produced in America is made by "busheling" scrap 
 into a pile, rolling it down and marketing, it as wrought iron. 
 This material is of good quality so long as the scrap from which it 
 is made is good, but when the scrap is collected form almost any 
 source, and especially when it contains steel, as it sometimes 
 does, we should expect great differences in potential and there- 
 fore rapid corrosion. 
 
 Coating. — Wrought iron has one advantage over steelin the 
 case of articles which are to be coated, because its rough surface 
 gives a better opportunity for the paint to adhere than the com- 
 paratively smooth and even surface of steel. 
 
 Summary, — Badly made steel and badly made wrought iron 
 corrode faster than any other material; next in order come well- 
 made steel and well-made wrought iron, between which two 
 classes the difference is probably very slight, and has not been 
 determined by any suflaciently lengthy or convincing series of 
 tests; the next in order is probably cast iron, although as yet 
 we cannot be quite certain that this corrodes more slowly than 
 wrought iron and steel, except in so far as it is protected by its 
 natural or artificial coating or both. Steel and wrought iron 
 are both liable to pitting, which may 'greatly shorten their life 
 in service even though the average rate at which they corrode is 
 slow. Several causes may produce this pitting, such as blow- 
 
THE CORROSION OF IRON AND STEEL 419 
 
 holes, segregation, bad welding in places, particles of oxide, 
 scale or dirt, etc. When pits or holes are found with a smooth, 
 hollow surface, it is altogether probable that they are due to 
 blowholes and sometimes these pits may be an inch or more in 
 diameter and extend an eighth of an inch into the plate, even 
 before the remainder of the surface has become severely attacked. 
 Numerous efforts are being made at the present time to improve 
 the quality of steel and to turn out a more uniform grade of 
 product. It is quite certain that painting is not as good to-day 
 as it was in years past and that the quality of paint used is also 
 worse, on the average, as shown by the fact that paint does not 
 last as well on wooden struci/ures. The better adherence of paint 
 gives an advantage to wrought iron, which, however, only applies 
 where work is so placed as to be capable of being painted. 
 
 Preservative Coatings for Iron and Steel 
 
 The oxidized surface which all steel retains after hot working 
 is in itself a certain protection against corrosion, but this is only 
 limited in effect, because the oxide is more or less porous and so 
 allows the corrosive agencies to penetrate it and attack the 
 surface underneath. Moreover, the scale does not adhere firmly 
 in some places, and as its coefficient of expansion and contraction 
 is different from that of the metal, it is liable to loosen and fall 
 off in places and so expose the iron or steel. 
 
 Preparation of Surfaces for Coating. — It is not advisable to 
 coat wrqught iron or steel until its surface has been carefully pre- 
 pared, for any rust or other product of corrosion, scale, grease, 
 dirt, .or moisture underneath the coat will either start corrosion 
 or else, by becoming loosened, will cause the paint or galvanizing 
 to fall off and so expose the metal underneath. The opposite is 
 the case with cast iron, because when this metal is poured liquid 
 into the mold a skin, is formed, consisting of a chemical union of 
 silica and oxide of iron, very firmly united to the metal and 
 capable of being relied upon to adhere beneath the paint or other 
 coating and serve as additional protection from corrosion. So 
 much mill scale as adheres to wrought iron and steel with great 
 force is permitted by some engineers to remain underneath the 
 protective coating, but others insist upon the removal even of 
 this upon the ground that it may become loosened later by ex- 
 pansion atid contraction, and then spall off. 
 
420 THE METALLURGY OF IRON AND STEEL 
 
 Priming Coat. — A difference of opinion exists as to the ad- 
 visability of having wrought-iron and steel surfaces prepared 
 at the mill and there given a priming coat, or else having the 
 priniing coat applied under the direction of the engineer of con- 
 struction, or else of omitting the priming coat altogether and not 
 painting the structure at all (except such parts of it as will be 
 inaccessible after erection) until the metal has been exposed 
 so long that the weather has loosened all the scale. This will 
 occupy perhaps six months to a year, depending upon the cor- 
 rosive conditions. During that time no great damage can be 
 done by corrosion although the structure will, of course, look 
 very shabby. After that period the scale is removed with sand 
 blast, wire brushes, pneumatic hammers, or chisels, and after the 
 surface is perfectly clean and dry, it is given a priming coat and 
 at least two other good coats of paint, each one being allowed 
 to dry thoroughly before the next is spread. For indoor work 
 one coat of paint upon the priming coat is often considered 
 sufficient. 
 
 Shop vs. Field Painting. — The advantage of painting the steel 
 at the shop is that it can be done inside of some building and there 
 is therefore less liability of hygroscopic moisture under the 
 coating. If the shop coat is put on with care and skill it unques- 
 tionably has certain advantages, but it is too true that the shop 
 painting and the preparation of surfaces at the shop is often 
 carelessly done, for the n:ianufacturer has not the interest in 
 preserving the metal from decay that the consumer feels. Fur- 
 thermore it undoubtedly saves expanse to allow the structure to 
 stand from six months to a year, provided that it is then very 
 thoroughly cleaned and well painted when absolutely dry. If 
 otherwise the whole work may have to be done over again at the 
 end of a short interval.^ 
 
 Pickling. — To remove scale it is customary in many cases to 
 pickle the steel or wrought iron, i.e., to immerse it in dilute sul- 
 phuric acid (say 10 per cent.) preferably heated to boiling so as to 
 act more quickly. After a few minutes the scale is removed, 
 when the metal is washed once in boiling water, once in cold 
 water, and finally in lime water to neutralize the last traces of 
 the acid. It should remain in the lime water until it is ready for 
 the .application of the coating, when it should first be washed 
 
 1 In which event it is customary for all interested parties to blame the paint, except 
 the paint-maker who is in the minority. 
 
THE CORROSION OF IRON AND STEEL 421 
 
 free of lime, and then heated slightly above 100° C. (212° F.) to 
 drive off all the moisture. Pickling is therefore applicable only 
 when the metal is to be coated at the shop, either with the prim- 
 ing coat of paint or with zinc, tin, etc. 
 
 Comparison of Methods. — ^Pickling costs less than the other 
 methods of removing scale and accomplishes the work very 
 thoroughly. Sand blast is the next cheapest method. This 
 latter does pot get off all the scale unless it is very thorough, while 
 if it is too thorough, it leaves the surface in a smooth condition so 
 that the paint does not stick so well. On the other hand, pickling 
 must be done with great care or it may leave hydrogen upon the 
 surface of the metal which will greatly hasten corrosion so that 
 pickled surfaces sometimes corrode more rapidly than those which 
 have been cleaned in any other way. Cleaning with wire brushes 
 is more expensive than sand blasting, but if performed with great 
 care is more effective and leaves the surface in a rougher condition 
 which assists~the adheren.ce of the paint. 
 
 Kinds of Paint Used. — There is a great difference of opinion as 
 to the best paint for preserving iron and steel, but some few 
 things seem certain: (1) That no one kind of paint is suitable 
 protection against all corrosive influences. For example, the 
 best paint to withstand the action of the open air may fail when 
 exposed to the elements in a damp tunnel, or when used on the 
 parts of pries under sea water, while a good protection against 
 this latter influence might be inefficient when exposed to the 
 oxidizing gases in locomotive smoke, etc.; (2) that whatever 
 paint is used must be sufficiently elastic to expand and contract 
 with the changes in temperature of the metal without cracking; 
 and (3) that it must contain nothing which will attack the metal 
 and so commence corrosion. In this last connction we must 
 especially avoid all oxidizing influences. The reader will find in 
 the Annual Proceedings of the American Society for Testing 
 Materials a very full interchange of opinions between experts in 
 paint manufacture and engineering which will doubtless help 
 him to form an opinion as to the best paint to use in each case. 
 
 There are two parts to every paint: (1) The vehicle which car- 
 ries the, pigment and undergoes a change to the solid state when 
 the paint dries, and (2) the pigment or originally solid part of the 
 preservative coating. These two must form ^ firm impervious 
 coating upon the surface of the metal, but must not be so solid 
 as to be inelastic or brittle. 
 
422 THE METALLURGY OF IRON AND STEEL 
 
 Linseed Oil. — Linseed oil is a very good and common vehicle.^ 
 It is what is known as a "drying oil." That is to say, an oil 
 which when exposed to the atmosphere will change from a liquid 
 to an elastic or leathery consistency. This action takes place not 
 by evaporation but by a process of oxidation, whereby the oil 
 absorbs oxygen to the extent of form 10 to 18 per cent, of its 
 weight and expands in volume, so that a coat of linseed oil spread 
 upon glass will wrinkle up upon drying. It is because linseed is 
 the best of all the drying oils that it is so much preferred as a paint 
 vehicle, but when allowed to dry in the raw state it requires too 
 long a time and the drying is therefore hastened by boiling it and 
 adding some oxidizing agent, known as a "drier," of which the 
 best, as far as iron and steel preservation is concerned, are the 
 salts of lead or manganese, used without rosin. It will readily 
 be understood how dangerous it is to indiscriminately use driers 
 (i.e., oxidizing agents) in steel paints, because so many of them 
 will oxidize the steel and so cause the very corrosion which it is 
 the object of the paint to prepent. 
 
 Purity of Linseed. — This brings us to the question of the purity 
 of linseed oil, for the usual adulterants are all harmful to steel 
 work and cause in the end more painting and more of the expens- 
 ive cleaning of structures to receive the coating, out of all 
 proportion to their lesser first cost. Freedom from adulterants 
 probably 'cannot be obtained expect by constant watchfulness 
 and frequent chemical analysis on the part of the consumer. 
 Some impurity arises from the presence of a few per cent, of 
 foreign seeds with the linseed, which is not always avoidable, but 
 the greater harm comes from the fact that the oil is obtained 
 from the linseed by pressing it while hot, in which way a larger 
 amount of product is extracted than if cold pressure is applied, 
 because some of the solid part of the seeds are thereby extracted 
 together with the oil. Cold pressed linseed oil has a golden 
 yellow color and remains clear in cold weather as distinguished 
 from the yellowish-brown color of the hot pressed oil which also 
 has a more acrid taste, is not so fluid, and contains more solid 
 fats, solid organic matter, and fatty acids, all of which are harm- 
 ful either because they attack the metal or else because they 
 make a pervious paint. 
 
 Pigments. — The pigment is not as important as the vehicle 
 and many different ones can be chosen, provided they are chem- 
 
 1 See page 336 of No. 163; page 41 of No. 164. 
 
THE CORROSION OF IRON AND STEEL 423 
 
 ically inert to the steel. Red lead has been very much used and is 
 very good especially for the priming coat, for it seems to form 
 with the linseed oil a very dense, impervious coating. For the 
 outer coats, however, it is generally well to mix the red lead with 
 some substance that shall reduce its weight, such as graphite. 
 Ferric oxide, FcgOg, and other oxides of iron in the form of iron 
 ore are very cheap and withstand the action of sulphur gas better 
 than the red-lead paints. They are very good for outer coats 
 where locomotive smoke and similar gases are liable to be present. 
 Sulphate of lead, white lead (a mixture of oxide and sulphate of 
 lead with often some sulphate of zinc) and sulphate of zinc are all 
 good white paints although expensive. Pulverized asphaltum 
 and other hydrocarbons are also used with success as pigments> 
 especially where the metal is exposed in damp ground or under 
 water. 
 
 Other Paints. — Pipe is often coated very cheaply by dipping 
 it in melted asphaltum or pitch. The objection to this coat is 
 that it is very hard and brittle when cold and in time it forms a 
 network of myriads of cracks through which the atmosphere 
 attacks the metal. For cast-iron pipe it is very useful, however, 
 because this is protected by its natural skin. Dipping in tar 
 would form an elastic coating, but unfortunately tar contains 
 certain acids and oxidizing agents which attack the metal. There 
 is a paint made by distilling off the creosote and other volatile 
 components of tar until the solid asphaltum is left. This is then 
 redissolved in two of the distillates, neither of which will attack 
 iron work, and thus a paint is obtained which is said to be prac- 
 tically tar without any of its harmful constituents. It forms a 
 very elastic coating which does not crack after years of exposure 
 nor does it disintegrate under the action of the sun as the linseed 
 oil paints sometimes do. 
 
 Galvanizing. — Galvanizing is the process of coating with 
 metallic zinc and where this coating adheres firmly, it undoubt- 
 edly forms a very efficient means of protecting iron from corro- 
 sion. As zinc is electrically positive towards iron, whatever 
 electrolysis exists would tend to corrode the zinc and protect the 
 iron. Indeed this fact is taken advantage of by some engineers 
 who hang pieces of zinc in their boilers, by means of a wire con- 
 nected to the steel work, so that the electrolytic action shall 
 corrode the zinc and protect the wrought iron or steel. 
 
 Galvanizing is usually applied to wire and wire products, thin 
 
424 THE METALLURGY OF IRON AND STEEL 
 
 sheets, especially corrugated sheets used for the outside of build- 
 ings, etc., tubes, hollow ware, and a great variety of articles, 
 after the surfaces have been cleaned by pickling. There are three 
 methods by which the galvanizing is effected, known respectively 
 as cold galvanizing, hot galvanizing, and dry galvanizing. 
 
 Cold Galvanizing, — In the cold galvanizing process zinc is 
 deposited electrolytically upon the surface of metallic articles 
 which are made the cathode of an electro-plating cell. The zinc 
 is first dissolved in sulphuric acid and water and this solution is 
 made the electrolyte. The anode is a piece of zinc, so that as 
 fast as the electricity deposits zinc upon the surface of the article 
 being galvanized, it replenishes the electrolyte by dissolving zinc 
 from the anode. The coating is about 0.0003 to 0.0005 in. 
 thick, equivalent to about 0.2 to 0.3 oz. of zinc per square 
 foot of surface. 
 
 Hot Galvanizing, — In the hot galvanizing process, which is the 
 commonest one used, the articles to be galvanized are dipped into 
 a bath of molten zinc at a temperature of 425° to 460° C. (800° to 
 860° F.), i.e., slightly above the melting-point (419° C. =Y86° F.). 
 The metal is exposed to the zinc bath usually about 1 1/2 to 7 1/2 
 minutes, depending upon the thickness of coating desired, which 
 will vary between 0.0003 and 0.0010 in. or about 0.2 to 0.6 oz. 
 of zinc per square foot of surface covered, or about 0.3 to 0.6 
 oz. per pound of wire. In. the caSe of wire the iron or steel 
 is drawn slowly through the bath of melted zinc and usually 
 passes over a wiper as it. comes out, which removes the still moltin 
 zinc and causes the zinc remaining to stick a little more firmly 
 and have a more uniform thickness. The coating on this wiped 
 wire is not so liable to crack and break off when the wire is bent 
 and twisted as the coat of unwiped wire but, on the other hand, 
 it is thinner and gives but little protection against corrosion. 
 Sometimes articles 'to be galvanized are first dipped in a bath of 
 melted lead and then in the melted zinc. This gives a cheaper 
 coating. 
 
 Dry Galvanizing. — The process of dry galvanizing is a recent 
 invention and consists in heating the articles to be galvanized 
 inside a closed vessel and while they are covered with what is 
 known as "blue powder," which is a zinc dust containing some 
 oxide of zinc and relatively cheap in price because it is a by- 
 product in the metallurgy of zinc. The temperature is about 
 300° C. (575° F.), and, although this is below the melting-point 
 
THE CORROSION OF IRON AND STEEL 425 
 
 of zinc and of iron, it is sufficiently high to produce an alloy 
 between the two, forming, it is said, a very resisting coating 
 which is more thoroughly attached to the surface of the metal 
 and therefore much more durable against cracking off. 
 
 Comparison of Galvanizing Methods, — Cold galvanizing deposits 
 a thinner coating of zinc which, if improperly performed, is 
 liable to be porous or spongy, but it gives a better connection 
 between iron and zinc and therefore a more durable coating. 
 Hot galvanizing necessitates the use of a flux on the bath of 
 melted zinc in order that the zinc may not be oxidized by air, and 
 these fluxes probably have the effect of sometimes beginning the 
 corrosion of the iron underneath the layer of zinc. The process 
 of dry galvanizing is too new yet for any comparison to be 
 drawn. 
 
 Tinning, — A large amount of metal is coated with tin in order 
 to give protection against organic acids, such as those present in 
 cooked foods, and also in order to give a more effective resistance 
 to the elements. Thus cooking utensils, roofing sheets, tin cans 
 for preserves, and many such articles are coated with tin in pref- 
 erence to zinc, either because zinc would not withstand so long 
 the corrosive influence, or else would not resist it at all. In the 
 tinning operations the metal sheets are usually drawn through a 
 bath of liquid tin by four to six pairs of rolls which are immersed 
 in it. Each pair of rolls presses the tin which has solidified on 
 the surface of the iron firmly upon the metal and the result 
 is a smooth, bright, adhering coat which protects the metal 
 very successfully. Tin plating is more expensive than galvan- 
 izing, chiefly on accoui^t of the additional cost of the tin. 
 
 Terne Plate, — Sometimes sheet metal is coated with a mixture 
 of two-thirds lead and one-third tin and then goes under the name 
 of terne plate, which is used very largely for roofing and out door , 
 purposes. It is applied by the same method as tinning but is 
 less expensive. 
 
 Nickel Plating. — Articles requiring a very high polish and 
 which are to be subjected to handling, etc., are often plated with 
 nickel. This is an electrolytic process, similar to the general 
 operation described under the electrolytic galvanizing. Nickel 
 plating is more expensive than galvanizing or tinning, but gives 
 a more highly resisting surface. 
 
 Oxidized Coating, — There are one or two processes by which a 
 black oxidized surface can be given to iron and steel which will re- 
 
426 THE METALLURGY OF IRON AND STEEL 
 
 sist rust for years and form what are known as "black iron"- 
 objects. It is used chiefly for fancy iron work in house decora- 
 tions, etc. 
 
 Enameling. — A number of articles, such as bath tubs, wash- 
 bowls, cooking utensils, are made of cast iron or steel and then 
 coated with a white or variously colored film known as enamel. 
 Enameling processes are more or less secret, but usually consist in 
 powdering the enamel upon the surface of the metallic article 
 which has been heated to a red heat. At this temperature the 
 mixtures forming the enamel melt and spread themselves uni- 
 formly over the surface where they chill and harden. Enamel 
 must be insoluble in water and in chemicals with which they are 
 liable to come in contact, and must also be sufficiently elastic to 
 expand and contract with the metal without breaking off. 
 
XVII 
 THE ELECTRO-METALLURGY OF IRON AND STEEL 
 
 In the electric smelting and refining of iron and steel, six 
 modifications in practice are produced: 
 
 L Practically any desired temperature in reason may be 
 obtained. A corrolary t'o this is, that a "super-refining'' of steel 
 is possible, whereby we may reduce sulphur to a lower content 
 than may be ordinarily achieved in the older processes; likewise, 
 dissolved gases, occluded slags, etc., may be eliminated from the 
 liquid metal by means of heat, instead of by means of manganese, 
 silicon, titanium, etc. In brief, the strength of Bessemer or 
 basic open-hearth steel may be increased some 10,000 lbs. per 
 sq. in. by a "super-refining'' in the electric furnace. 
 
 2. The impurities introduced with the usual fuels are avoided. 
 
 3. The temperature is regulated with much greater accuracy. 
 
 4. The cost of the process is greater than Bessemer or open 
 hearth.^ 
 
 5. Very little, if any, fuel is required, and that only for re- 
 duction, since heat is produced by electrical means (with certain 
 qualifications hereafter mentioned) . 
 
 6. Oxygen is not necessarily introduced into the furnace in 
 which the operation is carried on. 
 
 For many years the first three modifications have been taken ad- 
 vantage of in the production of ferro-alloys — e.g., ferro-tungsten, 
 ferro-chrome, ferro-molybdenum, etc.^that is, alloys of pig iron 
 and some other metal which is used for recarburizing in the man- 
 ufacture of alloy steels. The high temperature necessary for 
 the production of these alloys gives electric smelting especial 
 advantages while the high price at which they can be sold enables 
 their manufacturers to stand the additional expense with profit. 
 But in the year 1900 a number of important electro-metallurgists 
 in Europe and America began to use electric processes also for 
 the production of pig iron and steel on a commercial scale, and 
 from this time the industry dates. Because the electric processes 
 have apparently/secured for themselves a permanent place in the 
 metallurgy of iron and steel; because of the great interest which 
 
 * 8600 K. W. hours =0= to heat in about 1 ton of coal. 1 H. P. = 744 Watts = 
 approximately 3/4 K. W. 
 
 427 
 
428 THE METALLURGY OF IRON AND STEEL 
 
 they have evoked on account of the new- and valuable metal- 
 lurgical principles involved; because the production now amounts 
 to figures well worthy of consideration, and finally, because the 
 electric steel seems to rival, or even surpass, crucible steel in 
 quality, and to ultimately replace it in large part — we could well 
 devote more space to them than the limits of this volume permit. 
 Entire books could, and have been, profitably devoted to the 
 subject, to which readers must be referred for a more extended 
 discussion. 
 
 Iron and steel electro-metallurgical processes naturally divide 
 themselves into four classes: 
 
 (1) Ore smelting for the production of pig iron; 
 
 (2) Refining of pig iron to produce steel; 
 
 (3) Super-refining of steels made by other processes, and 
 
 (4) Electrolytic refining of steel or wrought iron to produce 
 almost chemically pure iron. 
 
 The first three classes are electro-thermic; that is to say, they 
 use electricity for conversion into heat; the fourth class is elec- 
 trolytic, that is to say, the electric current serves to produce 
 chemical changes. 
 
 Electro-thermic Iron Ore Smelting 
 
 Electric ore smelting to produce ordinary pig iron (not in- 
 cluding ferro-alloys) has been successful where coke and pig iron 
 are both costly, but ore is readily obtainable. The explanation 
 for this is that the consumption of coke in the ordinary iron blast 
 furnace is seldom less than 2000 lbs. per ton of pig iron produced, 
 and may be much more than that; whereas electric smelting may 
 reduce this consumption to well under a third of that figure. 
 It may also replace coke by charcoal for reduction, in localities 
 where the latter fuel is cheaper. J. W. Richards^ has shown 
 that electric furnace smelting cannot use more than one-third as 
 much carbon as ordinary blast furnace smelting, nor less than 
 two-ninths. The ore-smelting furnace about which the largest 
 amount of data is available, is the so-called "Domnarfvet" 
 furnace in Sweden, and we may use this as an example of this 
 process: 
 
 Domnarfvet Furnace, — The construction of this furnace is 
 shown in Fig. 327, and its operation briefly indicated: Alter- 
 
 ^ Transactions of the American Electro-Chemical Society, 1909, xv, page 56. 
 
THE ELECTRO-METALLURGY OF IRON AND STEEL 429 
 
 nating current of 3 phase and 25 cycles, and capable of adjust- 
 ment by small intervals from 20 to 80 volts, is introduced 
 through three carbon electrodes between which the de- 
 scending charge is made to pass. The total power avail- 
 
 ■Three tuyeres like this 
 blow cool tunnel-head 
 eases into the free space 
 above the ore, for the 
 purpose of cooling the 
 roof. They also give a 
 better distribution of 
 heat throughout the 
 charge in the furnace 
 shaft. 
 
 FIG. 327.— SECTIONAL ELEVATION OF ELECTRIC SHAFT FURNACE. 
 AT DOMNARFVET, SWEDEN. 
 
 able is 800 hp. The fuel in the charge conducts the current, 
 and the resistance of the charge transforms the electric into 
 heat energy. By means of -the heat so produced, the car- 
 bon reduces the iron from the ore, and is itself oxidized. The 
 
430 
 
 THE METALLURGY OF IRON AND STEEL 
 
 resulting gases pass upward, giving their heat to the charge in the 
 furnace shaft and also performing some preliminary reduction. 
 The tunnel-head gases are returned to th& crucible for cooling 
 purposes, and, by varying the speed of their circulation, the re- 
 lative amount of COg to CO may be made approximately equal, 
 whereby the temperature of gases at the top will be kept at 
 about 200 to 300° C. (392 to 572° F.) . There is much ore reduced 
 in the upper part of the furnace, just as in the blast furnace, but 
 the efficiency of reduction is greater in this electric furnace, as is 
 indicated by the larger relative proportion of COg to CO. 
 
 FIG. 328.— PLAN VIEW OF DOMNARFVET FURNACE WITH SHAFT AND 
 ELECTRODES REMOVED. 
 
 Designed by Gronwall, Lindblad and Stalhane. (Sometimes also known as the furnace of 
 Aktiebolaget Electrometall.) 
 
 The crucible does not support the shaft, but the latter rests on 
 six columns, as shown. Shaft and crucible are filled with ore, 
 except where the angle of rest of the charge leaves a space under 
 the roof of the crucible, as indicated by the dotted lines. This free 
 space is essential, and avoids the contact of electrodes with the 
 charge adjacent to the furnace brickwork. By this means, de- 
 
THE ELECTRO-METALLURGY OF IRON AND STEEL 431 
 
 struction of the lining, which is rapid when the electrode contact 
 with the ore is near the wall, is greatly reduced. 
 
 The melting zone is located between the three electrodes, and 
 the control of temperature (by means of varying the electric 
 tension) is excellent, as shown by the ability to regulate the silicon 
 in the pig iron at will, and the ready removal of sulphur by means 
 of a high heat and basic slag. From ores, containing 0.5 per cent. 
 of sulphur, iron was made with only 0.005 per cent, of that ele- 
 ment. An important factor in this removal is probably the cal- 
 cium carbide formed by the electric current. Phosphorus is not 
 under control, but substantially all that present in ores and fuel, 
 goes into the metal. With pure ores and charcoal for reduction, 
 low-phosphorous pig ir6n can be made, however. 
 
 The voltage and amperes under certain conditions are shown 
 in the following table: 
 
 Volts Current 
 
 ' across per electrode 
 
 Charge electrodes in amperes 
 
 Coke in excess 34 9,600 
 
 Cokfe not in excess 36 8,800 
 
 Charcoal in deficiency 60 6,300 
 
 Charcoal right 54 7,600 
 
 Charcoal in excess 48 7,600 
 
 Coke and charcoal in excess 35 9,200 
 
 Coke and charcoal right 48 7,600 
 
 The electric efficiency runs up to 58 per cent., and the average 
 consumption of electric energy in seven trial runs was 0.492 hp. 
 year per metric ton of pig iron produced; the best figure obtained 
 was 0.383 hp. year. If metal is allowed to accumulate in the 
 hearth, the voltage drops; the usual practice is to cast every 
 6 hours. The capacity of the furnace is described as about 6 
 tons per day. With 100 kilos of ore the furnace will use about 
 4 kilos of calcined lime, together with 21 to 28 kilos of charcoal; 
 or else 22 to 24 kilos of coke in place of the charcoal. The 
 total carbon consumption per metric ton of pig iron produced, 
 under favorable conditions, will be in the neighborhood of 560 
 lbs., and may be calculated from the following formulae: 
 
 321.43 (100-fe) + 10 fc 
 
 When smeltmff hematite: 
 
 ^ 100 + m 
 
 . 285.7 (100-k) +10k 
 
 When smeltmg magnetite: 
 
 100 4- m 
 
432 THE METALLURGY OF IRON AND STEEL 
 
 where k is the percentage of carbon in the iron, and m is the 
 volumetric percentage of COj in the tunnel-head gas. The con- 
 sumption of electrodes — always a costly item in electric furnaces 
 using carbon for this purpose — varies from 24.5 to 42.7 lbs. per 
 metric ton of iron produced. If means could be devised of using 
 the "stub" of the electrodes, the consumption might be reduced 
 by 40 per cent. On the basis of the practice described above, 
 the cost for smelting a metric ton (2204 lbs.) of pig iron will be 
 as follows, omitting the cost for ore, limestone and patent 
 royalties, which can be figured for each individual case: 
 
 Charcoal at $8 per ton $ 1 .92 
 
 0.5 E. H. P. year at $ 12 6.00 
 
 30 lbs. electrodes at 3 cents per lb 90 
 
 Miscellaneous $1 . 50 
 
 $10.32 
 
 In order that blast furnace smelting might cost as much as this, 
 we should have to assume a price for coke of over $7.00 per ton. 
 Such a price actually obtains, or is exceeded, in many localities. 
 
 Noble, or HerouU, Ore Smelting Furnace. — At H^roult, Cali- 
 fornia, where coke is costly, electric ore smelting is carried on in 
 a furnace originally designed by P. H^roult, but now modified. 
 The general form and arrangement at present is similar to the 
 Domnarfvet furnace, and a similar circulation of gases ife em- 
 ployed to cool the crucible roof, although they are here tapped 
 from the middle of the shaft. The power consumption is said 
 to be^ only 0.2 hp. year per ton of pig iron and the carbon con- 
 sumption close to the theoretical amount. Smelting is appar- 
 ently a commercial success at this plant. 
 
 Refining of Pig Iron and Melting in the Electric Furnaces. — 
 There are no technical or practical difficulties in the way of re- 
 moving carbon, silicon, manganese, phosphorus, sulphur, gases, 
 slag, oxides and other foreign substances in the electric furnace, 
 and one can find substantial evidence in the literature and at 
 steelworks, (especiallyin'Germany and America) of such practice 
 in furnaces .of the Roechling-Rodenhauser, the H^roult, the 
 Girod, the Stassano, the Giffre, and other types. The principles 
 by which the purification is accomplished are the same as those 
 applied in the processes with which we are more familiar; for ex- 
 ample, the blast furnace (as to sulphur only), the Bessemer, open- 
 
 ^The Eng. and Min. Jour., Aug. 6, 1910, page 271. 
 
THE ELECTRO-METALLURGY OF IRON AND STEEL 433 
 
 hearth and puddling furnaces, the crucible process (as to slag, 
 oxides), etc. On the other hand, commercial considerations 
 strictly limit refining in electric furnaces, because carbon, silicon 
 and manganese can be much more cheaply eliminated in the 
 puddling, Bessemer and open-hearth furnaces. Even the re- 
 moval of phosphorus is cheaper in the puddling furnace and 
 basic Bessemer and open-hearth processes, and just as effective 
 as the present electric processes. Depending upon the propor- 
 tion of phosphorus present, it requires an expenditure of about 
 40 to' 200 kw. hours to dephosphorize one ton of metal previously 
 melted, besides heavy costs for electrodes, repairs, etc. These 
 expenses give the older processes the advantage in cost under all 
 usual conditions. When it comes to both melting and refining, 
 the situation is still more against electric energy, because at 
 least 400 to 600 kw. hours are required to melt a ton of solid 
 iron or steel, besides the long operation during which other ex- 
 penses accrue. 
 
 Only when melting steel containing costly ingredients which 
 will be oxidized and wasted in the open-hearth furnace — as, for 
 instance, high-priced alloy steels, etc. — do commercial considera- 
 tions favor the electric melting process with its possibility of 
 melting in a neutral or reducing atmosphere. 
 
 Super-refining in Electric Furnaces 
 
 These considerations relegate the chief field of electric refining 
 processes to the performance of operations which are beyond 
 the scope of the Bessemer and open-hearth processes, and the 
 most important of these fields are desulphurizing and deoxidizing 
 the bath, and removing from it entangled particles of slag, oxide, 
 etc. The electric processes can accomplish these objects by 
 virtue of their ability to heat a liquid bath of metal in a non- 
 oxidizing or actually reducing atmosphere. In this respect the 
 electric processes resemble the crucible process, but they have 
 two advantages over crucible heating, of which the first is an 
 ability to desulphurize by means of attaining high temperatures, 
 and the second is greater economy of operation. The amount 
 of capital required to install electric furnaces, the large product 
 necessary to obtain economical operation, and the established 
 position of crucible steel in the trade, as well as the capital in- 
 vested in crucible plants, will probably keep this process alive 
 
 28 
 
434 THE METALLURGY OF IRON AND STEEL 
 
 for a long time, just as the puddling process has persisted for 
 decades since the development of the Bessemer and the open- 
 hearth process. 
 
 The usual super-refining in the electric furnace, of metaJ 
 previously made in the open-hearth or Bessemer furnace, consists 
 of an oxidizing period, during which the phosphorus is reduced 
 to any desired point, followed by a reducing period during which 
 the sulphur and occluded substances are removed. 
 
 Where acid Bessemer steel is super-refined, the necessity for 
 the oxidizing period for the removal of phosphorus is obvious, 
 but where electric super-refining follows the basic open-hearth 
 operation, it would seem as if the oxidizing period in the electric 
 furnace might more economically be replaced by a continuance 
 of the dephosphorization to the desired point on the basic hearth. 
 
 Practice in Super-refining. — The metal before leaving the 
 Bessemer or open-hearth furnace is reduced as low as practicable 
 in carbon, silicon, and manganese, but is not recarburized. It 
 is often poured directly into the electric furnace in liquid form, 
 although sometimes the use of an intermediary mixer is ad- 
 vantageous. To the liquid bath in the electric furnace is now 
 added materials for making a basic, oxidizing slag, rich in iron, 
 resembling more the puddling furnace slags than those ^of the 
 open-hearth. The advantage of such a slag is that iron oxide 
 is not only retentive of phosphorus, but it will also oxidize phos- 
 phorus. We have already pointed out that phosphorus must be 
 oxidized in order to be eliminated from the metal and it must also 
 be absorbed by a slag that is retentive of it. A slag that is basic 
 by virtue of lime will retain phosphorus but will not oxidize it. 
 Moreover, iron oxide is a cheaper base than lime, and has the 
 further advantage that the reactions which it produces result in 
 the formation of metallic iron which increase the yield of the 
 process. It is obvious that 'the metal treated in the electric 
 furnace must begin by being low in carbon, else the oxidizing 
 agents will attack the carbon at this high temperature in pref- 
 erence to the phosphorus. Moreover, if there is much phos- 
 phorus to remove, we must use more than one slag, skimming off 
 the first slags as they become so high in phosphorus that they no 
 longer absorb it readily from the metal. Likewise, if the phos- 
 phorus is to be reduced from a moderate point to a very low 
 point, we must skim off the first slags formed, because there is a 
 certain ratio of distribution of phosphorus between metal and 
 
THE ELECTRO-METALLURGY OF IRON AND STEEL 435 
 
 slag which prevents a highly phosphorized slag from reducing 
 the phosphorus in the metal to a low point. 
 
 After the phosphorus in the metal has been reduced to the 
 desired proportions, the oxidizing slag is removed as completely 
 as possible and then a new slag is formed containing the smallest 
 possible amount of oxides of iron and manganese, and running 
 as high as 70 to 90 per cent. lime. At this point we add carbon 
 to commence the deoxidation of the metal. This is the cheapest 
 agent we can use for the purpose, and as it acts chiefly upon the 
 slag in which it floats, its chief function is to deoxidize this slag 
 and therefore make it more attractive for the oxides in the metal, 
 for there is also a ratio of distribution for oxides between metal 
 and slag similar to the ratio of distribution of phosphorus 
 already mentioned. The carbon deoxidation may be supple- 
 mented by the use of silicon, manganese and aluminum, if desired. 
 
 Only when both slag and metal are relatively free from re- 
 ducible oxides (i.e., chiefly oxides of iron and manganese) can 
 the reactions begin whereby the bulk of the sulphur is removed: 
 
 1. C + (Fe.Mn)S + CaO = CaS+Fe.Mn-}-CO; 
 
 2. Si + 2(Fe.Mn)S+2CaO = 2CaS + 2Fe.Mn + Si03. 
 
 This point is recognized by the two circumstances that the sUtg 
 becomes snow-white and that it falls to powder on cooling, both 
 of which are proof of its very low content in oxides of iron and 
 manganese. The presence of calcium .carbide is a further in- 
 dication. 
 
 It is this desulphurizing by means of calcium sulphide which 
 is the great advantage of the electric furnace in respect to sulphur 
 removal. In the basic open-hearth furnace also we may reduce 
 sulphur by means of a very basic slag, and by virtue of the ratio 
 of distribution of sulphur between metal and slag. This ratio 
 of distribution depends upon the basicity of the slag, its fluidity, 
 and its volume in reference to the volume of metal, as well as its 
 paucity in iron oxide. These conditions may be maintained 
 to a certain degree in the open-hearth furnace, except only that 
 the heat there is not sufficient to maintain an extremely basic 
 slag in a fluid condition, while the oxidizing conditions there 
 also oppose a low content of iron oxide in the slag. Moreover, 
 the sulphur in the producer gas of the open-hearth process is an 
 important factor, counteracting the removal of the last traces. 
 On these accounts it is ordinarily impracticable to reduce the 
 
436 THE METALLURGY OF IRON AND STEEL 
 
 sulphur below 0.02 to 0.03 per cent. The presence of manganese, 
 as already noted, assists in the removal of sulphur, because it 
 would seem that manganese sulphide increases the coefficient of 
 distribution of sulphur in favor of the slag. Even in the electric 
 furnace the presence of manganese assists in the removal of 
 sulphur, because manganese sulphide migrates more readily to 
 the slag than does iron sulphide, and therefore tends to bring the 
 sulphur in contact with the lime which makes possible reactions 
 (1) and (2) given above. 
 
 As regards the ability of the electric furnace to free the metal 
 from occluded oxides/ we need hardly do more than refer to the 
 circumstance already mentioned of the slowness with which 
 these microscopic particles rise to the surface of the liquid metal, 
 commensurate, in some degree with the rising of cream on milk. 
 The quiescent period of deoxidation and desulphurization in the 
 electric furnace produces the conditions which permit of this 
 separation, and the absence of any subsequent deoxidation by 
 means of manganese, silicon, or aluminum removes the liability 
 to the formation of those oxides which might become entangled. 
 
 Electric Refining Furnaces 
 
 Induction Furnace. — The simple induction furnace is based 
 upon the principle of the ordinary static transformer, whereby 
 alternating electric current is transformed to lower voltage. It 
 was independently developed by E. A. Colby, of the United 
 States, and F. A. Kjellin, of Sweden, whose American rights 
 have been joined under one management. A sectional elevation 
 of the furnace is shown in Fig. 329, in which CCCC is the core of 
 an electro-magnet, around one leg of which a coil of wire, AA, 
 passes. When an alternating current goes through the coil, AA, 
 it sets up an alternating magnetic field in the core, CCCC, and 
 this in turn sets up a secondary current in the circle, BB^ parallel 
 to the coil, AA. In other words, an alternating current passing 
 through the coil, AAj induces an alternating current in the coil, 
 BBj without there being any metallic connection between the two. 
 This is a well-recognized phenomenon in electrical engineering and 
 requires no further comment here. In the furnace operation, the 
 circle, BBj is a hollow ring in the brickwork into which melted 
 metal is poured. The resistance offered by this melted metal to 
 the passage of the induced current generates heat which will 
 
r 
 
 THE ELECTRO-METALLURGY OF IRON AND STEEL 437 
 
 ^^^M'.^}c^M.4^x^^:.s.^xv^.-^.^;<$:x^..^.^^^ 
 
 '**■■■' ^ ''■' J^ 
 
 1 1 t I ■ 1 
 
 , , , 2|MeterB 
 
 FIG. 329.-SECTION OF KJELLIN INDUCTION FURNACE. 
 
 FIG. 330.-ROECHLING-RODENHAUSER FURNACE. 
 
438 
 
 THE METALLURGY OF IRON AND STEEL 
 
 maintain the temperature or raise it to any desired point. The 
 slot, BBj is really an annular crucible into which pig iron, steel 
 scrap, iron ore and flux may be charged as if it were an open- 
 hearth furnace, and the operation of steel making is practically 
 the same in principle except that electric heat is employed instead 
 
 ■ = t^i^l^l^i?^^ = 
 
 ^^3& 
 
 XiOngitudinal Sections-A B &-C D 
 FIG. 331.-HEROULT REFINING- FURNACE. 
 
 of regenerated gas and air. We may charge solid metal if desired, 
 but, in such a case, it is well to leave a shallow circle o/ metal in 
 the bottom of the slot, BBj after each operation is ended, to serve 
 to carry the induced current during the beginning of the next 
 operation, until the solid charge is melted. 
 
 Roechling-Rodenhauser Furnace, — The combination induction 
 furnace is shown in Fig. 330. This is in effect a combination of 
 
THE ELECTRO-METALLURGY OF IRON AND STEEL 439 
 
 three simple induction furnaces, the coils of the metallic baths 
 joining in a central pool. By means of this pool, a more effective 
 refining may be employed, for the pool serves the purpose of a 
 bath upon which slags may be charged, etc. It has also several 
 steel terminal plates imbedded in the lining, which are connected 
 to a few heavy turns of copper placed outside the primary coils, 
 
 Transyers Section E F 
 FIG. 332.-HEROULT REFINING FURNACE. 
 
 which collect and feed to the terminal plates the induced current 
 in these turns, thus suppressing magnetic leakage. 
 
 HSroult Furnace, — The design and operation of the H6roult 
 steel furnace is even more like an open-hearth than the induction 
 furnace, although here again electric heat is substituted for com- 
 bustion. The general form of the furnace is shown in Figs. 331 
 to 332. There are two electrodes of carbon, one of which leads 
 the current to the charge and the other conducts it away. The 
 
440 
 
 THE METALLURGY OF IRON AND STEEL 
 
 electrodes do not touch the charge, but the current arcs from one 
 electrode to the charge, through which it passes, and thence arcs 
 to the negative electrode. 
 
 Girod Furnace.— The design of the Girod furnace is very similar 
 to that of H6roult, but electrodes in the bottom carry away the 
 current, instead of double electrodes at the top. See Fig. 333. 
 
 FIG. 333.— GIROD FURNACE. 
 
 Electrolytic Refining of Iron 
 
 The greatest amount of work on the electrolytic refining of 
 iron has been done by C. F. Burgess, of the University of Wis- 
 consin, who produces an iron that is almost chemically pure, 
 the chief foreign element being hydrogen, whose presence renders 
 the metal very hard and brittle. The hydrogen is driven off by 
 heating the metal to a'high temperature, but this is accomplished 
 with the result of vitiating the iron with traces of carbon and 
 sometimes other impurities, for iron has such a great affinity for 
 
THE ELECTRO-METALLpRGY OF IRON AND STEEL 441 
 
 carbon that it will absorb it at a red l\eat from coke, charcoal, 
 and even from gases and oil vapors. Indeed, if a piece of electro- 
 lytic iron, or other iron very low in carbon, be heated in contact 
 with steel or wrought iron higher in carbon, there will be a small 
 amount of transfer of carbon from the low- to the high-carbon 
 metal. The same difficulty is met with in melting the metal, 
 and as its melting-point is 1507° C. (2745° F.), there are not many 
 kinds of crucibles that will stand the heat necessary and yet fail 
 to yield carbon, silicon, or some other impurity to the iron. As 
 I understand it, it is this difficulty which has been the chief 
 obstacle in the electrolytic process, for the electrolysis itself 
 seems to be accomplished by Burgess with success and economy. 
 If the cost of , the electrolytic product could be brought some- 
 where near that of Swedish iron and the other very pure forms, 
 it is probable that it would become a commercial commodity on 
 account of its high magnetic permeability, electric conductility 
 and softness. 
 
 Electrolytic Process. — In Burgess's process the electrolyte is 
 a mixture of ferrous sulphate and ammonium sulphate and it 
 is kept at a temperature of 30° C. (86° F.) . The anode is ordinary 
 wrought iron or steel and the primary cathode upon which the 
 first metal is deposited is a thin strip of sheet iron. The depos- 
 ited metal sticks so slightly to this that no difficulty is met with 
 in separating them after the operation is finished. The electric 
 current, with a density of from 6 to 10 amperes per square of 
 cathode surface, deposits the dissolved iron upon the cathode 
 with an efficiency of nearly 100 per cent., and cathode plates 
 averaging about three-quarters of an inch thickness are produced 
 in a four weeks' run. 
 
XVIII 
 THE METALLOGRAPHY OF IRON AND STEEL 
 
 Metallogkaphy in its larger sense is the description or study 
 of the structure of metals. That branch of the subject which 
 comes under the head of microscopic metallography is, however, 
 the most important because the structure of most metals, espe- 
 cially iron and steel, is discernible only when magnified. We 
 shall see, however, that the observation of, structure by eye-^- 
 known as macroscopic metallography — is not . without great 
 value. 
 
 Microscopic metallography has now reached that stage, of im- 
 portance where it is viewed almost on a par with chemical analy- 
 sis and physical testing. In the United States practically every 
 large steel works is well equipped for the microscopic analysis of 
 its product, and important laboratories of the universities ai;id 
 of consulting metallurgists devote much attention to the study. 
 Although only a little more than twenty years have elapsed since 
 the art first received attention, it has, advanced so far as. to 
 have become by now another and a very serviceable tool in the 
 hands of the expert. I take this opportunity, however, of offer- 
 ing a word of warning: reputations have more than once suffered' 
 severely, because of erroneous deduction made from microscopic 
 evidence, and history has shown that those who "rush in where 
 angles fear to tread" are sure to be caught sooner or later. The 
 wise man is he who never bases an opinion upon a sample whose 
 chemical analysis is unknown to him: who never bases an opinion 
 upon a microphotographic negative or print, and who polishes 
 and etches his own specimens, or has these operations performed 
 by some one well known to him and working under his immediate 
 direction. "With these precautions the microscope is a very 
 reliable index to an experienced mind. 
 
 Preparation of Samples for Microscopic Examination 
 
 Samples of iron and steel for microscopic analysis can be cut 
 out of soft samples by means of a hacksaw, lathe, or other ma- 
 chine, and broken out of hard and brittle samples by means of a 
 hammer. A good rule to follow is to have the surface that is to 
 
 442 
 
THE METALLOGRAPHY OF IRON AND STEEL 443 
 
 be polished about 3/8 to 1/2 in. on a side. If larger than 
 that, it requires excessive labor for polishing. It requires about 
 sixteen times as much labor to polish a sample an inch square 
 as to polish one 1/2 in. square, and it requires about sixty-four 
 times the labor to polish one 2 in. square. On the other hand it 
 is not advisable to polish to small an area, because the surface will 
 be liable to convexity and therefor^ very difficult to get into 
 focus, especially for high powers. 
 
 Rough Polishing, — After the proper size of a specimen i& ob- 
 tained the next step is to give its surf ace a bright mirror-like polish, 
 free from any scratches discernible even with high powers of the 
 microscope (1,000 diameters or so), and this result is achieved 
 with the greatest economy in labor by proceeding by gradual 
 steps: 
 
 If' the surface contains deep gouges, or marks produced by cut- 
 ting or breaking it out, it should first be brought to a plane sur- 
 face by rubbing across a rough file, a grindstone, or an emery 
 wheel. For my own work I much prefer to rub the specimen 
 across a file rather than to put it in a vice and draw a file across it, 
 for I believe the former method produces a more even — i.e., less 
 rounded — surface and therefore conduces to economy of labor 
 in the later operations. It is best to hold the specimen lightly 
 in the fingers and draw it back and forth across the file in a 
 straight line,' avoiding any circular motion and therefore having 
 the polishing marks all parallel and straight across the specimen. 
 
 When a plane surface has been produced in this way the speci- 
 men should be rubbed on a very smooth file. In this operation 
 the specimen should again be held in the fingers and rubbed in a 
 straight line back and forth, and should be turned 90° from the 
 first rubbing so that the marks now made will cut vertically 
 across the first scratches. In this way it is very easy to tell when 
 the scratches made by the first file are entirely eliminated and the 
 operation on the second file should be continued to at least this 
 point no matter how short and faint the old scratches may prove 
 to be. It only takes a minute or two to remove the last scratches 
 on this smooth file, but it would take several minutes to remove 
 them by means of one of the later polishing mediums, and the 
 greatest economy is obtained by having each stage of the opera- 
 tion absolutely complete. 
 
 After coming off the smooth file, the specimen is again turned 
 90°, so that the marks now to be made run in the same direction 
 
444 THE METALLURGY OF IRON AND STEEL 
 
 as those made on the rough file, and rubbed across a sheet of 
 ordinary 00 emery paper cut to about 33/8 in. wide by 9 in. long 
 and pinned with thumb tacks upon a piece of planed, 1/2-inch 
 board. This operation is continued until the last marks from the 
 smooth file are removed. 
 
 Fine Polishing. — The polishing then proceeds in the same 
 manner by steps upon French emery paper of gradually increas- 
 ing fineness, each piece being cut to about 33/8 in. by 9 in. and 
 mounted on a smooth board. Of the Hubert brand the grades- 
 are designated 0, 00, 000, and 0000. 
 
 After the 0000 French emery, the surface is very smooth and 
 bright and is given a final burnishing by rubbing across a piece of 
 broadcloth or baize stretched over a piece of wood, moistened 
 with water and covered with a very thin liquid paste of water 
 and best washed rouge. This should leave the specimen polished 
 as bright as a mirror and free from all scratches. It may be, 
 however, that the eye or a hand magnifying glass would discover 
 microscopic rounded furrows in the s-pecimen. These are due to 
 scratches made in the early stages of polishing, which have not 
 been eliminated, but whose corners have been rounded off by the 
 finer grades of polishing mediums. Never permit a specimen 
 containing these furrows to be used as the basis of any opinion. 
 
 Preparation of Rouge. — The best jeweler's rouge purchasable 
 is not good enough for polishing, as it contains very fine particles 
 of dirt and grit which produce scratches in the surface of the 
 specimen during the final stages of the process. It is best 
 washed by the metallographer himself. This is best done upon 
 samples of not more than one teaspoonful of rouge at a time 
 stirred in about a glassful of water in a flat pan or dish until 
 thoroughly wetted. After allowing to settle for about five 
 minutes the water is poured into an ordinary chemist's wash 
 bottle where it is kept until ready for use. The last dregs of the 
 water, containing a good de^l of coarse rouge and grit, is thrown 
 away. A little experience soon teaches one to get the maximum 
 amount of good rouge from a sample, without any particles that 
 would produce scratches. The rouge in the wash bottle is pro- 
 tected from dust and dirt and can be poured out of the glass tube 
 on to the broadcloth polishing board as needed. For the prepara- 
 tion of special powders for the very finest grades of polishing, the 
 reader is referred to the references at the end of this chapter. 
 
 Precautions as to Polishing. — Do not rub the specimen too 
 
THE METALLOGRAPHY OF IRON AND STEEL 445 
 
 hard on the polishing mediums. This does not produce the 
 desired effect any more rapidly and may distort the structure 
 of the metal so as to lead to erroneous conclusions. 
 
 Do not allow the specimen to become heated by the polishing. 
 This is especially true of hardened steel and other heat-treated 
 specimens which may become tempered and so altered even upon 
 gentle heating. 
 
 Rub a piece of hard steel over each piece of polishing paper and 
 broadcloth before using it for polishing your specimen. This is 
 to get rid of grit. 
 
 Do not lay polishing boards down where dust will get on them, 
 but let them stand with the full height upward inside a closed 
 box or a small closet. 
 
 Never form an opinion upon a specimen that retains scratches 
 or polishing marks. 
 
 Mechanical Polishing. — Instead of rubbing the specimen across 
 different mediums by hand, we can press them against the emery 
 papers and rouge-cloth mounted upon wooden discs about 8 in. 
 in diameter, revolving at speeds of about 600 r.p.m. This 
 method saves time and the necessary apparatus is very simple to 
 make, or can be purchased complete.^ Hand-polishing is 
 preferable except upon the rouge, because of the liability to 
 heating the specimens with the higher surface-speed, and to 
 damage of the specimen by having it snatched out of one's 
 fingers. , 
 
 Developing the Structure for Examination 
 
 To develop the structure of iron and steel so as to differentiate 
 between the constituents, four methods are available: 
 
 (1) Polishing in bas-relief. 
 
 (2) Etching with chemicals. 
 
 (3) "Polish attack," and 
 
 (4) Heat tinting. 
 
 Polishing in Bas-relief. — Where some of the constituents are 
 less durable than others the method of polishing that I have de- 
 scribed, upon a soft background, produces a bas-relief, since the 
 softer constituents are worn to a greater depth than the harder 
 ones. The parts thus worn down appear darker than the higher 
 places which reflect the light better. It is by this method that 
 
 1 Consult references ISO. 184 and 185. 
 
446 THE METALLURGY OF IRON AND STEEL 
 
 graphite is best distinguished in pig iron and slag in wrought iron, 
 because the attack by acids is liable to produce other dark spots 
 which may not be readily differentiated. It is by the bas-relief 
 method that.F. Osmond developed so beautifully the structure of 
 pearlite shown in Fig, 247. This method has the disadvantage, 
 however, of rounding of the edges of the harder constituents and 
 so causing the softer ones to appear larger than they really are. 
 
 Etching with Chemicals — Nitric Acid. — This method is probably 
 the commonest one of developing the structure of steel. Many 
 different strengths of acid are used by different metallographers 
 from 0.1 per cent, up to 20 per cent., and the length of time that 
 is necessary to expose the specimen will depend upon this factor 
 and upon the amount of carbon present. High-carbon steel will 
 require a longer time than soft steel and may take as much as 
 2 1/2 minutes, although this is very rare. The nitric-acid solu- 
 tions are usually made with alcohol instead of with water, in order 
 to dry more quickly. Some metallographers prefer to immerse 
 their specimens in the acid for a given length of time, and others 
 prefer to hold the specimen in the hand with the polished surface 
 upward, and then deposit a few drops of the etching fluid upon 
 it with a piece of rubber tubing on the end of a glass rod. In 
 any event it is usually better to etch a shorter time than is 
 estimated as suitable, examine under the microscope, and then 
 etch again if necessary. After every etching it is necessary 
 to wash the specimen off with alcohol and dry as quickly as 
 possible. This drying is best accomplished by absorbing all 
 the liquid left on the surface after the alcohol wash with a piece 
 of soft cloth and then holding in a stream of air. Some metal- 
 lographers wash the sj)ecimen after etching with alkali, and 
 then with water and then with alcohol. Sauveur immerses his 
 specimen in the strongest nitric acid for a few seconds and then 
 places it in a stream of running water, washes with alcohol and 
 dries. As the strong nitric acid puts the steel in a passive state, 
 the length of attack by this method is only momentary, so that 
 it is necessary to repeat it a few times. It results in a very even 
 etching of the surface. 
 
 Iodine Etching. — Some investigators use iodine for etching, 
 by rubbing a few drops over the surface until a film covers it, and 
 allowing it to remain until the iodine color has disappeared, 
 after which the iodine is washed off with alcohol and dried. A 
 good iodine solution for this purpose is the ordinary tincture of 
 
THE METALLOGRAPHY OF IRON AND STEEL 447 
 
 iodine diluted with an equal volume of alcohol. It is also well to 
 have an auxiliary solution of about one-fourth this strength 
 for lighter etching work. 
 
 Picric Acid Etching, — I have found the 5 per cent, of picric 
 acid in alcohol, which Igevsky used for hardened and annealed 
 steels, very useful also for wrought iron, very low carbon steels, 
 and pearlite. 
 
 Polish Attack. — The method of "polish attack" advised and 
 used by Osmond, consists in rubbing the specimen upon a piece 
 of parchment stretched over a piece of soft wood and moistened 
 with a 2 per cent, solution of ammonium nitrate. This method 
 gives very beautiful results in Osmond's hands and is especially 
 valuable for developing the structure of martensite, troostite, 
 pearlite, and sorbite. 
 
 Heat Tinting, — Heat tinting consists in warming the steel 
 until it becomes oxidized. The different constituents are oxi- 
 dized at a different rate and so may be distinguished from one 
 another. It is most serviceable in distinguishing the phosphide 
 of iron from the carbide, for by heating until the carbide is red, 
 the phosphide (FegP) will be yellow. The phosphorus eutectic 
 may be distinguished from pearlite in the same way, because the 
 former will be yellow when the latter is blue. The heat tinting 
 must be accomplished in such a way as not to expose the metal 
 directly to a flame. The simplest method is to put the specimen 
 upon an iron plate heated from beneath. 
 
 Microscope and Accessories 
 
 It is hardly desirable to describe here the different forms of 
 microscope and accessories used for metallographic work, as the 
 manipulations cannot well be learned, except by practice, and 
 the catalogues of the manufacturers of scientific instruments of all 
 iron-producing countries now give full data. A photograph of a 
 common illuminating device, microscope and camera, is shown in 
 Fig. 338. Those who expect to take up the subject should pro- 
 cure a book upon it and study it much more fully than we have 
 space for here. The chief difference between the microscopic 
 examination and photography of iron and steel, and of biological 
 specimens, botanical specimens, thin rock sections, etc., arises 
 from the non-transparency of the metaL It is necessary to 
 illuminate them from above, since we cannot cause the light to 
 
448 
 
 THE METALLURGY OF IRON AND STEEL 
 
THE METALLOGRAPHY OF IRON AND STEEL 
 
 449 
 
 pass through the specimen and into the instrument. This 
 necessitates special forms of illuminators and powerful lights. 
 For magnifications of about 500 diameters, the Welsbach mantle 
 gives sufficient illumination. The Nernst lamp is also very- 
 useful, but for very high powers it is necessary to use the electric 
 arc lamp. This introduces especial difiiculties, because the focus 
 
 FIGS. 339 TO 342.— RAILS ETCHED FOR SEVERAL HOURS WITH DILUTE 
 
 HYDROCHLORIC ACID. 
 FIG. 339.— RAIL WITH SOFT INTERIOR; ROLLED FROM TOO NEAR TOP 
 
 OF INGOT 
 FIG. 340.— RAIL WITH BLOWHOLES AND SOFT INTERIOR. 
 
 which gives good definition to the eye by means of the arc light 
 will be blurred upon the photographic plate. The use of a light- 
 yellow screen in front of the light assists in this difficulty, but a 
 good deal of experience is required to get good results. A 
 mercury light avoids this difficulty. 
 
 29 
 
450 THE METALLURGY OF IRON AND STEEL 
 
 Magroscopic Metallography 
 
 An experienced eye may get a very good idea of the size of 
 crystals in iron and steel by examining a freshly broken fracture, 
 and this is one of the branches of magroscopic metallography. 
 For -the most accurate results, however, it is not as good as 
 microscopic examination. 
 
 It is also possible to get other information by etching a polished 
 surface for many hours with dilute hydrochloric acid. For this 
 purpose the polishing need only go as far as the commercial 00 
 emery paper, following the smoother file. In this way the center 
 of ingots, because of their looser texture, will be eaten away much 
 more rapidly, and this will be evident to the unaided eye. Also 
 the interior of sections of large area will be attacked more than 
 the outside, which has become harder through the work of rolling. 
 Some blowholes, which cannot be seen by eye or have been par- 
 tially welded up, may often be discovered, because the etching 
 action is more severe in their neighborhood, and the same is true 
 of spots where segregation has occurred. After etching and 
 examination, permanent records may be kept of the indications 
 by covering the etched surface with printer's ink and then press- 
 ing it with a letter press on to a piece of cardboard or heavy paper. 
 A few examples of records made in this way are shown in Figs. 
 339 to 342. 
 
XIX 
 
 METALLURGICAL FUELS 
 
 Metallurgical fuels may for convenience be divided into three 
 classes: (1) solid fuels, which include wood, charcoal, anthracite 
 anthracite coal, bituminous coal and coke; (2) liquid fuels 
 comprising chiefly crude petroleum and its products and tar, 
 and (3) gaseous fuels, including natural gas, producer gas, 
 water gas and special manufactured gases. 
 
 Wood. — Wood burns rapidly with a long flame and is there- 
 fore adapted for reverberatory furnace heating, but requires a 
 good deal of labor for firing. The flame from wood has the 
 advantage of being relatively free from sulphur which is a 
 harmful impurity in the flame from bituminous coal. In Table 
 XXXV is shown Gruener's Classical Classification of Solid 
 Fuels, from which it will be observed that the volatile matter 
 of wood is relatively very high and that the fixed carbon is 
 relatively low in comparison with the other solid fuels. 
 
 Charcoal. — Charcoal is made from wood by driving off the 
 volatile matter and leaving the fixed carbon. This process of 
 manufacture which is known as "charring" or carbonizing, is 
 accomplished by heating the wood out of direct contact with 
 the air. The resulting product is a porous, black, non-lustrous 
 material, pulverulent, without strength. It consists chiefly of 
 carbon with traces of residual volatile matter and small amounts 
 of ash, the latter representing the inorganic solid matter initially 
 contained in the wood. Charcoal burns substantially without 
 flame, and is therefore inapplicable to reverberatory and similar 
 furnaces, but is valued in direct-contact furnaces, such as crucible 
 furnace, cupolas, blast furnaces, etc., on account of its porosity 
 and also its relative freedom from objectionable impurities, such 
 as sulphur and phosphorus. The lack of strength of charcoal 
 limits its usefulness in cupolas and blast furnaces because it 
 cannot support a heavy burden of superincumbent material. It 
 will be remembered that the solid fuel is the only unmelted 
 material below the melting zone of blast furnaces and cupolas, 
 and that it therefore has to support as much of the weight pi 
 
 451 
 
452 
 
 THE METALLURGY OF IRON AND STEEL 
 
 
 2 
 
 8 
 1 
 
 1 
 
 • 
 
 i 
 
 
 
 h^ 
 
 
 3 
 
 
 3 
 
 J ^ 
 
 (^ 
 
 5 
 
 ^ 
 
 
 
 
 H] 
 
 
 1 
 1 
 
 
 1 
 1 
 
 1 
 
 1 
 
 d V 
 
 3^ 
 
 ll 
 
 It 
 
 g-i 
 
 t 
 s 
 
 1 
 
 
 
 
 '^ 
 
 ■■ 
 
 
 ^ 
 
 cd o 
 
 <S cL 
 
 C a 
 
 ^ 
 
 
 
 
 p. 
 
 f^ 
 
 ^ 
 
 ft 
 
 Ph " 
 
 O 
 
 O ** 
 
 ^ 
 
 P=< 
 
 Ai 
 
 
 
 c 
 
 iC 
 
 
 
 o 
 
 CO 
 
 -* 
 
 01 
 
 CM 
 
 (M 
 
 
 
 ^ 
 
 I 
 
 K 
 
 
 
 o 
 
 o 
 
 :t 
 
 Z 
 
 CM 
 
 O 
 
 
 
 
 o 
 
 K 
 
 « 
 
 lO 
 
 CD 
 
 CO 
 
 t>- 
 
 00 
 
 OS 
 
 
 
 
 
 tr. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 I— 1 
 
 o 
 
 t- 
 
 «o 
 
 <M 
 
 
 
 
 
 
 
 
 1-1 
 
 
 w 
 
 <N 
 
 tH 
 
 rH 
 
 rH 
 
 00 
 
 
 
 
 
 ss 
 
 
 4 
 
 
 4 
 
 1 
 
 o 
 
 i 
 
 «!, 
 
 J, 
 
 1 
 
 CM 
 
 
 
 o 
 
 
 o 
 
 
 a 
 
 
 (N 
 
 (N 
 
 (N 
 
 .-I 
 
 rH 
 
 i-l 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t-H 
 
 !;l 
 
 
 
 
 
 
 
 
 
 
 
 
 
 % 
 
 1 
 
 1 
 
 
 '^ 
 
 
 Tt< 
 
 »o 
 
 cq 
 
 o 
 
 
 
 
 
 o 
 
 t— 1 
 
 a 
 
 u 
 
 
 i> 
 
 
 r-i 
 
 •-I 
 
 iH 
 
 tH 
 
 "? 
 
 ^ 
 
 
 
 1 
 
 
 J: 
 
 
 ci 
 
 at) 
 
 J> 
 
 ^ 
 
 <^ 
 
 Js 
 
 
 
 & 
 
 9 
 
 
 10 
 
 
 tH 
 
 tH 
 
 rH 
 
 1-) 
 
 l-H 
 
 
 
 
 CQ 
 
 S 
 
 
 
 
 
 
 
 
 
 
 
 
 
 D3 
 
 < 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f> 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 lO 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 t 
 
 
 1-1 
 
 "P 
 
 7 
 
 j 
 
 T 
 
 ? 
 
 
 
 p^ 
 
 
 S 
 
 
 ^ 
 
 
 ^ 
 
 
 •o 
 
 CO 
 
 i-< 
 
 iH 
 
 
 
 ^ 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 p 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 
 •ts 
 
 
 W 
 
 
 o 
 
 o 
 
 (N 
 
 CD 
 
 QO 
 
 o 
 
 
 
 ^ 
 
 ^ 
 
 
 2 
 
 
 2 
 
 
 
 S 
 
 C^ 
 
 3 
 
 i 
 
 > 
 
 
 
 i> 
 
 
 CD 
 
 lO 
 
 ^ 
 
 CO 
 
 . N 
 
 1-H 
 
 1-1 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 ^. 
 
 
 
 
 
 
 
 
 
 
 tti 
 
 X 
 
 
 6 
 
 
 
 
 
 
 
 
 
 
 l> 
 
 
 d 
 
 
 
 
 
 lO 
 
 W3 
 
 CO 
 
 cq 
 
 1-1 
 
 rH 
 
 T-t 
 
 o 
 
 ^ 
 
 1 
 
 ^ 
 
 N 
 
 
 c^ 
 
 
 4 
 
 1 
 cc 
 
 1 
 
 
 
 rH 
 
 h^ 
 
 
 (^ 
 
 
 
 
 
 
 
 
 
 
 
 PQ 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 a 
 
 03 
 
 « 
 
 
 lO 
 
 o 
 
 ^-1 
 
 IC 
 
 »o 
 
 
 
 
 1 ' 
 
 O 
 
 
 ^ 
 
 1 
 
 I— 1 
 
 "-1 
 
 en 
 
 C3J 
 
 1 
 
 i 
 
 I— ( 
 
 
 7 
 
 U3 
 
 CO 
 
 
 1 * 
 
 
 
 ■* 
 
 CO 
 
 CI 
 
 
 tH 
 
 r-t 
 
 CO 
 
 « 
 
 
 
 
 
 m 
 
 lO 
 
 
 Ui 
 
 
 »o 
 
 lO 
 
 
 
 
 M 
 O 
 
 w 
 
 iH 
 
 CO 
 
 ^ 
 
 « 
 
 7 
 
 CO 
 
 CO 
 
 i 
 
 U5 
 
 CO 
 
 
 ^ 
 
 uj 
 
 
 : 
 
 
 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 
 
 b- 
 
 
 
 
 
 
 
 
 
 
 
 
 ■* 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 T] 
 
 .H 
 
 K 
 
 i? 
 
 o 
 
 U5 
 
 OS 
 
 fH 
 
 CO 
 
 U5 
 
 
 
 
 3 
 
 US 
 
 
 :i 
 
 2 
 
 O 
 
 ob 
 
 
 ? 
 
 CO 
 
 
 
 
 
 >o 
 
 
 CO 
 
 1>- 
 
 00 
 
 00 
 
 CO 
 
 OS 
 
 m 
 
 
 
 
 
 
 
 
 a' 
 
 OJ 
 
 o tj 
 
 Q> 
 
 u 
 
 
 
 
 
 
 
 
 
 
 E 
 
 
 S<H 
 
 E 
 
 s- 
 
 
 
 
 1 
 4) 
 
 
 
 
 
 "rt 
 
 if, 
 
 •2 g 
 
 Its- « 
 
 
 
 «g 
 
 Q bO 
 
 d 
 
 1 
 
 
 
 
 i 
 
 
 I 
 
 1 
 
 ■^ 
 
 i 
 
 a 
 
 
 cao 
 
 
 i2 " 
 
 11 
 
 -■§ 
 
 
 
 
 1 i 
 
 ! J 
 
 
 jBOQ snotnnni!).!^ 
 
 
 "l 
 
 < 
 
 1 1 
 
METALLURGICAL FUELS 453 
 
 the material in the shaft as is not held up by the pressure of the 
 blast. Charcoal is used in some iron blast furnaces of small 
 size because of its purity. This is especially true in Sweden, 
 where the advantage of pure iron ores would be to some extent 
 nullified if an impure material like coke were used for smelting. 
 
 The manufacture of charcoal is a very simple process, and 
 consists in a slow combustion of the gases "Which the wood gives 
 off when heated. The process may be carried on in heaps or 
 kilns. In the former practice, the wood is piled up in a regular 
 heap and covered on all sides with earth and sods. Openings 
 are left at the bottom for the admission of a small amount of 
 air and the burning is started by introducing fire at these 
 points. The admission of air thereafter is regulated with great 
 care to the different parts of the heap, so that the gases will 
 slowly burn without there being much fixed carbon oxidized 
 at the same time. The burning occupies many days, and then 
 the heap is torn down and the charcoal is ready for use. Instead 
 of burning in heaps, the wood may be piled in kilns, which have 
 the advantage of permanency. In the latter case, we may 
 carry away the gaseous products of combustion of the charring 
 operation and separate many valuable by-products, such as 
 wood tar, wood alcohol, acetic acid, etc. 
 
 Peatj Lignite^ Etc. — ^Peat, lignite, brown coal, etc., represent 
 stages in the natural carbonization" of wood to coal in the bosom 
 of the earth. In the majority of cases, these fuels are too impure 
 and too high in ash to be burned on grates. They are sometimes 
 converted into gas in gas producers, and they are often crushed, 
 washed and agglomerated into briquettes, which makes an 
 admirable substitute for bituminous coal and is rapidy growing 
 yearly in the amount used. 
 
 Bituminous Coals, — On account of the volatile matter in 
 bituminous coals, they burn with a flame, and are therefore well 
 adapted to use in reverberatory furnaces. The richest in gas 
 are often used for the manufacture of illuminating or city gas, by 
 a simple process of driving off volatile matter by means of heat, 
 but this is a too expensive gas to use ordinarily for metallurgical- 
 purposes. Some of the bituminous coals on being heated soften 
 and swell as they give off the volatile components. The softening 
 is sufficient to cause these coals to cake together under the 
 pressure of their own weight, and to consequently form porous, 
 coherent masses of a silvery-gray material, known as coke, which 
 
454 THE METALLURGY OF IRON AND STEEL 
 
 contains only traces of residual volatile matter, together with 
 ash, representing chiefly the inorganic matter originally con- 
 tained in the coal. The chief chemical constituent of coke is 
 carbon, amounting often to 87 or more per cent, of the mass. 
 Coke bears approximately the same relation to bituminous coal 
 that charcoal does to wood. 
 
 Coke. — Coke is made by distilling the volatile matters from 
 bituminous coal; this distillation takes place in chambers in 
 which only enough air is admitted to supposedly burn the vola- 
 tile matters, or else in chambers from which the air is excluded 
 completely. The chambers into which a limited amount of air 
 
 FIG. 350.— STANDARD AMERICAN BEEHIVE OVEN. 
 
 is admitted are called "beehive ovens", on account of their form, 
 which is shown in Fig. 350. The chambers into which no air is 
 admitted are known as "retorts'', of which a cross-section is 
 shown in Fig. 351. In Europe and England, the retort oven is 
 commoner, but America makes more coke in beehive ovens, which 
 is a far more wasteful method, since only 66 per cent, of coke is 
 obtained from a ton of coal, and no "by-products"; whereas, in 
 retort ovens, we obtain a yield of 75 per cent, of coke and also by- 
 products consisting of tar, ammonia, excess gas, etc., equal in 
 total value to about one-third of the value of the coke. In the 
 United States there are some 104,000 beehive ovens, producing, 
 in 1910, an average of 376 tons of coke each per year; and about 
 
METALLURGICAL FUELS 
 
 455 
 
 4000 by-product ovens, producing, in 1910, an average of 1762 
 tons of coke per year. The total production of by-product ovens 
 was 17 per cent, of all in the United States in 1910. 
 
 Coking in Beehive Ovens. — The crushed coal is poured into the 
 
 FIG. 351— HORIZONTAL SECTION THROUGH A RETORT OF A BY-PRODUCT 
 
 OVEN. 
 
 oven through the "eye" at the top, and carefully levelled. 
 Combustion is started at the middle of the top circle of coal, and 
 gas begins to be distilled off by the heat; air is admitted through 
 the arch of the lower door, and flame soon fills the dome and 
 sends a black, sooty smoke out of the eye. The heat from the 
 
 FIG. 352.— BATTERIES OF BEEHIVE OVENS. 
 
 flame distills the volatile matter from the coal, advancing layer by 
 layer from the top; this causes the coal to swell and rise slightly. 
 At the end of the coking a five-ton charge of coke 28 in. thick wiU 
 have risen to 34 in. thick. When distillation is complete to the 
 
456 THE METALLURGY OF IRON AND STEEL 
 
 bottom of the charge, the combustion is stopped by quenching 
 the coke with a stream of water, this action taking place inside 
 the oven itself in order that air may not oxidize the surface of the 
 coke and dull its silvery color. The coke breaks up into long 
 pieces, as shown in Fig. 353. A small amount (say 2 to 2 1/2 
 per cent.) of fine particles, or " breeze," is formed which cannot be 
 marketed. 
 
 Coking in Retort Ovens, — In retort ovens the coal is completely 
 enclosed during distillation, which accomplishes three purposes: 
 (1) it enables the gas distilled from it to be led off into by-pro- 
 
 FIG. 353.— STRUCTURE OF BEEHIVE COKE- 
 
 duct-saving apparatus; (2) it protects the fixed carbon in it from 
 contact with air, which would burn it and reduce the yield, and 
 (3) it puts it under greater pressure during the swelling operation, 
 so that some coals which will not cake into a coherent mass in bee- 
 hive ovens, will make good strong coke in retorts. The retorts 
 are made of fire-brick, and are heated by burning gas outside of 
 them on three sides. Good coking coals will yield from 5000 to 
 12,000 cu. ft. of gas per ton. In the beehive oven this is all 
 burned; in retort ovens we burn about one-half of it to produce 
 the heat required for the coking operation, and the remainder is 
 available for sale, or other purposes. In all the retort ovens of 
 the United States in 1910, an average of nearly 3000 cu. ft. of 
 excess gas was obtained per ton of coal used; together with about 
 7 gal. of tar, and ammonia products in value equal to over 40 
 
METALLURGICAL FUELS 
 
 457 
 
 cents. The total value of by-products recovered is equal to 
 about one dollar per ton of coal coked. 
 
 Surplus Gas Obtained. — The analysis of the gas given off by 
 the coal will vary with the time of. coking — that is, the first gas 
 distilled off which is given up more readily, will be richer than 
 that given off later. This is shown by Table XXXVI.* 
 
 From these figures it will be noticed that the coal also parts 
 with its volatile matter more freely at first — there is more gas 
 given off per hour; secondly, the components of the gas which 
 
 FIG. 354.— OTTO-HOFFMANN BY-PRODUCT, OR RETORT. COKE OVENS. 
 
 give the most illuminating power (CmHn) are relatively higher in 
 the gas first given off than in the later gas, and therefore the 
 candle-power of the gas decreases as coking progresses; thirdly, 
 that the calorific power of^ the gas decreases also, but not at so 
 rapid a rate as the candle-power, especially excepting the last 
 few hours of the operation. Now, if all the gas distilled off in 
 coking a gharge of coal was stored in a reservoir it would not be 
 rich enough to use for illuminating purposes for cities, but, if we 
 should store the first gas given off in one reservoir, and the later 
 gas in another, we would have a supply of sufficiently rich gas 
 and a supply of poorer gas. This admirable arrangement, a 
 result of the genius of Dr. Schniewind, is in fact adopted at 
 
 1 Taken from pr. F. Schniewind's paper on the "Production of Illuminating Gas from 
 Coke Ovens," read before the Gaa Section of the Engineering Congress at Glasgow. 
 
458 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXXVI.— THE ANALYSIS OF GAS GIVEN UP BY COAL 
 DURING COOLING 
 
 1 
 
 Analysis of gas produced. 
 
 *ow 
 
 1 
 
 II 
 
 I 
 
 II 
 
 •g« 
 
 o 
 
 Is 
 
 as produced during 
 ach hou]>-per long 
 ton of dry coal. 
 Cubic feet 
 
 Heat value of gas 
 
 produced during 
 
 ach hour^-per long 
 
 ton of dry coal 
 
 B.H.U. 
 
 g Value of 
 Bed of CO2 
 tiring each 
 ong ton of 
 landle feet 
 
 1 
 
 o 
 
 
 a 
 
 
 d 
 
 6 
 
 ^ 
 
 
 ■Pifl 
 
 K 
 
 o o 
 
 W 
 
 o 
 
 u 
 
 O 
 
 ^ 
 
 
 ^ 
 
 o 
 
 o« 
 
 o 
 
 1 
 
 5.8 
 
 40.3 
 
 34.3 
 
 6.8 
 
 3.9 
 
 0.2 
 
 8.7 
 
 .552 
 
 707 
 
 18.4 
 
 413 
 
 292,000 
 
 7,599 
 
 2 
 
 5.8 
 
 41.2 
 
 33.6 
 
 6.7 
 
 3.8 
 
 0.3 
 
 8.6 
 
 .547 
 
 707 
 
 18.4 
 
 333 
 
 236,000 
 
 6,127 
 
 3 
 
 5.8 
 
 41.8 
 
 34.0 
 
 6.5 
 
 3.7 
 
 0.3 
 
 7.9 
 
 .539 
 
 736 
 
 18.4 
 
 295 
 
 217,000 
 
 5,428 
 
 4 
 
 5.7 
 
 41.5 
 
 34.1 
 
 6.3 
 
 4.2 
 
 0.4 
 
 7.8 
 
 .555 
 
 699 
 
 16.2 
 
 312 
 
 218,000 
 
 5.054 
 
 5 
 
 5.7 
 
 41.5 
 
 34.8 
 
 5.8 
 
 3.9 
 
 0.3 
 
 8.0 
 
 .539 
 
 709 
 
 15.4 
 
 347 
 
 246,000 
 
 5,344 
 
 6 
 
 5.4 
 
 40.7 
 
 36.1 
 
 5.7 
 
 3.7 
 
 0.4 
 
 S.O 
 
 .527 
 
 703 
 
 14.4 
 
 357 
 
 251,000 
 
 5,141 
 
 7 
 
 5.4 
 
 39.6 
 
 37.6 
 
 5.8 
 
 3.5 
 
 0.4 
 
 7.7 
 
 .615 
 
 693 
 
 13.8 
 
 342 
 
 237,000 
 
 4,720 
 
 S 
 
 5.2 
 
 38.8 
 
 39.0 
 
 5.7 
 
 3.4 
 
 0.3 
 
 7.6 
 
 .502 
 
 683 
 
 14.5 
 
 350 
 
 239,000 
 
 5,075 
 
 9 
 
 5.0 
 
 37.8 
 
 40.2 
 
 5.9 
 
 3.3 
 
 0.3 
 
 7.5 
 
 .492 
 
 671 
 
 13.7 
 
 386 
 
 259,000 
 
 5.288 
 
 10 
 
 4.9 
 
 37.4 
 
 41.2 
 
 5.9 
 
 3.2 
 
 0.3 
 
 7.1 
 
 .486 
 
 666 
 
 13.3 
 
 344 
 
 229,000 
 
 4,575 
 
 U 
 
 4.9 
 
 az.o 
 
 41.5 
 
 a.2 
 
 3.2 
 
 0.3 
 
 6.9 
 
 .484 
 
 667 
 
 13.2 
 
 324 
 
 216.000 
 
 4^77 
 
 12 
 
 4.8 
 
 36.5 
 
 42.0 
 
 6.3 
 
 3.3 
 
 0.3 
 
 6.8 
 
 .482 
 
 666 
 
 13.2 
 
 362 
 
 241,000 
 
 4,778 
 
 13 
 
 4.7 
 
 36.2 
 
 41.9 
 
 6.1 
 
 3.4 
 
 0.4 
 
 7.3 
 
 .485 
 
 672 
 
 13.0 
 
 363 
 
 244,000 
 
 4,719 
 
 14 
 
 4.5 
 
 35.7 
 
 41.9 
 
 6.2 
 
 3.5 
 
 0.3 
 
 7.9 
 
 .485 
 
 665 
 
 12.3 
 
 337 
 
 224,000 
 
 4,145 
 
 15 
 
 4.4 
 
 35.3 
 
 42.1 
 
 6.2 
 
 3.3 
 
 0.4 
 
 8.3 
 
 .483 
 
 638 
 
 12.1 
 
 362 
 
 231,000 
 
 4,380 
 
 16 
 
 4.1 
 
 34.9 
 
 42.4 
 
 6.3 
 
 3.3 
 
 0.4 
 
 8.6 
 
 .478 
 
 624 
 
 11.9 
 
 364 
 
 227,000 
 
 4,331 
 
 17 
 
 3.8 
 
 34.7 
 
 42.9 
 
 6.5 
 
 3.0 
 
 0.4 
 
 8.7 
 
 .478 
 
 642 
 
 11.9 
 
 318 
 
 204,000 
 
 3.784 
 
 IS 
 
 3.7 
 
 34.7 
 
 42.9 
 
 6.6 
 
 3.2 
 
 0.4 
 
 8.5 
 
 .477 
 
 628 
 
 11.8 
 
 360 
 
 226,000 
 
 4,248 
 
 19 
 
 3.5 
 
 35.2 
 
 42.6 
 
 6.5 
 
 3.2 
 
 0.4 
 
 8.6 
 
 .478 
 
 623 
 
 11.1 
 
 345 
 
 215,000 
 
 3.829 
 
 20 
 
 3.4 
 
 35.1 
 
 42.7 
 
 6.5 
 
 3.0 
 
 0.4 
 
 S.4 
 
 .473 
 
 624 
 
 10.6 
 
 330 
 
 206.000 
 
 3,498 
 
 21 
 
 3.2 
 
 35.4 
 
 44.0 
 
 6.5 
 
 2.7 
 
 0.4 
 
 7.8 
 
 .458 
 
 624 
 
 11.0 
 
 346 
 
 216,000 
 
 3,806 
 
 22 
 
 3.1 
 
 34.8 
 
 45.4 
 
 6.3 
 
 2.5 
 
 0.4 
 
 7.5 
 
 .447 
 
 625 
 
 10.8 
 
 344 
 
 215,000 
 
 3,715 
 
 23 
 
 2.8 
 
 34.1 
 
 46.8 
 
 6.1 
 
 2.4 
 
 0.4 
 
 7.4 
 
 .434 
 
 608 
 
 10.3 
 
 396 
 
 241,000 
 
 4,079 
 
 24 
 
 2.5 
 
 32.3 
 
 48.7 
 
 5.9 
 
 2.2 
 
 0.4 
 
 8.0 
 
 .425 
 
 583 
 
 10.3 
 
 378 
 
 221,000 
 
 3.893 
 
 25 
 
 2.0 
 
 30.3 
 
 50.5 
 
 5.7 
 
 2.0 
 
 0.3 
 
 9.2 
 
 .415 
 
 556 . 
 
 9.0 
 
 320 
 
 178.000 
 
 2,880 
 
 26 
 
 1.5 
 
 26.8 
 
 54.5 
 
 5.5 
 
 1.8 
 
 0.3 
 
 9.6 
 
 .390 
 
 534 
 
 7.8 
 
 294 
 
 157,000 
 
 2,293 
 
 27 
 
 1.3 
 
 22.7 
 
 59.9 
 
 5.4 
 
 1.5 
 
 0.3 
 
 8.9 
 
 .361 
 
 510 
 
 6.3 
 
 286 
 
 146.000 
 
 1,802 
 
 28 
 
 1.0 
 
 19.2 
 
 64.0 
 
 5.4 
 
 1.3 
 
 0.3 
 
 8.8 
 
 .343 
 
 469 
 
 4.5 
 
 275 
 
 129,000 
 
 1,237 
 
 29 
 
 0.7 
 
 16.9 
 
 66.2 
 
 5.4 
 
 1.1 
 
 0.3 
 
 9.3 
 
 .341 
 
 456 
 
 3.8 
 
 228 
 
 104,000 
 
 866 
 
 30 
 
 0.6 
 
 15.2 
 
 66.4 
 
 5.6 
 
 0.8 
 
 0.2 
 
 11.2 
 
 .328 
 
 440 
 
 3.7 
 
 168 
 
 74,000 
 
 622 
 
 31 
 
 0.5 
 
 13.3 
 
 66.3 
 
 5.8 
 
 0.7 
 
 0.1 
 
 13.3 
 
 .330 
 
 389 
 
 3.8 
 
 144 
 
 56,000 
 
 547 
 
 32 
 
 0.4 
 
 12.0 
 
 66.2 
 
 6.3 
 
 0.8 
 
 0.1 
 
 14.2 
 
 .330 
 
 362 
 
 3.6 
 
 127 
 
 46,000 
 
 457 
 
 33 
 
 0.4 
 
 10.8 
 
 66.2 
 
 6.6 
 
 1.0 
 
 0.2 
 
 14.8 
 
 .335 
 
 362 
 
 3.5 
 
 94 
 
 34,000 
 
 329 
 
 34 
 
 0.2 
 
 9.6 
 
 66.7 
 
 6.4 
 
 1.3 
 
 0.2 
 
 15.6 
 
 .338 
 
 565(?) 
 
 2.5 
 Total, 
 
 46 
 
 26.000 
 
 115 
 
 
 10,390 
 
 6,501.000 
 
 122,981 
 
 several by-product coke plants, the rich gas being sold to cities 
 for illuminating purposes, and the poorer gas used for heating the 
 retorts in the coking operation. Before either gas is used, how- 
 ever, it is freed from tar and ammonia. When retort coke is 
 made at metallurgical works, the surplus gas may be used in open- 
 hearth furnaces, or in heating furnaces at the rolling plant, or 
 burned in gas engines for the production of power. When using 
 
METALLURGICAL FUELS 
 
 459 
 
 this gas in furnaces, the high content of hydrogen causes some 
 difficulty in burning out the brick-work unless care and experi- 
 ence is employed. Failures on this account have been frequent, 
 but the possibility of success is proven by practice in Germany 
 
 FIG. 355.— LONGITUDINAL SECTION OF SEMET-SOLVAY RETORT COKE OVEN". 
 
 and elsewhere, although the life of side-walls and roof is shorter 
 when coke oven gas is used than is the case with natural gas- 
 producer gas or oil. 
 
 Section SectTon Section 
 C C BB A_A_ 
 
 FIG. 356.— TRANSVERSE SECTION OF SEMET-SOLVAY OVEN, 
 
 Desiderata in Retort Coke Ovens. — The desirable features of 
 retort ovens are; 
 
 1. Uniform heating in all parts of the retort in order that the 
 operation may not be unduly prolonged or the coal unevenly coked. 
 On account of the length of the retorts, this feature has been 
 secured only after some failures and much experimenting by the 
 
460 
 
 THE METALLURGY OF IRON AND STEEL 
 
 pioneers of the art. It is now secured in the prominent types 
 of ovens by three means: 
 
 a. Uniform draught in all the flues ; 
 
 b. Effective regulation of combustion at all points; 
 
 c. Shortest possible passages for the burning gases. 
 
 2. Rapid conduction of heat through the 
 retort walls to the coal. This is secured in 
 various ways: 
 
 d. Thin retort walls; but this is qualified to 
 some extent by the third desirable feature 
 mentioned below. 
 
 e. High combustion temperatures; this is 
 obtained in many cases by regenerating the 
 air, -as in the open-hearth furnace. 
 
 3. Strong retort walls to resist the pressure 
 of the swelling of the coal. 
 
 4. Large output. This is obtained chiefly 
 by rapid conduction of heat to the coal, and 
 it is a very desirable feature, because retort 
 ovens are very costly to construct, and this 
 heavy overhead expense must be met by 
 large product. 
 
 A beehive oven will cost normally not more 
 than $300 erected, while each retort of a by- 
 product plant will cost from $2000 to $3500, 
 which will be raised to $4500 to $7000 in- 
 cluding the by-product-saving apparatus. 
 (f) Simplicity of operation and (g) use of 
 mechanical devices are other means by which 
 the output is increased, and also labor costs 
 decreased. 
 
 5. Economical heating. — This is obtained 
 by recovering the waste heat in the outgo- 
 ing products of combustion by regenera- 
 tion or recuperation, by uniform heating, by effective regulation 
 of combustion, and by complete combustion; i.e., absence of 
 smoke. All of these desiderata are of minor importance com- 
 pared to rapid conduction of heat through the retort walls 
 already mentioned, 
 
 6. Economy in Repairs. — ^This is another very desirable 
 feature, because the repairs are not only a large factor in the cost 
 
 FIG. 357.— CROSS-SEC- 
 TION OF RETORT. 
 Structure of By-product 
 Coke. 
 
METALLURGICAL FUELS 
 
 461 
 
 of coking, but time used for repairing reduces output. The 
 means by which this economy is most commonly secured are: 
 
 h. Ready accessibility of all parts, for renewal and inspection. 
 
 i. Simple and strong construction. 
 
 j. Avoiding damage to one part of the oven by the operation 
 of another; for example: if the regenerators are placed under- 
 neath a heavy weight of brick, their expansion and contraction 
 
 (a) 
 
 FIG. 358.— THE OTTO-HOFFMANN COKE OVEN. (SCHNIEWIND TYPE). 
 
 damages the brickwork above, and the weight of the latter brings 
 a burden on the regenerative chambers while they are hot, and 
 therefore weak. 
 
 Different types of retort ovens have been devised to secure these 
 desirable features, but space is not available here to describe 
 them in detail. The following ovens are in use in the United 
 States: 
 
 Semet-Solvay 1387 ovens. 
 
 United Otto 2104 ovens (Otto-Hoffmann type). 
 
 Rothenberg 307 ovens. 
 
 Koppers 280 ovens. 
 
462 
 
 THE METALLURGY OF IRON AND STEEL 
 
 Producer Gas. — If air be blown through red-hot carbon the 
 following reaction takes place: 
 
 0+20=00^; 
 
 but if the bed of fuel is deep, the carbon dioxide enters into a 
 further reaction, as follows: 
 
 C0„+C=2C0. 
 
 .^>C«.'«0*«y^-«^»v'«.-«;»^>K-«C«M'N'*C«.-t»Ct^^^ 
 
 FIG. 359.-OTTO-HOFFMANN OVEN. 
 
 In other words, if air be made to blow through a deep bed of red- 
 hot carbon, there will be produced carbon monoxide gas which has 
 combustible value: 
 
 CO + O-COg (generates 68,040 calories). 
 
 Producer gas for open-hearth and reverberatory furnaces is usu- 
 ally thus made from bituminous coal, because the hydrocarbons 
 contained in this coal enter into the gas and thus give it illuminat- 
 ing power, which makes it much more efficient in the furnace, 
 
METALLURGICAL FUELS 
 
 463 
 
 because the heating takes place by radiation chiefly. Such a gas 
 will contain 3 to 5 per dent, of hydrocarbons, 20 to 25 per cent, of 
 CO, 55 to 60 per cent, of nitrogen, and 2 to 8 per cent, of carbon 
 dioxide. 
 
 The two latter components produce no heat and are therefore 
 worse than useless, because they carry heat away from the 
 furnace up the chimney stack. The nitrogen comes from the 
 air, of course, but the CO 2 is theoretically absent. Its presence 
 
 B (c) 
 
 FIG. 360.-OTTO-HOKFMANN OVEN. 
 
 is due to irregularities in the gas producer operation, such as 
 vertical channels forming in the bed of fuel, up which the COg 
 gas passes without being brought into contact with carbon; or 
 the rapid passage of the gas does not permit time for the reactions 
 to be completed; or irregularities in the fuel bed, whereby the 
 fuel will be red-hot much higher on one side than on the other. 
 
 The air is usually blown through the fuel by means of a steam 
 jet, which results in a certain amount of steam passing into the 
 producer with the air; but this is rather an advantage than other- 
 
464 
 
 THE METALLURGY OF IRON AND STEEL 
 
 wise as the steam is decomposed by the red-hot carbon and 
 enriches the producer gas: 
 
 H2O+C-2H + CO; 
 (2H + = H20: generates 58,060 calories.) 
 
 Gas Producers, — The gas producers are the furnaces in which 
 the fuel is contained while the air is passed through it. The main 
 objects to be accomplished are: (1) To pass the air uniformly 
 c 
 
 ^■v ■■ 
 
 v'^^-:';-'a::l' 
 
 
 
 vi?--:: 
 
 ;V/.':::\i?i-^;' 
 
 ;■■■: ■■■::/<^'-' 
 
 -» (d) 
 
 FIG. 361.-OTTO-HOFFMANN OVEN. 
 
 through the bed; (2) to remove ashes and charge fresh fuel with- 
 out interrupting the production of gas; and (3) to preserve the 
 deep bed of incandescent carbon, having level upper and lower 
 surfaces. There are three horizontal zones in the gas producer: 
 The first is the ash zone, which is deep in order that the air may be 
 slightly preheated in passing through it and that any unburned 
 carbon which gets into it may have a strong liability of being 
 burned. Next above that is the CO 3 zone, in which the oxygen 
 
METALLURGICAL FUELS 
 
 465 
 
 and carbons are first combined; and above that the CO zone, in 
 which the CO 2 is reduced by more carbon. The top of this zone 
 should be at a dull-red heat. 
 
 There are many different forms of producer, which are exten- 
 sively used for open-hearth work, but these may be divided into 
 two general types: In the first or water-sealed type, the bottom 
 of the producer dips into a pool 
 of water and thus the tools may 
 be introduced for the removal of 
 the ashes at will. In this type 
 there are sometimes steel arms 
 extending into the bed of fuel, 
 either from the top or from a 
 central shaft, by the rotation of 
 which the bed is poked, lumps 
 and channels are broken up, etc. 
 The second type has a mechan- 
 ical grate, by which the ashes 
 can be scraped down into the 
 chamber underiieath without in- 
 terrupting the producer opera- 
 tion for the purpose. 
 
 Grate Area. — The total grate 
 area of all the producers supply- 
 ing gas to a furnace should be 
 about 3.5 sq. ft. per ton of 
 furnace capacity; some producer 
 plants run even higher than this, 
 and up to 6.25 sq. ft. Another 
 method of figuring the grate area 
 is that 1 sq. ft.> should be supplied for every 7.5 to 12.5 lbs. 
 of coal burned per hour, although with expert gas makers and 
 good coal the combustion may be much greater than this, and 
 higher values (to 22 lbs.) are claimed by the makers of the gas 
 producers. 
 
 Volume q,nd Calorific Power. — The volume of producer gas 
 obtained from a ton of coal will be about 150,000 to 170,000 cu. ft., 
 having a calorific power of 33 to 36 calories per cu. ft., or 130 to 
 145 B.t.u. per cu. ft.^ These figures will, of course, depend 
 upon the quality of the coal gasified, but the calorific power is no 
 
 1 For calculation of this relation, see page 473, 
 30 
 
 FIG. 362.— TAYLOR REVOLVING 
 BOTTOM GAS PRODUCER. 
 
466 
 
 THE METALLURGY OF IRON AND STEEL 
 
 FIG. 363.— MORGAN WATER-SEALED GAS PRODUCER. 
 
METALLURGICAL FUELS 
 
 467 
 
 more important than the amount of heat that it will radiate, 
 which depends upon the luminosity of the flame. 
 
 Luminosity, — The luminosity of flames depends upon the 
 amount of hydrocarbons, and especially of heavy hydrocarbons, 
 
 FIG. 364.— HUGHES MECHANICALLY POKED GAS PRODUCER. 
 
 burned to produce them. It is therefore necessary, if the pro- 
 ducer gas is made from bituminous coal low in hydrocarbons or 
 from coke or anthracite, to increase its illuminating power by 
 spraying oil into it. The luminosity is produced by the deposition 
 
468 
 
 THE METALLURGY OF IRON AND STEEL 
 
 of a myriad of tiny particles of carbon, which are heated to incan- 
 descence and then radiate energy in the form of light. It is prob- 
 able that this action is produced by the relatively light hydro- 
 carbons, such as methane (CH^), breaking up first into ethylene 
 (C2H4), and then into acetylene (C2H2), which deposits the carbon 
 particles or soot. It is for this reason that the pure acetylene 
 flame has such intense luminosity. 
 
 Gas Mains, — The gas mains leading from the producer plant 
 to the open-hearth furnace should be lined with brick and be at 
 
 FIG. 365.- 
 
 -BUTTERFLY REVERSING VALVE. 
 (See also Fig. 28, page 49.) 
 
 least large enough for a man to pass through. Beyond this, a 
 good rule is 1 sq. ft, of area of cross-section of gas main for every 
 8 sq. ft. of total combined area of gas-producer grates. The gas 
 loses heat by radiation in the mains and deposits the tarry con- 
 stituents, i.e., the nearly solid hydrocarbons, in both ways 
 losing heating power. 
 
 Natural Gas. — In those districts, like Pittsburg, where natural 
 gas occurs, it is a great boon to the open-hearth steel industry, 
 because of its high calorific power and the cheapness with which 
 
METALLURGICAL FUELS 
 
 469 
 
 it may be obtained, and about 80 wrought-iron plants and 90 
 steel plants in America use it. It is drawn from the earth, and 
 has a calorific power of 970 to 1010 B.t.u. per cu. ft. (equivalent 
 
 FIG. 366.— MUSHROOM REVERSING VALVE. 
 
 - .Furnace Flue ChlmneT Flue Tuiuaae Flua 
 
 FIG. 367.— WATER-SEALED REVERSING VALVE. 
 
 to 225 to 250 calories per cu. ft. or 8600 to 9000 calories per cu. 
 meter).* In the Pittsburg district the amount of natural gas 
 
 ^ The calcuatioa of this relaUoa will be found on page 473. 
 
470 
 
 THE METALLURGY OF IRON AND STEEL 
 
 used per ton of steel made ranges from 4000 to 11,000 cu. ft., 
 averaging about 5500. This includes the gas used in the 
 furnace and the small amount necessary for heating the ladles. 
 The natural gas is usually introduced through two pipes at each 
 end of the furnace, which are located close to the bottom of the 
 gas-ports and deliver the fuel about 3 1 /2 feet back from the hearth. 
 The natural gas is therefore applied without any loss in gas 
 mains, reversing valves, regenerators, etc., and is cheaper in 
 labor on the furnace floor itself. The Pittsburg natural gas 
 averages about 70 to 90 per cent, of methane (CH^) or marsh 
 gas, the remainder being hydrogen with a fraction of a per 
 cent, of carbon dioxide and a very few per cent, of nitrogen. 
 
 Detail of Washer 
 
 'BiibbatHoea / 
 CoanectatHon 
 
 FIG. 368— OIL BLOW-PIPE FOR OPEN-HEARTH FURNACE. 
 
 On account of its high calorific power, it is not necessary to pre- 
 heat this gas, and this is the more fortunate, because the amount 
 of hydrocarbon is so great that if the gas be passed through 
 regenerative chambers, it decomposes and deposits soot on the 
 checker-work. All the open-hearth furnaces in the Pittsburg 
 district are so arranged that they can be put on producer gas in 
 case it should become necessary, because the supply of natural 
 gas has been decreasing for many years. In the early part of 
 1908, some furnace plants of a very important company in this 
 district began the use of the manufactured gas, because of the 
 increased cost of natural gas. 
 
 Oil. — On the continent of Europe, and in parts of the United 
 States distant from the natural gas and bituminous coal regions. 
 
METALLURGICAL FUELS 
 
 471 
 
 niany open-hearth furnaces are heated by petroleum. The 
 United States has many deposits of fuel oil, besides which it is 
 sometimes possible to obtain a refuse from the oil refineries which 
 is excellent for this purpose. There are therefore many different 
 grades employed, but they will usually average from 7.8 to 8.3 lbs. 
 per gallon, with a calorific power of 14,000 to 17,000 B.t.u. per 
 pound. H. H. Campbell^ states that a rough comparison may be 
 made by assuming that 50 gal. of oil will give the same amount 
 of heat as about 1000 lbs. of soft coal, and he has had a valuable 
 amount of experience with this kind of fuel. It would seem, 
 
 Elevation. 
 
 D Q 
 
 Fofeplate Level 
 
 Charg ing Floor 
 
 FIG. 369.— FURNACE ARRANGED TO USE OIL BLOW-PIPE. 
 
 however, as if this value for oil was somewhat high for safety in 
 making calculations, and that a more conservative estimate 
 would be to say that from 35 to 60 gal. of oil would be required 
 per ton of steel treated in the open-hearth furnace. Eight 
 furnaces using oil averaged 38 to 42 gal. per ton of steel made. 
 The crude petroleum is vaporized by atomizing it with a jet of 
 steam or compressed air, and it is not common practice in the 
 United States to pass this vapor through a regenerative chamber. 
 The usual method of application is by a blow-pipe introduced 
 through the brickwork at the end of the furnace, as shown in 
 Fig. 369, and special forms of blow-pipe are now on the market for 
 this purpose. It is necessary to pump the oil to the blow-pipe or 
 
 * See page 247 of No. 2. Old edition. 
 
472 THE METALLURGY OF IRON AND STEEL 
 
 else to store it in an overhead tank, from which gravity will cariy 
 it, but the labor in connection with this is much less than the 
 labor on gas producers. In the rare cases when the oil vapor 
 is preheated, it must be introduced into the hot part of the 
 regenerative chamber, because if it cools it condenses. More- 
 over, when introduced into a cold furnace or. with a cold charge 
 in the furnace, combustion will be retarded. 
 
 Whether oil is used or not will depend principally upon freight, 
 because it may be transported much more cheaply than any other 
 form of fuel. It gives a longer flame than either natural or pro- 
 ducer gas, and one very great advantage of using it is a saving of 
 the roof of the furnace; the oil flame may be directed so accurately 
 by means of the blow-pipe that it does not impinge directly on the 
 roof, and the brickwork therefore lasts very much longer. On 
 the other hand, it spreads out horizontally and causes a greater 
 wear on the front and back walls of the furnace. It gives a 
 more uniform beat, a more oxidizing flame, and no danger of 
 losses or difficulties, in case there is a leak in the walls between 
 the gas and air regenerators, which is not an infrequent occur- 
 rence with gas. 
 
 Water-gas. — If steam be made to pass through a bed of red- 
 hot carbon, the product is a gas containing slightly less than 50 
 per cent, each of hydrogen and carbon monoxide and having a 
 high calorific power: 
 
 C + H20=C0 + H2 (absorbs 28,900 calories). 
 
 The result of this reaction is a reduction of the temperature of 
 the fuel bed, which is rapidly reduced until another reaction 
 begins to take place: 
 
 C + 2H20=C02 + 4H (absorbs 18,920 calories). 
 
 Therefore some means must be employed to r^ise the tempera- 
 ture at intervals, and this is ordinarily accomplished by inter- 
 rupting the passage of steam and passing air through the fuel bed, 
 which raises the temperature and at the same time forms pro- 
 ducer gas which is used for other purposes. Usually, it is neces- 
 sary to blow air through for 12 to 15 minutes, and then steam 
 for 4 or 5 minutes. Consequently, the manufacture of water- 
 gas is intermittent and the method is not very satisfactory 
 for open-hearth work, although the gas is used to some extent 
 on account of its high calorific power. This operation produces 
 roughly 35,000 cu. ft. of water-gas and 80,000 cu. ft. of producer 
 
METALLURGICAL FUELS 
 
 473 
 
 gas per ton of coal. Water-gas has about 2600 cals. per cubic 
 meter, or say, 290 to 300 B.t.u; per cubic foot.^ 
 
 Dellwik-Fleischer System, — The Dellwik-Fleischer system is a 
 modification of the ordinary water-gas system, in that the amount 
 of air blown through for heating up the fuel is so very large that 
 carbon dioxide is produced instead of carbon monoxide. This 
 
 Air 
 
 Mzm>:M. 
 
 '^^W/My0^JWMWMM ' ' 
 
 FIG. 370.— WAT^R-GAS PRODUCER. 
 
 gas is therefore altogether wasted, but the formation of carbon 
 dioxide generates so much more heat that the blowing up does 
 not take so long, and usually lasts from 1 1/2 to 2 minutes, after 
 which water-gas is made for 8 to 12 minutes. In this way about 
 80 per cent, of the calorific value of the fuel is converted into 
 
 ' One cubic meter =35.3 cubic feet. 
 One calorie = 3 . 968 B. t. u. 
 
 One calorie per cubic meter =8.9 B. t. u. per cubic foot 
 One calorie per gram = 1 . 8 B. t. u. per lb. 
 
474 THE METALLURGY OF IRON AND STEEL 
 
 water-gaS; and several steel-works in Europe have adopted the 
 method. 
 
 Mond Gas, — In the Mond process a mixture of water-gas and 
 producer gas is made continuously. For every ton of fuel burned 
 there is forced into the producer about 3 tons of air and 2.5 tons 
 of steam, the latter being produced by absorbing the sensible heat 
 of the gas in the boiler. It gives about 150,000 cu. ft. of gas per 
 ton of fuel burned, containing about 25 per cent, of hydrogen, 
 12 1/2 per cent, of CO, 45 to 50 per cent, of nitrogen, and 12 1/2 
 per cent, of CO2. This gives a slightly higher calorific power 
 than ordinary producer gas, and is used in some steel-works. 
 
XX 
 
 CHEMISTRY AND PHYSICS INTRODUCTORY TO 
 METALLURGY 
 
 Chemical Changes. — If a piece of coal be burned, it ceases to 
 exist as such. It disappears from sight, except for its slight resi- 
 due of ash, and apparently has been wiped out of existence for- 
 ever. Likewise, if a'piece of steel be attacked by some acid, it 
 disappears as such, and only a coloration of the acid gives evi- 
 dence to the eye of the metal previously present. Lastly, if a 
 piece of bright iron or steel be exposed to the weather, it is soon 
 converted into reddish-brown rust, which bears but little resem- 
 blance to the original metal. In the first example the solid coal 
 has been combined with oxygen of the air and converted to the 
 form of an invisible gas; in the second case, the iron has been 
 combined with the acid and water and converted into a liquid; 
 in the third case the iron has been combined with oxygen and 
 converted into a powder. In no case has there been any loss in 
 total amounts or weights, but only a difference in composition 
 or substance. These changes in composition are chemical 
 changes. 
 
 Physical Qhanges. — Changes may occur in form or properties 
 without any change in composition; For instance, water may be 
 converted into ice by mere cooling, and no change in composition 
 will take place. Or it may be converted into steam by heating 
 and will' still be composed of the same elemental constituents as 
 when it was in the form of ice or water. Iron may be liquid or 
 solid; it may be cold or hot; it may be magnetic or non-magnetic, 
 and all without change in substance. These changes in proper- 
 ties are known as physical changes. They may consist of changes 
 in form, strength, heat, light, magnetism, electricity — in fact, 
 everything but composition. 
 
 Relation Between Chemical and Physical Changes. — Every 
 chemical change produces one or more physical changes. Thus, 
 chemical changes are always accompanied by a loss or gain of 
 heat, and in the examples cited in paragraph 1, there were also 
 observed changes in from, in color, etc. Conversely, we shall 
 
 475 
 
476 THE METALLURGY OF IRON AND STEEL 
 
 see too that physical changes are often the cause of starting chem- 
 ical changes: heat is necessary to start the chemical action of the 
 burning of fuel; pressure produces the explosion of dynamite; 
 electricity breaks up many chemical compounds; light pro- 
 duces the chemical changes that make the photograph. 
 
 Chemical Compounds and Mechanical Mixtures, — When coal 
 is burned it is chemically unitedto the oxygen of the air, and the 
 gas formed contains properties entirely different from anything 
 belonging to either of the substances that compose it. Likewise 
 when steel is dissolved in an acid, or is converted to rust. This 
 is the essential characteristic of chemical action: that the sub- 
 stances acting upon each other lose their individual properties 
 and produce a new substance with different properties. . Not so 
 when the substances are merely mixed together, no matter how 
 intimate the mixture may be. If finely ground sulphur be 
 mixed with finely ground iron' no new properties are produced, 
 but if the mixture be heated until a chemical action takes place 
 and the two substances unite, a new compound is formed, with 
 new and different properties. Moreover, once the chemical union 
 has taken place the iron and sulphur cannot be separated except 
 by chemical means, whereas the mixture of the two could be sep- 
 arated by mechanical means, such as blowing the sulphur away 
 with a slow blast of air, or picking up the particles of iron with a 
 magnet. 
 
 Chemical Affinity. — The power that causes substances to unite, 
 and that holds them together afterward is known as "chem- 
 ical affinity." Like gravity, magnetism and some other great 
 forces, its nature is not understood but its influence is very evi- 
 dent. Some substances have seemingly no affinity for each 
 other (for instance, mercury and iron will not form a compound), 
 while others have tremendous affinity — such as sodium and 
 oxygen, which cannot be brought into each others' presence with- 
 out uniting violently and generating much heat. Some sub- 
 stances have almost universal affinities, like oxygen which unites 
 with every other elemental substance known except one, while 
 others are relatively inert and form few compounds. 
 
 Conditions under which Chemical Action will Occur. — To start 
 chemical action it is sometimes only necessary to mix the sub- 
 stances, as, for instance, sodium and water; in other cases we must 
 apply heat, or pressure, or electricity, etc., even though the sub- 
 stances have great affinity for each other and unite vigorously 
 
CHEMISTRY AND PHYSICS 477 
 
 lafter the action is once started. In other cases chemical action 
 is very slow, no matter how started as, for instance, the rusting 
 of iron, etc. As a general thing an increase in temperature 
 increases all chemical affinites, up to a certain point, but it in- 
 creases some faster than others. 
 
 The Elements, — In the universe there are millions of chemical 
 compounds, and these are mixed together to produce animal and 
 plant forms, the earth, sea, etc. All these compounds are differ- 
 ent combinations of only about eighty elemental substances, 
 which are known as "the elements." The compounds can all 
 be separated into their component parts by chemical means 
 (perhaps aided by elebtricity) , but the elements have so far re- 
 sisted every attempt to break them down into simpler substances, 
 and they are therefore considered as the simple substances and 
 the basis of all matter. These eighty elemental substances are 
 therefore of great importance, but some much more so than 
 others, for only eleven of them form the great bulk of the earth as 
 we know it, and the remainder are less abundant. The crust of 
 the earth is made up of the following elements: 
 
 Oxygen 47 . 29 per cent. 
 
 Silicon 27 . 21 per cent. 
 
 Aluminum 7 . 81 per cent. 
 
 Iron 5 . 46 per cent. 
 
 Calcium 3 . 77 per cent. 
 
 ^ Magnesium 2 . 68 per cent. 
 
 Sodium 2 . 36 per cent. 
 
 Potassium 2 . 40 per cent. 
 
 All others 1 . 02 per cent. 
 
 The air is 77 per cent, nitrogen and 23 per cent, oxygen. 
 Pure water is 89 per cent, oxygen and 11 per cent, hydrogen. 
 Plant and animal forms are composed chiefly of combination of 
 carbon, hydrogen, oxygen and nitrogen. 
 
 The air is the only one of the foregoing bodies in which the 
 elements are not severally united in the form of compounds of a 
 more or less complicated nature, and it is because the oxygen of 
 the air is in the free, or elemental, state that it is capable of per- 
 forming the chemical work of supporting life and burning fuels. 
 
 Summary. — We have now found that the whole universe, 
 so far as we know it, is built up of only about eighty different 
 simple substances, which we call elements, and which are 
 sometimes mixed together and sometimes chemically united. 
 
478 THE METALLURGY OF IRON AND STEEL 
 
 forming many millions of different combinations. We have also 
 learned that a chemical compound has properties different from 
 those of any of the substances composing it, and that the ele- 
 ments of the compound are held together by what is known as 
 " chemical affinity." We have learned that chemical action does 
 not always take place when two or more substances are mixed 
 together, but we have often to start it by heat, electricity or 
 some others means. Lastly we have learned that all chemical 
 action is accompanied either by a production or consumption 
 of heat. 
 
 Synthesis. — When two or more elements are chemically united 
 to form a compound, or when two or more compounds are chem- 
 ically united to form a further compound, the building-up process 
 is called a "synthesis." 
 
 Analysis. — On the other hand, when a compound is separated 
 into the elements which compose it, the breaking-down process 
 is called a "decomposition." 
 
 Definition of Metallurgy. — Metallurgy is the art of extracting 
 metals from their ores and adapting them to their intended service. 
 As iron occurs in the earth combined with one or more other 
 elements, the process of extracting it consists in decomposing the 
 compounds and obtaining the metal from them. But this is not 
 all of the metallurgy of iron for, after it is extracted, it must be 
 adapted to the service for which it is intended, and for this pur- 
 pose various other elements are combined with it in proportions 
 depending upon the uses to which the metal is to be put. Even 
 one part of some of the added elements in ten thousand parts 
 of iron will make a great difference in its properties, so that 
 these syntheses must be performed with great care. Then too 
 the physical treatment which metals are given affects some of 
 them greatly. 
 
 Qualitative and Quantitative Chemical Analysis. — Chemical 
 analysis is important in metallurgy in another connection, because 
 every substance put into the furnaces for smelting, and every 
 substance produced must be carefully "analyzed" in order that 
 we may know exactly what is in them and guide our operations 
 accordingly. A "qualitative analysis" will show what elements 
 are in any substance, while a "quantitative analysis" will show 
 how much of each is present. 
 
CHEMISTRY AND PHYSICS 479 
 
 Oxygen 
 
 Occurrence, — Oxygen occurs free in the air; combined with 
 hydrogen in water, and combined with silicon, with metals and 
 with many other elements in the earth's crust. It is the most 
 abundant element known to us. 
 
 Uses. — Oxygen reacts chemically with coal, coke, oil, gas and 
 other fuels to produce the heat for our fires, among which we 
 must include prominently the fires so necessary in metallurgy. 
 A familiar, if somewhat special, reaction due to oxygen may 
 .bring this home to us. When breathed into our lungs oxygen 
 in the air performs certain chemical reactions which produce the 
 heat that keeps us alive, and which purify the blood. 
 
 Preparation. — Oxygen is very easily prepared in a concen- 
 trated form in a great variety of ways. If, by means of combined 
 pressure and cold, air be converted into a liquid, its two com- 
 ponents may be separated by centrifugal force, or else the nitrogen 
 may be allowed to evaporate, leaving the liquid oxygen behind. 
 No chemical processes are necessary for this separation because 
 the elements are not conibined. Where a compound exists, 
 other means must be employed. For example, water may be 
 decomposed into oxygen and hydrogen by an electric current, 
 and from 100 parts by weight of water we can alwuys obtain 89 
 parts of oxygen and 11 parts or hydrogen, no weight being 
 lost in the change. 
 
 Chemical Action. — Oxygen is an odorless, colorless, tasteless 
 gas. At the ordinary temperature it forms few chemical reac- 
 tions, but, when heated, is one of the most active of the elements, 
 vigorously attacking, for example, hydrogen and carbon, as well 
 as their joint compounds in the form of gases, when once a 
 chemical reaction is started with a match, electric spark or 
 similar means. It also attacks almost all the metals at a red 
 heat, and some of them at lower temperatures — iron for instance, 
 which is coated with an " oxide" upon being heated about twice as 
 hot as boiling water. All the simple compounds of the elements 
 with oxygen go under the name of " oxides," and their formation 
 is'^accompanied'with the production of heat. 
 
 Phlogiston Theory. — Centuries ago it was observed that lead, 
 when melted and exposed to the air, became an apparently new 
 substance, and at the same time gained in weight. Starting with 
 the same weight of lead and allowing the action to go on until 
 
480 THE METALLURGY OF IRON AND STEEL 
 
 complete, always resulted in the same gain in weight. The an- 
 cients believed that this action was the transfer from the fire to 
 the metal of a certain indefinable substance which they called 
 ''phlogiston/' They learned that if the lead and "phlogiston" 
 were later heated with charcoal, metallic lead was again produced, 
 and they said that the "phlogiston" was driven out of it. We 
 now know that it was oxygen from the air that attacked the 
 lead when melted and formed "lead oxide," and that when the 
 lead oxide was heated with charcoal (carbon) the charcoal robbed 
 the lead of its oxygen and formed "carbonic oxide" leaving 
 metallic lead again. 
 
 Oxidation and Rednction. — When oxygen attacks an element 
 and forms an oxide the process is said to be an " oxidation," and 
 when an oxide is deprived of its oxygen and reduced to a metal 
 the process is said to be a "reduction." Iron is "reduceil" from 
 its ores, which are usually oxides. Oxidation and reduction are 
 therefore opposite actions in chemistry, the first adding some- 
 thing to a substance and the second taking something away. At 
 first the terms were used in connection with oxygen alone, but 
 are now applied to adding or taking away anything. In metal- 
 lurgy oxidation and reduction are all-important, for everything 
 that is reduced goes with the metals, and everything that is 
 oxidized passes away with the impurities. In the example cited 
 where melted lead was oxidized, the oxide separated itself from 
 the metal just as fast as formed, floating upon the top of it, and 
 when the reduction with charcoal was effected, the lead separated 
 itself from the mass and dropped down into the bottom of the 
 furnace as fast as it was reduced. This latter was then a metal- 
 lurgical operation. 
 
 Combustion. — Combustion is a form of oxidation, in which a 
 "combustible" is chemically united with oxygen, and, as we 
 know, this combustion, or burning, is our chief means of obtaining 
 heat. When there is just the right amount of both oxygen and 
 combustible to enter into combination, we have "perfect com- 
 bustion," but if the compound is formed and there is either 
 oxygen or combustible left over, we have "incomplete combus- 
 tion." Incomplete combustion always means waste of heat. 
 Thus, we know that too much air passed through a fire-bed will 
 carry waste heat up the chimney, or if we have too little oxygen 
 and carry a combustible gas up the chimney, or leave unburned 
 fuel on the grate, we again waste heat. 
 
CHEMISTRY AND PHYSICS 481 
 
 Therm 0-CHEMisTRY 
 
 Chemical Energy. — If two or more substances combine and 
 produce heat, they have chemical energy, and they transform 
 this chemical energy into hdat energy, which can be transformed 
 in turn into other forms and into work. Not all chemical actions 
 produce heat, but some are accompanied by a consumption of 
 heat, and therefore use up energy, or rather, they transform, 
 energy into chemical work. But the heat energy so transformed 
 into chemical work is not lost, for we can get it back again by 
 reversing the action. For instance, if lead oxide be reduced, a 
 consumption of heat occurs, and, the same amount of heat will be 
 produced if we burn the lead and produce the oxide again. So 
 it is in every case: The heat produced by the formation of a com- 
 pound is the same in amount as the heat consumed when the 
 compound is broken up, and the heat produced by any chemical 
 reaction is the same in amount as that consumed when the action 
 is reversed. The science that treats of the heat changes accom- 
 panying chemical changes is called "thermo-chemistry." As a 
 general thing syntheses and oxidations are accompanied by a 
 production of heat, while decompositions and reductions are 
 accompanied by an absorption of heat. Thermo-chemistry is 
 very important in metallurgy because metallurgy is chemistry 
 carried on at high temperatures, and the metallurgist must know 
 how much heat is required for all his reactions, and by what 
 reactions he may obtain it. 
 
 Maximum Affinity,— li iron is heated in air the oxide of iron 
 is formed, and if this be mixed with powdered aluminum and the 
 action started with a fuse, the aluminum will rob the iron of its 
 oxygen and unite with it instead. The reason for this is that 
 aluminum has a greater affinity for oxygen than iron has. This 
 process of selection is a common one in chemistry and any sub- 
 stance will decompose a compound provided it can form a new 
 compound with greater chemical affinities. Likewise, if we have 
 a limited amount of a substance in the presence of two others, it 
 will combine with the one for which it has the greatest affinity. 
 For example, if we have liquid iron and aluminum in the presence 
 of oxygen, none "of the iron will be oxidized until all the aluminum 
 has been oxidized. 
 
 Net Heat of Chemical Reactions. — When iron oxide is formed, 
 195,600 units of heat are evolved; when aluminum oxide is formed 
 
 31 
 
482 THE METALLURGY OF IRON AND STEEL 
 
 392,600 units of heat are evolved. Therefore, when aluminum 
 decomposes iron oxide and forms aluminum oxide instead, the 
 net heat effect of the reaction is to evolve (392,600-195,600=) 
 197,000 units of heat. But suppose a reaction occurs in which 
 the decomposition brought about consumes more heat than the 
 compound formed produces? For example, if iron oxide is 
 attacked by carbon and deprived of its oxygen there will be a net 
 loss of heat, instead of a gain, because carbonic oxide generates 
 only 29,160 heat units in its formation, while 195,600 heat units 
 are consumed in the decomposition of iron oxide. Such a 
 reaction would not go on unless we constantly supplied heat to 
 the bodies. This is important in metallurgy because it means 
 that when we smelt iron oxide with coke (carbon) we cannot 
 reduce the iron unless we continually heat the bodies. In the 
 case where aluminum reduced the iron it was only necessary to 
 start the reaction, but with carbon smelting it is not only neces- 
 sary to start the action with heat, but also to continually supply 
 the (195,600-29,160=) 166,440 units of heat that are absorbed. 
 
 Temperatures. — Temperature is the degree of heat. There are 
 two scales by which it is commonly measured, known respec- 
 tively as the Fahrenheit and the Centigrade scale. In both of 
 these the freezing- and boiling-points of water are taken as the 
 standards. In the Fahrenheit scale, 32° is the freezing-point of 
 water, and 212° is the boiling-point. Each degree is therefore 
 1/180 of this interval. In the centigrade scale 0° is the freezing- 
 point and 100° is the boiling-point of water. Each degree is 
 therefore i/100 of this interval. One degree Centigrade equals 
 1 4/5° Fahrenheit. A table of comparison is shown in Table 
 XXXVIII, page 504. 
 
 Heat Units. — The amount of heat in a body is different from 
 its temperature; it takes much more heat to raise a pound of 
 water to 200° F. than it does to raise a pound of iron, and more 
 to raise iron than copper, lead, or gold. There are two standard^ 
 by which amounts of heat are measured: A British Thermal Unit 
 (known as B. t. u.) is the amount of heat required to raise one 
 pound of water one degree Fahrenheit; a calorie is the amount 
 of heat required to raise one gram of water one degree Centigrade. 
 In both cases the water must start at its maximum density, which 
 is at 39.1° F. (=4° C). One B. t. u. equals 252 calories. ^ In 
 
 * One pound avoirdupois =453.59 grams; one ounce avoirdupois =23.3495 grams. 
 
CHEMISTRY AND PHYSICS 483 
 
 this book I shall generally use calories, as that is the ordinary- 
 system in scientific work. One Calorie =1,000 calories. 
 
 Summary. — We have now learned that iron and steel metal- 
 lurgy is chemistry at high temperatures, and that when iron is 
 reduced from its ores, we must decompose the ores, which con- 
 sumes a great deal of heat. We have also learned how heat is 
 obtained from chemical reactions, and chiefly from combustion. 
 We have learned that oxygen forms oxides with many of the 
 elements, and that some of the elements have a greater afiinity 
 *for oxygen than others, so that they will keep, or eVen take, the 
 oxygen away from them. Lastly we have learned that any 
 reduced substances will join with the metal in our furnaces, and 
 any oxidized ones will join with the impurities, and that the 
 reduced substances will not ordinarily mix with the oxidized ones, 
 except in minute quantities. 
 
 Chemical Equations 
 
 Combining Weights. — When elements unite in compounds they 
 always do so in certain definite proportions. Compounds of 
 oxygen and iron contain oxygen in multiples of 16 and iron in 
 multiples of 56 parts; those of oxygen and calcium contain 
 oxygen in multiples of 16 and calcium in multiples of 40. In 
 brief, iron enters compounds in multiples of 56, oxygen in mul- 
 tiples of 16,- calcium in multiples of 40. Sulphur enters in 
 multiples of 32, carbon in multiples of 12, etc. These character- 
 istic combining weights are known as "atomic weights" for a 
 reason that will be evident shortly. A list of about one-half 
 of the known elements with their atomic weights is given in the 
 table on the next page, those which are of the least importance 
 in metallurgy of iron and steel being omitted, and those which 
 are of greatest importance being printed in small capitals. 
 
 The Atomic Theory. — The atomic theory supposes that all of 
 the elements are made up of a myriad of tiny particles called 
 atoms. The atoms are the smallest particles of matter that can 
 exist; too small to be even conceived of by the imagination, and 
 yet all having a definite size and weight and incapable of being 
 divided into finer particles. All the atoms in any one element 
 are alike in composition, size, and weight, but differ in these three 
 properties from the atoms of all of the other elements. When 
 two or more elements combine chemically the atoms of one are 
 
484 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXXVII 
 
 Aluminum Al 27 
 
 Antimony Sb 120 
 
 Arsenic As 75 
 
 Barium Ba 137.4 
 
 Bismuth Bi 208 
 
 Boron B 11 
 
 Calcium. Ca 40 
 
 Carbon C 12 
 
 Chlorine CI 35:5 
 
 Chromium Cr 52 
 
 Cobalt '. Co 59 
 
 Copper Cu 63.6 
 
 Fluorine F 19 
 
 Gold Au 197 
 
 Hydrogen H 1 
 
 Iodine I 127 
 
 Iron Fe 56 
 
 Lead Pb 207 
 
 These combining weights are used in the laboratories of chemical analysis, 
 
 in calculating furnace burdens and similar work, A list should be kept for 
 convenient reference. 
 
 Magnesium Mg 24 
 
 Manganese Mn 55 
 
 Molybdenum Mo 96 
 
 Nickel Ni 59 
 
 Nitrogen N 14 
 
 Oxygen O 16 
 
 Phosphorus P 31 
 
 Potassium K 39 
 
 Silicon Si 28. 
 
 Silver Ag 108 
 
 Sodium Na 23 
 
 Sulphur S 32 
 
 Tin Sn 118. 
 
 Titanium Ti 48 
 
 Tungsten Wo 184 
 
 Vanadium V 51 
 
 Zinc... Zn 65. 
 
 locked with bonds of chemical affinity to the atoms of the others, 
 and thus the weight of each element entering the compound is in 
 proportion to the number and weight of its atoms. 
 
 Chemical Symbols. — In writing the elements it is customary 
 to represent each by one or two initial letters, instead of writing 
 the name out in full. This is a sort of shorthand of chemistry. 
 The representative letters are taken from th^ Latin name of the 
 elements, and the same symbols are employed in every civilized 
 country of the world. The symbol for iron is Fe, because the 
 Latin name for iron is ferrum. That for ali^minum is Al; for 
 calcium, Ca; for carbon, C; for hydrogen, H; for magnesium, Mg; 
 for manganese, Mn; for oxygen, 0; for phosphorus, P; for silicon, 
 Si; and for sulphur, S. These eleven symbols should be learned 
 as they will be used frequently in the body of the work. Others 
 are shown in Table XXXIII. 
 
 Multiple Proportions of Atomic Weights. — It has been found 
 that some of the elements can combine in different numerical 
 ratio in some compounds from that they do in others. For 
 instance, one atom of carbon may combine with one atom of 
 oxygen, and again one atom of carbon may combine with two 
 atoms of oxygen. In the first compound there will be 12 
 
CHEMISTRY AND PHYSICS 485 
 
 weights of carbon and 16 of oxygen; in the second, 12 weights of 
 carbon and 32 of oxygen. Likewise, one atom of iron may com- 
 bine with one atom of oxygen, or two atoms of iron may combine 
 with three atoms of oxygen. Each of these compounds will have 
 different properties. It is possible to represent these compounds 
 in a very simple way by using the symbols for the elements, for 
 each symbol designates one atom of the element. To represent 
 the first compound of carbon and oxygen we write their symbolic 
 letters together — thus, CO. To represent the compound con- 
 taining one atom of carbon and two of oxygen we write, COO or, 
 COj. To represent the first iron oxide we write — ^FeO. To 
 represent the second one, FojOg. Then the formulae for these 
 compounds tell iis not only what elements make up the com- 
 pound, but also how much of each is present. For example, in the 
 first iron oxide we have. 56 parts of iron and 16 parts of oxygen; 
 in the second we have 112 parts of iron and 48 of oxygen. 
 
 Molecules. — When two or more atoms are held together by 
 chemical affinity the particle formed is known as a molecule. The 
 symbol, CO, represents a molecule of carbonic oxide;, FcjOg rep- 
 resents a molecule of iron oxide. 
 
 Chemical Equations. — Chemical shorthand may be used to 
 represent chemical reactions, and will indicate at a glance what 
 is taking place. A synthesis will be shown as follows: 
 
 C + O = CO; 
 
 Carbon and oxygen produce carbonic oxide. 
 
 The decomposition of water would T^e written: 
 
 H^O = 2H + 0; 
 
 Water gives hydrogen and oxygen. 
 
 The reduction of iron oxide by carbon: 
 
 FeO + C = Fe + CO. 
 
 Iron oxide and carbon give iron and carbonic oxide. 
 
 The reduction of iron oxide by aluminum: 
 
 Fefi, + 2A1 = 2Fe + Al^Og; 
 
 Iron oxide and aluminum give iron and aluminum oxide. 
 
 Indestrv^tihility of Matter. — In the equations written above 
 it will be noticed that there are always as many atoms of each 
 element on the left-hand side of the equation mark as on the right. 
 This is in accordance with the fundamental law of chemistry that 
 matter can neither be destroyed nor created. If we combine car- 
 bon with oxygen the weight of carbonic oxide formed is exactly 
 equal to the weight of carbon and oxygen together. Also, if we 
 
486 THE METALLURGY OF IRON AND STEEL 
 
 decompose water the total weight of hydrogen and oxygen will 
 be equal to the weight of water from which it came. 
 
 Hydrogen 
 
 Occurrence. — Hydrogen forms 11 per cent, of water, which is 
 a compound of hydrogen and oxygen, whose molecules contain 
 two atoms of hydrogen and one of oxygen, so that they have the 
 formula, HgO. Hydrogen also occurs in all living forms. 
 
 Properties. — Hydrogen is a colorless, tasteless, odorless gas, 
 and the lightest substance known, so that it would be very useful 
 for filling balloons except for its cost. It has a high chemical 
 affinity for oxygen and a stream of it when ignited will burn read-^ 
 ily in the air and produce water vapor with the evolution of 
 58,060 calories: 
 
 2H + = H20(+ 58,062 cals.). 
 
 It is therefore a good combustible, and an impure form of it is 
 indeed one of our important fuel gases, going under the name 
 of "water gas." It is also a good "reducing agent"; that is, it 
 will reduce substances by taking their oxygen away. 
 
 Hydrocarbons. — Hydrogen has a strong affinity for carbon, 
 and forms with it a long series of compounds known as "hydro- 
 carbons," of which there are about two hundred different com- 
 binations. These form the basis of mineral oil, or petroleum, 
 from which we get kerosene, gasoline, naphtha, benzene, lubri- 
 cating oils, vaseline, paraffine, etc. The "light hydrocarbons" 
 are found in kerosene, gasoline, etc., while the heavy hydro- 
 carbons" are found in the less volatile oils. Some of the more 
 important compounds are as follows: Methane, whose molecule 
 has the formula, CH^, is the chief constituent of natural gas; 
 when it burns the following reaction takes place : 
 
 CH, + 40 = C02 + 2H20. 
 
 Ethylene, CgH^, is a heavier hydrocarbon than methane because 
 its molecule contains more atoms of carbon, while acetylene, 
 C2H2, is heavier still and is the most powerful illuminating gas 
 known. Benzene, CeHg, has the same relation between the atoms 
 of hydrogen and carbon as acetylene, but a different number of 
 them in the molecule, so that it is an entirely different substance 
 with different properties. 
 
CHEMISTRY AND PHYSICS 487 
 
 Thermo-chemistry of the Hydrocarbons. — When methane burns 
 we get the following reaction: 
 
 CH^ +40= 2H2O + CO2. ( + 191,270 cals.) 
 
 -22,250 cals.+ 116,320 cals. + 97,200 cals. 
 
 If we put the heat of combination under each of the compounds 
 then we can readily calculate the net heat produced or consumed 
 by the reaction, because all the compounds on the left of the 
 equation mark are decomposed and all those on the right are 
 formed. Therefore the sum of the heats on the left is to be com- 
 pared with the sum of the heats on the right. If the right-hand 
 sum is greater, heat is produced; if the left-hand sum is greater, 
 heat is destroyed. In the burning of methane 191,270 calories 
 are produced and we therefore place (+191,270 cals.) at the end 
 of the equation. 
 
 Let us now consider for comparison the reduction of tin oxide 
 by carbonic oxide, as follows: 
 
 SnOa + CO = SnO + 00^ (-2,560 cals.) 
 
 -141,300 cals. -29,160 +70,700 cals. + 97,200 cals. 
 
 In this case we find that the sum on the left is greater, and 2,560 
 calories are consumed. We therefore place ( — 2,560 cals.) 
 next to the equation. 
 
 Preparation of Hydrogen. — The cheapest method of obtaining 
 hydrogen is by decomposing water. This may be done as 
 follows : 
 
 Na + H^O^H + NaOH ( + 33,700 cals.) 
 (sodium) - 69,000^ + 102,700 
 
 We see that this is a heat-producing reaction. Another way: 
 C + H20 = 2H + CO (-28,900 cals.) 
 -58,060^ +29,160 
 
 We do this by passing water vapor over red-hot carbon, but the 
 reaction consumes heat so we must frequently heat the carbon 
 or the reaction will not go on. 
 
 Still another way is to pass an electric current through a body 
 of water. Hydrogen gas appears at one electric connection and 
 oxygen gas at the other. This process is known as the " electrol- 
 ysis'' of water, and it is an operation in " electro-chemistry." If 
 it is carried out perfectly the amount of electric energy passed 
 into the water will be equal to the amount of heat energy required 
 for the decomposition — that is, 69,000^ calories. 
 
 1 The heat of formation of water- is 69,000 cals. in liquid form and 58,060 in gaseous 
 form. For other heats of reactions, see page 124, No. 63. 
 
488 THE METALLURGY OF IRON AND STEEL 
 
 Summary. — Now we have learned that each element is con- 
 stituted of infinitesimal particles called atoms which in any one 
 element are all identical in weight and composition, and that 
 when elements form compounds the atoms of one are joined to 
 those of the other by bonds of chemical affinity. We have also 
 learned that atoms of elements are represented by letter symbols, 
 and that we can express reactions between them by putting the 
 symbols together in molecules and then showing how they break 
 up and change places to form other molecules, and that there are 
 no atoms and no weight lost or gained in any of these changes, 
 but there is a gain or loss in heat energy to correspond exactly 
 with each loss or gain of chemical energy. And we have seen 
 how the loss or gain of heat may be determined by reckoning the 
 total heat of compounds decomposed as heat lost, and the total 
 heat of compounds formed as heat produced. From this point 
 we shall go on to consider the important elements more in detail 
 as to their chemical behavior. 
 
 Elements, Compounds, and Radicals 
 
 Metallic and Non-metallic Elements. — All the metals are ele- 
 ments, but some of the elements are not metals; for instance, we 
 know without being told that oxygen is not a metal. The dis- 
 tinction between metallic and non-metallic elements is not very 
 clear. It once was considered that all elements which looked like 
 metals should be classified as such; they were said to have 
 "metallic luster." And all others were classified as non-metals. 
 But this classification has been shown to be deceiving, and a 
 chemical one has taken its place. Now all the elements that form 
 "bases" are classified as metals, and all those that form "acids" 
 are classified as non-metals. 
 
 Acids and Bases. — Acids have certain characteristic chemical 
 properties and bases have certain other characteristic properties. 
 Acids will destroy the characteristic properties of bases and 
 neutralize them, and conversely, bases will neutralize acids. 
 Acids and bases have strong affinity for each other and either one 
 will attack the other if opportunity offers. That is why a basic 
 slag will attack an acid furnace lining, or an acid slag will attack 
 a basic lining. Metallurgical acids are " anhydrous " (water-free) . 
 
 Salts. — When an acid and a base just neutralize each other 
 they form what is known as a salt. Let us dissolve 36.5 grains of 
 
CHEMISTRY AND PHYSICS 489 
 
 hydrochloric acid, HCl, in water, and put a piece of paper soaked 
 in litmus in it; the paper will at once turn a brilliant red. Now 
 let us dissolve 40 grains of caustic soda, NaOH, in another vessel. 
 Caustic soda is a strong base; if we put one end of our piece of red 
 litmus paper in the solution it will turn blue. Now let us pour 
 the acid solution into the basic solution, and we will get the 
 following reaction: 
 
 HCl + NaOH = NACl + H^O. 
 36.5 40 58.5 18 
 
 Now, 36.5 is the molecular weight of HCl (because one atom of 
 hydrogen weighs 1, and one of chlorine weighs 35.5), and 40 is the 
 molecular weight of NaOH (one atom of Na = 23; one of = 16, 
 and one of H = 1) ; therefore there must be as many molecules of 
 HCl present as of NaOH, and a complete neutralization will occur. 
 Moreover it will be seen that the total weight of atoms at the 
 right of the equation mark is the same as that at the left. This 
 neutralization forms "sodium chloride," which is our common 
 table salt. The salt will be dissolved in the water used in the 
 experiment. If now we put in this salt solution the piece of 
 litmus paper, one end of which is red and the other blue, it will 
 not change its colors at all. 
 
 Radicals. — Let us consider the neutralization of caustic soda 
 by sulphuric acid, HjSO^: 
 
 2NaOH + H^SO, = Na,SO, + 2H2O. 
 80 98 142 36 
 
 In this reaction the SO4 has changed places with the OH. When 
 two or more atoms are joined together and travel around in com- 
 pany in this way, acting as if they were inseparable, they are 
 called "radicals." In this reaction, the SO4 is called an "acid 
 radical," and the OH is called the "hydroxide radical." For 
 the time being these radicals act as if they were elementary 
 substances. 
 
 Valence, — Let us consider four compounds with hydrogen, 
 as follows: 
 
 CIH OH2 NH3 CH, 
 
 Hydrochloric acid, water, ammonia, methane. 
 
 One atom of hydrogen can hold one of chlorine, but it takes two 
 to hold one atom of oxygen, three to hold one of nitrogen, and 
 four to hold one of carbon. Conversely, one of chlorine can hold 
 one of hydrogen, one of oxygen can hold two of hydrogen, one of 
 
490 THE METALLURGY OF IRON AND STEEL 
 
 nitrogen, three, and one of carbon, four. This capacity for hold- 
 ing numbers of atoms is called "valence." In the compounds 
 shown above, chlorine is uni-valent, oxygen is bi-valent, nitrogen, 
 tri-valent, and carbon, quadri-valent. In each case hydrogen is 
 uni-valent; indeed hydrogen is established as the standard of 
 valency, with a holding power of oiie. We can determine the 
 valence of other elements by learning how many atoms of hy- 
 drogen they will hold, or, if they do not form a compound with 
 hydrogen, we can compare them with some other element that 
 does. For example, calcium forms a very common oxide, CaO, 
 known as lime. , In lime calcium holds one atom of oxygen; but 
 it takes two atoms of hydrogen to hold one atom of oxygen; 
 therefore calcium is bi-valent. 
 
 Chemical Stability. — We have already seen enough compounds 
 to know that the valence of several of the elements is not a 
 constant quantity. For example, carbon and hydrogen atoms 
 unite in nearly two hundred different combinations. Likewise, 
 iron forms FeO and FegOg. In the first it has a valence of two; 
 in the second, of three. But there is a difference in the stability 
 of these compounds; the oxide, FeO, can only exist under strong 
 reducing conditions, and will take on more oxygen with the least 
 opportunity. Iron forms two sulphides, designated as FeS and 
 FeSj/ In the second compound it has a valence of four, but this 
 sulphide is not as strong a one as the other, and the second atom 
 of sulphur may be driven off by heating it slightly: FeSg^ 
 FeS + S. 
 
 Ferrous and Ferric Compounds, — The oxide, FeO, is called 
 "ferrous oxide"; while FeS is called "ferrous sulphide." The ox- 
 ide, FegOg, is called "ferric oxide," while FeSj is called ^'ferric sul- 
 phide." Manganese forms two oxides: MnO is called "mangan- 
 ous oxide," and MnOa is called "manganic oxide." So with all 
 compounds; that having the lower valence is given the sufl&x 
 -ous, and that with the higher valence, -ic. 
 
 Mono-, Bi~j Tri-j etc. — ^FeS is also called "iron mono-sul- 
 phide," from the Latin, meaning one; FeSg is sometimes called 
 "iron bi-sulphide." MnO is called "manganese monoxide," and 
 MnOa, " manganese bi-oxide" (di-oxide is sometimes used instead 
 of bi-oxide). HjO is "hydrogen monoxide"; H^Oj is "hydrogen 
 di-oxide." FcjOg is called "iron sesqui-oxide," from the Latin 
 meaning three halves. 
 
 Oxidizing Agents. — When oxygen is added to a compound 
 
CHEMISTRY AND PHYSICS 491 
 
 it is said to be oxidized. For example, FeO will be oxidized to 
 FegOg (2 FeO + 0=Fe203). Oxidation can only be produced by 
 means of some "oxidizing agent.'' The commonest oxidizing 
 agent in metallurgy is the oxygen of the air, and the next most 
 important in iron and steel processes is FejOg, and slags very 
 rich in FeaOgi 
 
 3Si+2Fe203=3Si02+4Fe 
 6P + 5Fe2O3 = 3P2O6 + 10Fe 
 
 Another important one is carbon di-oxide: 
 
 Fe+C02=FeO+CO. 
 
 Reducing Agents. — When atoms are taken out of the mole- 
 cule of a compound it is said to be reduced. Reduction can only 
 go on in the presence of some "reducing agent." The common- 
 est reducing agent in metallurgy is carbon in the form of coke, 
 charcoal, etc. 
 
 Si02+2C-Si+2CO. 
 
 Another one is carbon monoxide, and another is hydrogen: 
 
 Fe203 + CO = 2FeO + C02. 
 
 Manganese is also a reducing agent for iron: 
 
 FeO + Mn=Fe + MnO. 
 
 Chemical Reactions and Compounds 
 
 Organic and Inorganic Chemistry. — The chemistry of living 
 organisms, such as plants, animals, etc., is a very complex sub- 
 ject, and need not be discussed here. Because carbon enters into 
 all organisms we may describe organic chemistry as the chemistry 
 of the carbon compounds. Inorganic chemistry is the chemistry 
 of the metals and of compounds in which carbon enters in 
 relatively small proportions. Inorganic chemistry is the only 
 one that concerns metallurgists especially. 
 
 Wet and Dry Chemistry. — In, the analytical laboratories they 
 perform their chemical reactions by dissolving everything in 
 water and so getting them in the liquid form, because solids do 
 not unite with each other rapidly, and gases are not easily con- 
 trolled. This branc'h of chemistry is known as "wet chemistry.'' 
 In iron and steel metallurgy, however, we get everything in 
 liquid form by melting it. This is known as "dry chemistry." 
 The reactions that take place in dry chemistry are the same in 
 
492 THE METALLURGY OF IRON AND STEEL 
 
 principle as those of wet chemistry. The chief difference is that 
 we bring substances into the intimate contact necessary by 
 fusion instead of by solution. 
 
 Carbon. — Carbon occurs in the earth in the crystallized form 
 as graphite and as diamonds. Of these the diamond is the purer 
 variety, but both may be considered as pure carbon in diiferent 
 forms. It occurs in naure in the impure form of coal, in quanti- 
 ties vastly greater and commercially of the highest importance. 
 The element may be obtained in a massive, or uncrystallized, 
 form by burning organic matter, such as wood, when a black 
 residue of carbon (charcoal) will be left. The most abundant 
 occurrence of carbon is, however, in combination with other ele- 
 ments in the various forms of living matter, and also in inorganic 
 compounds with metals, known as carbonates, such as the car- 
 bonate of lime, CaCOg, called limestone, or, when in the crystal- 
 lized form, marble. 
 
 When bituminous coal is burned with the proper deficiency 
 of air, a silvery-gray residue is left which is an impure form 
 of carbon, called coke. Crystals of graphite often are present 
 on the surface of coke. This coke is one of the most import- 
 ant of all metallurgical reducing agents, as well as fuels. Car- 
 bon also forms a number of hydro-carbons which are used in the 
 form of gases as reducing agents and fuels, because both their 
 carbon and hydrogen will unite with oxygen. 
 
 Carbon forms two oxides — CO and COj. The first combina- 
 tion is accompanied with the production of 29,160 calories, and 
 the second, 97,200 calories. The formation of CO is not complete 
 combustion, because it will itself be further oxidized: C0 + = 
 CO2, with the evolution of 68,040 calories. The heat of formation 
 of C -}- together with that of CO + is just equal to that of the 
 reaction: 
 
 C + 20 =C02( + 97,200 cals.); 
 29,160 + 68,040 = 97,200. 
 
 Carbon di-oxide, CO2, is an acid radical and unites with many 
 bases to form "carbonates," of which the commonest are those 
 of calcium, CaCOg, magnesium, MgCOg, and sodium, NajCOg. 
 As CO2 is very volatile, the carbonates may be decomposed by 
 heat, which drives off the CO3 as a gas: 
 
 Na2003 = Na,0 + C02. 
 
 Chemical Behavior of Iron, — Iron is attacked by many of 
 
CHEMISTRY AND PHYSICS 493 
 
 the wet acids — sulphuric, nitric, hydrochloric, acetic, etc. 
 It is attacked by oxygen when heated,. and even when cold pro- 
 vided the air is damp. At a red heat, iron decomposes water 
 vapor (2 Fe + 3 H20=Fe203-l-6H). It has a high affinity for 
 oxygen, and also for small amounts of carbon, silicon, sulphur, 
 phosphorus, and hydrogen. The last-named gas will penetrate 
 solid iron very readily at a red heat and form a compound with it. 
 Iron practically never occurs native on the earth's surface except 
 combined with oxygen or some other elements. 
 
 Chemical Behavior of Silicon, — Silicon has a high affinity for 
 oxygen, with which it forms a very common oxide, SiOj, which 
 is known as silica. This compound is decomposed with great 
 difficulty; the following reaction takes place only when we get 
 to the very high temperature in the hearth of the iron blast fur- 
 nace: 
 
 SiO2+2C=-Si4-2CO(-121,680 cals.) 
 -180,000 +2X29,160 = 58,320. 
 
 We must remember the difference between silicon and its oxide, 
 silica. Silicon never occurs uncombined in the earth, but silica 
 is the most abundant constituent known to us. Quartz is a crys- 
 tallized form of pure silica, while flint, jasper, agate, etc., are 
 uncrystallized forms. Opal is silica combined with water. 
 
 Silicates. — Silica is the great acid of dry chemistry, and when 
 in the melted condition will neutralize every base with which it 
 comes in contact, forming a series of salts known as " silicates." 
 The great bulk of the earth's rocks are either pure silica or silicates 
 of the different metals, and all metallurgical slags are silicates. 
 The mono-silicate of iron has the formula FcgSiOi. But it is 
 more commonly written (FeO) aSiOg, which is the same as 
 FejOgSiOg. The series of commonest iron silicates are given 
 below: 
 
 FeO.SiO" Sub-silicate. 
 
 (FeO)2Si02 Mono-silicate. 
 (FeO)2(Si02)3 Sesqui-silicate. 
 
 FeO(Si02)2 Bi-silicate. 
 
 FeO(Si02)3 Tri-silicate. 
 
 With lime, CaO, and magnesia, MgO, a similar series is formed, 
 but the silicates of alumina, AlgOg, are more complicated in com- 
 position. The different metallic silicates have the property of 
 dissolving in each other when melted, and of dissolving the oxides 
 
494 THE METALLURGY OF IRON AND STEEL 
 
 of metals, and various other oxidized substances, but not of dis- 
 solving metals or reduced substances. 
 
 Feldspar. — With potassium and aluminum silica forms a series 
 of silicates known as the feldspars, which are common con- 
 stituents of the earth's crust. The feldspars are chiefly important 
 because when reduced to powdered form they become clay, which 
 has the peculiar property of becoming plastic when moistened. 
 The purer clays melt at a very high temperature and are therefore 
 used as the bond to hold together the material for the linings of 
 furnaces, but the clays that contain much potassium or sodium 
 melt relatively easily, and are not so " refractory." Clays contain 
 a certain amount of water of crystallization, that is, water chem- 
 ically combined with the molecule of the silcates. If they are 
 heated so hot that this water of crystallization is driven out of 
 them, they will not again become plastic. 
 
 Chemical Behavior of Aluminum. — Aluminum has great af- 
 finity for oxygen and is therefore used, like silicon, for the pur- 
 pose of de-oxidizing steel: 
 
 3FeO + 2A1 = 3Fe + AI2O3. 
 Indeed aluminum retains its oxygen more tenaciously than silicon, 
 and even the highest temperature of our fuel furnace does not 
 effect its reduction. The oxide of aluminum, AI2O3, is called alu- 
 mina, is a common constituent of rocks, and when nearly pure is 
 used as an ore of the metal, its reduction being effected in electric 
 furnaces. Alumina is very refractory, that js, it will stand a high 
 temperature without melting, and is neutral in character, that is, 
 it is attacked neither by acid nor basic slags. It is therefore used 
 as a neutral lining for some furnaces. 
 
 Chemical Behavior of Manganese. — Manganese has a higher 
 affinity for both oxygen and sulphur than iron has, and is there- 
 fore used as a de-oxidizer and de-sulphurizer ofiron and steel. If 
 sufficient manganese is present, and the metal bath kept liquid a 
 sufficient time, neither oxygen nor sulphur will be found com- 
 bined with iron: 
 
 FeO + Mn = MnO +Fe ; 
 FeS + Mn = MnS+Fe. 
 
 Chemical Behavior hf Sulphur. — Sulphur is found in the earth 
 native (that is, free from combination), especially in volcanic re- 
 gions, and also combined with metals as sulphides. Iron bi- 
 sulphide, called "iron pyrites," FeSg, is very abundant and is the 
 chief source of sulphuric acid manufacture, while the sulphides 
 
CHEMISTRY AND PHYSICS 495 
 
 of cppper, lead, and zinc are the principal commercial ores of 
 those metals. Iron sulphides are not used so much as ores on 
 account of the expense of ridding the iron of sulphur, which is 
 very harmful to it. Sulphur readily combines with oxygen at a 
 slightly elevated temperature to form SOg and SO3, and these com- 
 bine with water to form sulphurous and sulphuric acids (HjO H- 
 S02 = H2S03; and H20+S03 = H2SOJ. In wet chemistry sul- 
 phuric acid attacks metals to form sulphates: 
 
 2Fe + 3H2S04 = 6H+Fe2SgOi2 or Fe2(S04)3. 
 Ca + H2SO4 = 2H + CaSO^. 
 
 Phosphorus, — Phosphorus occurs in nature usually as metallic 
 phosphates, and chiefly as phosphate of lime, Ca2(P04) 2, a natural 
 mineral to which the name of apatite is given. It is in this form 
 that it ordinarily gets into the blast furnace with the iron ores 
 which it accompanies in the earth. Phosphate is necessary to 
 animal and vegetable life and a good part of bones and other hard 
 parts of living organisms are composed of it. The phosphates 
 that will dissolve easily are therefore valuable fertilizers. Con- 
 sequently certain slags which are used to remove the phosphorus 
 from steel can be sold for fertilizing purposes. 
 
 Phosphorus ajjts the part of an acid-forming element, and the 
 phosphate radical will form salts with many metallic oxides, but 
 especially with iron oxides, magnesium oxide and lime_, for the 
 latter of which it has great affinity. But it is a weaker acid 
 than silica, and silica wiU drive the phosphate radical away 
 from all the basic radicals until the silica has completely satis- 
 fied itself. For this reason phosphorus cannot be combined in 
 slags unless there is a superfluity of bases present over the amount 
 necessary to surfeit the silica. In our steel furnace slags this 
 means usually as least 40 per cent, of lime plus magnesia plus 
 iron oxide. 
 
 Calcium and Magnesium. — Calcium forms a very common 
 oxide, CaO, known as lime, and magnesia forms a similar one, 
 MgO; called magnesia. These occur in nature chiefly combined 
 with carbonic acid to form carbonates; CaCOg is called limestone, 
 and MgCOg, magnesite. The two carbonates often occur com- 
 bined together in a compound having the formula (CaMg)C03. 
 This type of formula is used to indicate that the calcium and 
 magnesium replace each other in the arbonate in almost any rela- 
 tive proportion. The natural rock, (Ca.Mg)C03, has the miner- 
 alogical name of " dolomite." 
 
496 THE METALLURGY OF IRON AND STEEL 
 
 Limestone is used as a material to add to the charge of the 
 iron blast furnace because the carbonic acid is driven off in the 
 upper levels of the furnace as soon as it begins to become hot 
 (CaCO3 + heat = Ca04-C02) and the lime so produced serves as a 
 base in the blast-furnace slag. Burnt limestone, that is, lime- 
 stone from which the carbonic acid has been driven off by heat, is 
 also added to the slags made in some of the steel furnaces, in 
 order to increase their basicity. Lime has the peculiarity of 
 absorbing moisture from the air and forming a hydrate [CaO + 
 H20 = Ca(OH)2]. This is known as "slacking," and it causes 
 the lime to lose its coherence. For this reason furnace linings 
 cannot be made of it. 
 
 Magnesia is made by burning magnesite (MgC03 4-heat = MgO 
 + CO2), and this is much used for making the basic linings of 
 furnaces. Burnt dolomite is used for patching basic furnace lin- 
 ings, but it is not as durable as magnesia for the original lining. 
 
 Chemical Solutions 
 
 Chemical Compoundsj Mechanical Mixtures, and Chemical 
 Solutions. — We have learned that the differences between me- 
 chanical mixtures and chemical compounds are: (1) The proper- 
 ties of compounds are different from those of its components; (2) 
 the formation of a compound is attended with the production or 
 absorption of heat; (3) the components of a compound are held to- 
 gether with bonds of chemical affinity, and (4) the components 
 always form the compound in the same definite proportions. 
 There is another class of combinations different from both com- 
 pounds and mixtures, and known as solutions. These have some 
 of the characteristics of compounds and also of mixtures. (1) 
 The properties of a solution are different from those of its com- 
 ponents, but not to as marked a degree as is the case with com- 
 pounds; (2) its formation is attended with the production or con- 
 sumption of heat; (3) the components of the solution are held to- 
 gether by bonds of chemical affinity, and can only be separated 
 by chemical means, or by electricity, but (4) unlike compounds, 
 and to a limited degree like mixtures, the the components of a solu- 
 tion may vary widely in proportions. In some cases the variation 
 is infinite, as with melted gold and silver, which will dissolve in 
 each other in any proportion; likewise, with melted copper and 
 
CHEMISTRY AND PHYSICS 497 
 
 silver. In other cases the limit of solubility is very narrow, as in 
 the case of melted iron, which will dissolve only about 5 per cent, 
 of carbon, while carbon will dissolve apparently only 1 per cent, 
 or so of iron. 
 
 Just as some substances resist all our efforts to make them 
 combine chemically, so others refuse to dissolve. For instance, 
 several salts and liquids will not dissolve in water, the best solvent 
 known, or rather, dissolve to such a slight degree that, for practi- 
 cal purposes, it may be neglected. The same is true of iron and 
 mercury, melted lead, and zinc, etc. 
 
 Essence of Solubility, — Just what the nature of the state of 
 solution is cannot be told at present, but the atoms, or molecules, 
 of the dissolved substance, known as the "solute," seem to be 
 held by the molecules of the solvent. One striking difference ex- 
 isting between a solution and a mixture is that the solute seems 
 to occupy no space. If we mix with hot water one-quarter of its 
 weight of table salt, the level of the water will rise in the contain- 
 ing vessel an amount equivalent to the bulk of the salt, but, as 
 soon as the water dissolves the salt, it will fall back to its original 
 volume. In short, we now have a quarter more weight of ma- 
 terial in the same bulk, so that the specific weight of the mass 
 increases 25 per cent. In several ways which we have not space 
 here to discuss we can know of the presence of a greater number 
 of molecules than ordinary in the same space when two or more 
 substances are dissolved in each other; for instance, "osmotic 
 pressure," surface tension, Metals dissolved in each other are 
 heavier than the same bulk of any of the metals alone. 
 
 Precipitation, — If we dissolve 27 per cent, of table salt in hot 
 water and then allow the water to cool, some of the salt will fall 
 out of solution again and crystallize. This action is called "pre- 
 cipitation. It is one of the most important actions in chemis- 
 try. We may cause a precipitation by another means : If we have 
 as much of any salt dissolved in water as it will take, and then 
 add to the solution a more soluble one, the water will dissolve the 
 new salt and precipitate the old one in corresponding amount. 
 The same applies in metallic solutions: If we have 5 per cent, of 
 carbon dissolved in iron and then add some metallic silicon, the 
 iron will precipitate graphite in flakes of " kish," and dissolve sili- 
 con. The commonest method of precipitating elements in wet 
 chemistry is by producing chemical change: Suppose we have 
 some table salt (NaCl) dissolved in water and add just enough 
 
 32 
 
498 THE METALLURGY OF IRON AND STEEL 
 
 silver nitrate (AgNOg) to react with it; we form silver chloride 
 which is insoluble, and which precipitates almost instantaneously: 
 
 NaCl + AgN03 = AgCI + NaN03. 
 This is only a partial precipitation, because sodium nitrate is still 
 left in solution, but is often of great service. 
 
 Solubility and Temperature. — As a general thing the higher the 
 temperature the greater amount of solute can be dissolved in any 
 given solvent. This rule is not universal, but usually applies 
 in practical metallurgical chemistry. Five per cent., or even 
 more, carbon will dissolve in iron at a high temperature, but some 
 of this precipitates as the metal cools near its solidification tem- 
 perature. 
 
 Alloys. — Metallic alloys are dependent upon solution for their 
 formation. Two metals which will not dissolve cannot be made 
 to form alloys, and, in practice, all alloys are made by dissolving 
 melted metals in each other. The only exception is certain al- 
 loys made by dissolving solid metals in each other under great 
 pressure. 
 
 Nature of Slags. — Slags are molten solutions, and, as a gen- 
 eral rule, they will dissolve all the oxidized substances with which 
 they come in contact in the furnace (except such as are 
 artificially kept too cold to be affected, such as the furnace walls) 
 and will precipitate all the reduced substances. For example, 
 metals will be precipitated as fast as reduced from their combina- 
 tions in the ores : phosphorus, if oxidized by being combined with 
 some base as a phosphate, will be dissolved, but, if silica takes 
 the base away from it, the reduced phosphorus will be precipi- 
 tated again. 
 
 Some Principles of Physics 
 
 Dalton and Gay Lussac Law. — Air, oxygen, nitrogen, hydrogen, 
 and a few other gases, are called permanent gases, and all these 
 expand and contract under change of temperature according to 
 the same law, which is, that they expand or contract 1/273 of 
 their volume at 0° C. for each 1° C. their temperature is raised or 
 lowered, provided the pressure upon them remains constant. 
 On the Fahrenheit scale, this coefficient is 1/493 instead of 1/273. 
 Thus air under a given pressure will have twice the volume at 
 273° C. that it has at 1° C. This is most easily explained and re- 
 membered in the form that the volume of permanent gases is 
 proportional to the " absolute temperature," whose zero is 273^ 
 
CHEMISTRY AND PHYSICS 499 
 
 below freezing on the Centigrade scale, and 493° below freezing 
 (i.e.) 461° below zero, on the Fahrenheit scale. This law is only 
 approximately true for vapors and gases other than permanent 
 gases, for example, steam, ammonia, etc, 
 
 BoyWs Law. — The volume of gases also increases or decreases 
 in proportion to lessening or increasing the pressure. Ordinarily 
 gases are under the atmospheric pressure, which is 15 pounds per 
 square inch. If we increase the pressure upon them to 30 pounds 
 their volume becomes one-half; if we increase it to 45 lb., it 
 becomes one-third, etc., UAg. 
 
 These laws are often consolidated into the statement that 
 the product of the volume and pressure is proportional to 
 the absolute temperature or P-}-V = RT where R is a constant 
 depending on the gas. 
 
 Specijw Gravity. — The specific gravity of bodies is the rela- 
 tive weights of a unit volume. For example, a cubic inch of iron 
 weighs nearly eight times as much as a cubic inch of ice, and a 
 cubic inch of lead or platinum weighs more still. The specific 
 gravity of solids and liquids are usually compared with water as a 
 standard, and water at its temperature of maximum density — 4°C. 
 (39.2° F.) — is given an arbitrary value of 1. Gases are usually com- 
 pared with air as a standard, and air at 0° C. (32° F.) and under 
 a pressure of 760 millimeters of mercury ( = practically 15 lb. 
 per square inch) is given an arbitrary value of 1. 
 
 Heat. — The atoms and molecules of all bodies are never in a 
 state of rest, even though the body itself appears to be quiet. 
 It is this constant and violent motion of the molecules which we 
 know under the name of heat. To raise the temperature of a 
 substance increases the motion and vice versa. In the case of 
 solids the vibration of each molecule is of course confined to a 
 very small space indeed, but in the case of gases, the mole- 
 cules travel until they strike against some other body with 
 force enough to resist them, as, for instance, some other molecule, 
 or the walls of the vessel in which they are contained, when they 
 rebound with equal velocity in another direction. It is, in fact, 
 the constant impact of molecules upon the walls of the containing 
 vessel that causes gases to exert pressure. It also explains why 
 gases expand when their temperature is raised, provided the 
 pressure under which they are confined remains constant, because 
 if their motion is more rapid they exert greater pressure against 
 the containing walls. 
 
500 THE METALLURGY OF IRON AND STEEL 
 
 Conservation of Energy, — The law of conservation of energy 
 tells us that energy can be neither created nor destroyed. We 
 can convert chemical energy into heat, or heat into motion, but 
 we cannot get energy out of anything into which we do not put an 
 equivalent amount in some form or another. We may waste 
 energy, such as energy lost in heat from friction which is useless 
 to us, but it does not cease to exist. 
 
 Physical Properties of Metals 
 
 Tensile Strength, — The tensile strength of a body is its resist- 
 ance to being pulled asunder. It is usually measured in pounds 
 per square inch; that is to say, a bar of wrought iron for example, 
 with 1 sq. in. cross-sectional area^ will support about 50,000 
 lb. weight. 
 
 Stress and Strain. — A stress is a force put upon a body, and a 
 strain is the deformation of the body produced by a stress. For 
 instance, if a bar of wrought iron 1 sq. in. in cross-sectional 
 area and 2 in. long be made to support a weight of 10,000 
 lb., it will stretch about 0.0007 inch; the 10,000 weight is the 
 , stress, and the 0.0007 in. is the corresponding strain. 
 
 Elastic Limit. — In the case just mentioned, if the 10,000 
 weight be removed the strain will be removed. That is, the bar 
 will return to its original length of 2 in. Now if the same bar 
 be loaded with 20,000 lb. it will stretch 0.0014 in., and again 
 this elongation will be lost when the weight is removed. If, how- 
 ever, we load the bar with 30,000, it will stretch a little more than 
 0.0021 in., and now it will not return to its original 2-in. length 
 when the weight is removed, but will be permanently elongated. 
 It has taken a "permanent set," as it is called. The "elastic 
 limit" of a body is the force necessary to produce the first per- 
 manent set. It is usually measured in pounds per square inch. 
 Another way of expressing the elastic limit is to say it is the 
 force beyond which the strain is not proportional to the stress. 
 
 Modulus of Elasticity. — The modulus of elasticity tells of the 
 rigidity or stiffness of a body, that is to say, its rate of yielding 
 under any stress up to the elastic limit. The modulus of elas- 
 ticity is obtained by dividing any stress (up to the elastic limit) 
 by the strain produced per inch of length. For example, the 
 wrought iron mentioned in the last paragraph stretched 0.0014 
 
 ^ Say a round bar about 1 l/S-in. diameter, or a square bar 1 in. on a side. 
 
CHEMISTRY AND PHYSICS 501 
 
 in. in a length of 2 in. = 0.0007 per inch of length, with a stress of 
 20,000 lb. per square inch; its modulus of elasticity will then be: 
 
 20,000 
 
 0.0007 = 28,500,000. 
 
 Percentage Elongation. — After a bar under tensile stress has 
 passed its elastic limit it begins to be permanently elongated in 
 the direction of the pull. A soft metal, like copper or mUd steel, 
 will stretch out somewhat like molasses candy before finally 
 breaking, and may be almost twice as long as it was originally. 
 The increase in length, divided by the original length, is the per- 
 centage elongation. It is usually measured on a length of 2 
 in., or of 8 in. 
 
 Reduction of Area. — When a bar is elongated it of course 
 shrinks in cross-section; finally, just before the bar breaks, it 
 usually "necks down'' on either side of the point of fracture. 
 This type of fracture occurs with soft metals. The original area, 
 minus the area of smallest cross-section after fracture is called 
 the "reduction of area" and this divided by the original area is 
 the "percentage reduction of area." 
 
 Ductility. — The percentage elongation and the percentage 
 reduction of area are usually taken together as the measure of the 
 ductility of a metal. 
 
 Compressive Strength. — The compressive strength of a body is 
 its resistance to crushing. It also is usually measured in pounds 
 per square inch.^ The terms " stress and strain," " elastic limit," 
 and "modulus of elasticity" all have the same meaning when 
 referred to compressive as to tensile stresses. 
 
 Transverse Strength. — If a bar 1 in. square be supported on 
 thin edges placed 12 in. apart its resistance to a force applied 
 half way between the supports is called its " trajisverse strength." 
 Here again we have the same terms, "stress and strain," etc. 
 
 Impact. — If a bar be supported on thin edges placed a certain 
 distance apart and then a falling weight be allowed to strike 
 upon it at a point midway between the supports, its resistance 
 to this force will give an indication of its strength under impact, 
 while the amount that it will bend before breaking will indi- 
 cate its ductility under impact, or under "shock." 
 
 Shearing Strength. — The resistance of a body to being cut in 
 
 * In Great Britain they often use long tons (2,240 lb.) per square inch, instead of 
 pounds, both for tensile and compressive stresses. 
 
502 THE METALLURGY OF IRON AND STEEL 
 
 two by a pair of knife edges is called its shearing strength. 
 Rivets are sometimes tested in this way because they are sub- 
 jected to this kind of stress in service. . 
 
 Torsion. — The resistance of a bar to being twisted like an augur 
 is called its torsional strength. The number of twists it will 
 endure before breaking gives an indication of its ductility under 
 this stress. 
 
 Repeated Stress. — If a certain kind of stress be applied to a 
 body, then relieved, applied again, and so on alternately, this 
 class of test is called " repeated stress." A metal will break under 
 many applications of a repeated stress much less in amount than 
 that required to break it if constantly applied. It is to be under- 
 stood, however, that the interval between the applications of the 
 stress must be very short so the metal will have no opportunity 
 to rest between applications. 
 
 Alternate Stresses. — If we place a body first under tension, then 
 under compression, and so on alternately, it produces what is 
 known as " alternate stresses." It is like in nature to bending a 
 wire back and forth, and metals wiU break under alternate stress 
 even less than their elastic limits under either tension or com- 
 pression alone. 
 
 Toughness. — The toughness of a metal is its resistance to 
 breaking after its elastic limit is passed. It is the direct opposite 
 of brittleness. 
 
 Brittleness. — The brittleness of a metal is the ease with which 
 it breaks after its elastic limit is exceeded. A very brittle steel 
 wiU have an elastic limit exactly equal to its ultimate tensile or 
 compressive strength; that is to say, it will take no permanent 
 elongation or reduction of area; its ductility will be zero. Some 
 metals are more brittle under shock than under constantly applied 
 stress, and vice versa. 
 
 Malleability. — Malleability is the quality of being deformed 
 under a hammer. Gold is the most malleable of metals and can 
 be hammered into sheets of extreme thinness without cracking. 
 
 Resilience. — Resilience is springiness. A very resilient metal, 
 that is, one with a small modulus of elasticity, would be unsuit- 
 able for a bridge even though strong, because its vibration under 
 a moving load would be so great. 
 
 Hardness. — The hardness of a metal is its resistance to being 
 scratched, or to wearing away under friction. In steel metallurgy 
 hardness is often used to mean brittleness, but this is no longer 
 
CHEMISTRY AND PHYSICS 503 
 
 advisable, because we are now making hard steels that are also 
 tough. 
 
 Allotropy. — AUotropy is the capacity that certain elements . 
 have of changing their properties without changing their com- 
 position or purity. For example, we may have pure carbon in 
 the form of diamond, graphite, or charcoal; we may have iron in 
 a magnetic or non-magnetic condition; we may have sulphur in 
 a brittle or in a p-asty state, etc. What the nature of allotropy is 
 we cannot at present tell. It may perhaps have to do with the 
 relations of the different atoms in the molecules of the elemgnt. 
 When elements form compounds the atoms of one are joined to 
 the atoms of the others, and even when elements are in the pure 
 state their atoms are often joined together to form molecules. 
 Allotropy may be a difference in the number of atoms that are in 
 each molecule, or perhaps in the form in which they are joined 
 together. An allotropic change is always accompanied by a loss 
 or gain of heat. 
 
 Crystallization. — The tendency of most elements and com- 
 pounds to arrange themselves into regular forms called crystals 
 is really a powerful force of nature, and one of the most wonderful 
 and charming studies imaginable. The crystalline forms of each 
 particular substance are usually the same, or very similar, but 
 different from almost all other substances. Each crystal is built 
 up of smaller crystals, and these in turn of still smaller ones. 
 The tendency to produce a regular form is well illustrated in the 
 case of alum: If a piece of an alum crystal be broken off and the 
 main part be immersed in a saturated alum solution, the crystal 
 will slowly repair itself and renew the lost part until it is again 
 perfect. Moreover, if the alum solution is impure, the crystal 
 will take to itself only the pure salt, and leave the impurity. 
 
504 
 
 THE METALLURGY OF IRON AND STEEL 
 
 TABLE XXXIV.— COMPARISON OF DEGREES CENTIGRADE AND 
 
 FAHRENHEIT 
 
 Below 
 
 zero 
 
 Above 
 
 zero 
 
 Above zero 
 
 Equivalents. 
 
 C. 
 
 F. 
 
 C. 
 
 F. 
 
 C. F. 
 
 C. F. 
 
 -200°=- 
 
 -328° 
 
 + 525° = 
 
 4- 977° 
 
 + 1,250°= 2,282° 
 
 1°= 1.8 
 
 150 = 
 
 •238 
 
 550 = 
 
 1,022 
 
 1,275 = 2,327 
 
 2 = 3.6 
 
 100 = 
 
 148 
 
 575 = 
 
 1,067 
 
 1,300 = 2,372 
 
 3 = 5.4 
 
 50 = 
 
 58 
 
 600 = 
 
 1,112 
 
 1,325 = 2,417 
 
 4 = 7.2 
 
 
 
 625 = 
 
 1,157 
 
 1,350 = 2,462 
 
 5 = 9.0 
 
 Above 
 
 zero 
 
 650 = 
 
 1,202 
 
 1,375 = 2,507 
 
 6 = 10.8 
 
 
 
 675 = 
 
 1,247 
 
 1,400 = 2,552 
 
 7 = 12.6 
 
 C. 
 
 F. 
 
 
 
 
 
 + 0°=- 
 
 f32° 
 
 700 = 
 
 1,292 
 
 1,425 = 2,597 
 
 8 = 14.4 
 
 25 - 
 
 77 
 
 725 = 
 
 1,337 
 
 1,450 = 2,642 
 
 9 = 16.2 
 
 50 = 
 
 122 
 
 750 = 
 
 1,382 
 
 1,475 = 2,687 
 
 10 = 18.0 
 
 75 = 
 
 167 
 
 775 = 
 
 1,427 
 
 1,500 = 2,732 
 
 11 = 19.8 
 
 100 = 
 
 212 
 
 800 = 
 
 1,472 
 
 1,525 = 2,777 
 
 12 = 21.6 
 
 125 = 
 
 257 
 
 - 825 == 
 
 1,517 
 
 1,550 = 2,822 
 
 13 = 23.4 
 
 150 = 
 
 302 
 
 850 = 
 
 1,562 
 
 1,57,5 = 2,867 
 
 14 = 25.2 
 
 175 = 
 
 347 
 
 875 = 
 
 1,607 
 
 1,600 = 2,912 
 
 15 = 27.0 
 
 200 = 
 
 392 
 
 900 = 
 
 1,652 
 
 1,625 = 2,957 
 
 16 = 28.8 
 
 225 = 
 
 437 
 
 925 = 
 
 1,697 
 
 1,650 = 3,002 
 
 17 = 30.6 
 
 250 = 
 
 482 
 
 950 = 
 
 1,742 
 
 1,675 = 3,047 
 
 18 = 32.4 
 
 275 = 
 
 527 
 
 1,000 = 
 
 1,832 
 
 1,700 = 3,092. 
 
 19 = 34.2 
 
 300 = 
 
 572 
 
 1,025 = 
 
 1,877 
 
 1,725 = 3,137 
 
 20 = 36.0 
 
 325 = 
 
 617 
 
 1,050 = 
 
 1,922 
 
 1,750 = 3,182 
 
 21 = 37.8 
 
 350 = 
 
 662 
 
 1,075 = 
 
 1,967 
 
 1,775 - 3,227 
 
 22 = 39.6 
 
 375 = 
 
 707 
 
 1,100 = 
 
 2,012 
 
 1,800.= 3,272 
 
 23 = 41.4 
 
 400 = 
 
 752 , 
 
 1,125 = 
 
 2,057 
 
 1,825 = 3,317 
 
 24 = 43.2 
 
 425 - 
 
 797 
 
 1,150 = 
 
 2,102 
 
 1,850 = 3,362 
 
 25 = 45.0 
 
 450 = 
 
 842 
 
 1,175 = 
 
 2,147 
 
 1,875 = 3,407 
 
 
 475 - 
 
 887 
 
 1,200 = 
 
 2,192 
 
 1,900 = 3,452 
 
 
 500 = 
 
 932 
 
 1,225 = 
 
 2,237 
 
 2,000 = 3,632 
 
 
APPENDIX A 
 
 Literature Referred to in the Text 
 
 General Text-books and Reference Books on the 
 Metallurgy of Iron and Steel 
 
 1.. H. M. Howe. "Iron, Steel, and Other Alloys," 1903. Pub- 
 lished by Sauveur & Whiting, Boston, Mass. This 
 book contains three chapters upon the ' 'Manufacture of 
 Iron and Steel" and ten chapters upon its constitution 
 and properties, especially from the standpoint of 
 metallography. Upon this latter subject it is without 
 an equal and, like the same author's larger work, bids 
 fair to remain the standard authority for many years to 
 come. 
 
 2. H. H. Campbell, ' 'The Manufacture and Properties of Iron 
 
 and Steel." Fourth Edition. New York and London. 
 1907. This is a great reference book by one of the best 
 of the practical American metallurgical engineers. It 
 is the best American reference book upon the manu- 
 facture of iron and steel, but is not intended especially 
 for beginners or those without technical education. 
 
 3. A. Ledebur. "Handbuch der Eisenhuettenkunde." Fifth 
 
 Edition. 1906. Leipzig. This is an excellent refer- 
 ence book for those who read German, and contains a 
 very complete account of the metallurgy of both iron 
 and steel, and of their properties. There are also 
 classified lists of the literature upon each of the branches 
 of the subject. 
 
 4. Sir I. Lowthian Bell. "Principles of the Manufacture of 
 
 Iron and Steel." London. 1884. This book well 
 accomplishes its aim, namely, to elucidate the princi- 
 ples of iron and steel manufacture, and no man can be 
 either so well informed or so ignorant as not to under- 
 stand the metallurgy of these metals better after reading 
 it. 
 
 505 
 
506 THE METALLURGY OF IRON AND STEEL 
 
 5. H. P. Tiemann. ''Iron and Steel. (A Pocket Encyclo- 
 
 pedia)." New York and London. 1910. There is a 
 wealth of general information in this book arranged for 
 ready reference because in alphabetical order. It is in 
 fact an encyclopedia of the iron and steel industry, 
 with a sufficient explanation of all the terms and pro- 
 cesses and products likely to be met with by those 
 interested in the industry or trade. 
 
 6. Joseph W. Richards. "Metallurgical Calculations." Part I, 
 
 Introduction, Chemical and Thermal Principles, Prob- 
 lems in Combustion. 1906. Part II, Iron and Steel. 
 1907. These problems not only teach how to calculate 
 many very important things in connection with fur- 
 naces and their efficiency, but give a good insight into 
 the principles of the processes themselves. 
 
 7. James M. Swank. ' 'Directory to the Iron and Steel Works of 
 
 the United States." Embracing the Blast Furnaces, 
 Rolling Mills, Steel Works, Forges, and Bloomaries in 
 Every State and Territory. Prepared and published 
 by The American Iron and Steel Association, Phila- 
 delphia. The first edition of this book appeared in 
 1873, and the seventeenth edition in 1907. The data 
 given are very complete and are classified for conven- 
 ient reference. 
 
 8. James M. Swank. "History of the Manufacture of Iron in 
 
 All Ages, and Particularly in the United States from 
 1585 to 1892." Philadelphia. 
 
 9. ''Ryland's Colliery, Iron, Steel, Tin-plate, Engineering and 
 
 Allied Trades' Directory (For Great Britain only) with 
 Brands and Trade Marks." 1906. Published by Eag- 
 land & Co., Ltd., London. 
 
 L6on Gages. "Trait6 de M6tallurgie du Fer." In two 
 volumes. Paris. 1898. The first volume covers the 
 manufacture of iron and steel, and the second, foundry, 
 mechanical treatment and properties. 
 
 John Percy. ''Metallurgy. Iron and Steel." London. 
 1864. This classical book is now chiefly valuable for 
 historical reasons, where its usefulness is often unex- 
 pectedly advantageous, as, for example, in patent 
 litigations, but at the time it was written it was a model 
 for wealth of information (although badly arranged). 
 
APPENDIX A 507 
 
 Hermann Wedding. ' 'Ausfuehrliches Handbuch der Eisen- 
 huettenkunde." Braunschweig. 1906. Four volumes. 
 This is an edited translation of Percy's ''Iron and 
 Steel," brought up to date and greatly enlarged with 
 especial reference to German practice, which is the . 
 second largest in the world. 
 
 Andrew Alexander Blair, ' 'The Chemical Analysis of 
 Iron." A complete Account of all the best known 
 Methods for the Analysis of Iron, Steel, Pig Iron, Iron 
 Ore, Limestone, Slag, Clay, Sand, Coal, Coke, and 
 Furnace and Producer-Gases, Sixth edition. Phila- 
 delphia and London. 1906. 
 
 10. The Journal of the Iron and Steel Institute, — ^Published in 
 
 London, Vol. i, 1869, This periodical appears twice 
 a year and contains not only many original articles of 
 great value, but also an almost complete collection of 
 abstracts of the literature of iron and steel that is 
 published anywhere, classified under headings for 
 convenient reference. Anyone beginning research in 
 any branch of iron and steel metallurgy should com- 
 mence with this journal, as soon as the text-books have 
 been consulted. 
 
 11. Stahl und Eisen. — ^Published in Dusseldorf. Vol. i, 1881. 
 
 This is the best German periodical on iron and steel, and 
 contains not only many valuable original articles and 
 abstracts, but also translations. It is particularly useful 
 in this latter connection, because of its translations 
 of many articles from the Swedish. 
 
 12. Revue de Metallurgie, Published in Paris. Vol. i, 1904. 
 
 This is a very valuable periodical for those who read 
 French, not only for its original articles, but also for its 
 abstracts. Upon the more scientific side of metallurgy, 
 that is to say the properties and constitution of iron 
 and steel, alloy steels, etc., it is without an equal, 
 
 13. The Mineral Industry. — Its statistics, technical and trade. 
 
 Published in New York. Vol. i, 1892. This contains a 
 review every year of the technology and trade of each 
 of the metals listed alphabetically, as well as the sta- 
 tistics of production, price, etc. The articles usually 
 include a review of the progress of the metallurgy 
 during the year. 
 
508 THE METALLURGY OF IRON AND STEEL 
 
 14. The Iron ^g^e.— Published in New York. Vol. i, 1869. 
 
 This is the oldest and largest of the American iron and 
 steel technical magazines, and deals not only with the 
 scientific and technical side of the subject, but also acts 
 as a sort of a weekly newspaper upon the condition of 
 the iron trade and recent happenings of interest. 
 
 15. Transactions of the American Institute of Mining Engineers, — 
 
 Published in New York. Vol. i, 1871. The American 
 Institute of Mining Engineers is the leading aggregation 
 of both mining engineers and metallurgists in America. 
 These transactions contain many original articles of 
 value. 
 
 16. James M. Swank. Annual Statistical Report of the Secretary 
 
 of the American Iron and Steel Associationj containing 
 detailed statistics of the American and foreign iron 
 trade. Published in Philadelphia. 
 
 17. Iron Trade Review. Published in Cleveland, Ohio. Al- 
 
 though this magazine aims to deal principally with iron 
 trade conditions, it contains also a great many technical 
 articles of importance. 
 
 18. Metallurgie. Published in Halle am See. Vol. i, 1904. 
 
 This German magazine contains a great many original 
 articles and abstracts. One should refer to Nos. 2, 3, 
 and the list given below: 
 
 References on the Manufacture of Iron 
 
 20. Robert Forsythe. "The Blast Furnace and the Manufac- 
 
 ture of Pig Iron." The best modern American book on 
 this subject. 
 
 21. Sir I. Lowthian Bell. "Chemical Phenomena of Iron Smelt- 
 
 ing." London. The classical treatise on the chemistry 
 of the blast furnace. 
 
 22. Thomas Turner. "The Metallurgy of Iron."* London. 
 
 1895. 
 
 23. A. de Vathaire. " Les Hautes Fourneaux." Paris. 
 
 24. H. Bauerman. "A Treatise on the Metallurgy of Iron." ^ 
 
 London, 1868. 
 
 25. M. A. Pavlov. "Atlas of Plans for Blast Furnace Construc- 
 
 tion." Gekatermoslov (Russia), 1902. Although these 
 
 1 Starred books refer to both pig iron and wrought iron. 
 
APPENDIX A 509 
 
 drawings are lettered in Russian, one gets much valu- 
 able information from them even without being able to 
 read the language. 
 
 26. Frederick Overman. "The Manufacture of Iron." ^ Phila- 
 
 delphia, 1850. 
 
 27. W. Truran. "The Iron Manufacture of Great Britain."^ 
 
 London, 1865. Revised by J. Arthur Phillips and W. 
 H. Dorman. 
 
 General Reference Books on Steel 
 
 30. Henry M. Howe. "The Metallurgy of Steel." Vol. i, 1890. 
 
 New York. This is the recognized standard authority 
 on the metallurgy of the, Bessemer and crucible steel 
 processes, and upon the properties of steel as far as they 
 were known and understood at the time when this book 
 was written. It will long remain a classic. There have 
 been many editions of different dates, but no change in 
 text since 1890. 
 
 31. F W. Harbord. "The Metallurgy of Steel." 1905. Lon- 
 
 don. With a Section on Mechanical Treatment by 
 J. W. Hall. Next to No. 30, this is the most complete 
 and thorough book on steel ever written in English. 
 The section on mechanical treatment is the best extant. 
 34. H. Noble. "Fabrication de I'Acier." Paris, 1905. As its 
 name indicates, this book deals chiefly with the manu- 
 facture of steel, the section on properties being very 
 small. 
 
 References on the Bessemer Process 
 
 See Nos. 2, 3, 4, 30, 31, and the following list: 
 
 50. Richard Akerman. "The Bessemer Process as Conducted 
 
 in Sweden." Trans. American Institute of Mining En- 
 gineers. Vpl. xxii, 1893, pages 277 et seq. 
 
 51. Henry M. Howe. "Notes on the Bessemer Process." Jour- 
 ' nalj Iron and Steel Institute, No. 11, 1890, pages 100 
 
 et seq. 
 
 52. Bradley Stoughton. "The Development of the Bessemer 
 
 Process for Small Charges." Trans. American Institute 
 of Mining Engineers. Vol. xxxiii, 1903, pages 846 et 
 seq. 
 
510 THE METALLURGY OF IRON AND STEEL 
 
 53. Friedrich C. G. Miiller. Untersuchungen ilber den deutschen 
 
 Bessemer process, Zeitschrift des Vereines deutscher 
 Ingenieure. Vol. xxii, 1878, pages 384-404 and 454- 
 470. This is one of the most comprehensive studies of 
 the metallurgy of the Bessemer process in any language. 
 
 54. Hermann Wedding. "The Basic Bessemer, or Thomas 
 
 Process." Translated into English by William B. 
 Phillips and Ernst Prochaska. New York, 1891. 
 This the the most complete account of the basic Besse- 
 mer process. 
 
 References on the Open-Hearth Process 
 
 See Nos. 2, 31, and: 
 
 61. M. A. Pavlov. "Album of Drawings Relating to the Manu- 
 
 facture of Open-Hearth Steel." St, Petersburg, 1908. 
 (See also No. 25.) 
 
 62. W. M. Carr. "Open-Hearth Steel Castings." 1907. This 
 
 contains a concise, simple and readily intelligible dis- 
 cussion of the acid and basic open-hearth processes and 
 practice. 
 
 References on Defects in Ingots 
 
 See Nos. 1, 3, and: 
 
 70. C. A. Caspersson. Reviewed by Richard Akerman. "The 
 
 Influence of the Temperature of the Bessemer Charge on 
 the Properties of the Steel Ingots." Stahl und EiseUj 
 1883, pages 71-76. 
 
 71. Henry M. Howe. "Piping and Segregation in Steel Ingots." 
 
 Transactions, American Institute of Mining Engineers, 
 1907. Advance proof. 
 
 72. Henry M. Howe and Bradley Stoughton. "The Influence 
 
 of the Conditions of Casting on Piping and Segregation, 
 as Shown by Means of Wax Ingots." Transactions, 
 American Institute- of Mining Engineers, 1907. Ad- 
 vance proof. 
 
 73. N. Lilienberg. "Piping in Steel Ingots." Transactions, 
 
 American Institute of Mining Engineers, vol. xxxvii, 
 1906, pages 238-247. 
 
 74. Benjamin Talbot. " Segregation in Steel Ingots." Journal, 
 
 Iron and Steel Institute, No. 11, 1905, pages 204-247. 
 
APPENDIX A 511 
 
 75. N. Lilienberg. " The Compression of Semi-liquid Steel 
 
 Ingots." Journal of the Franklin Institutej February, 
 1908. 
 
 76. E. von Maltitz. "Blowholes in Steel Ingots." Transac- 
 
 tions, American Institute of Mining Engineers, vol. 
 xxxviii, 1907, pages 412-447. 
 
 References on Foundry Practice 
 
 90. George R. Bale. ' 'Modern Iron Foundry Practice." Part I. 
 
 Foundry Equipment, Materials Used, and' Processes 
 Followed. 1902. Part II. Machine Molding and 
 Molding Machines, Physical Tests of Cast Iron, Methods 
 of Cleaning Castings, Foundry Accounting, etc., etc. 
 1905. London. This is the most scientific book on 
 iron foundry practice published. 
 
 91. The Foundry, Published monthly. Vol. xxxi, 1907, Cleve- 
 
 land, Ohio. 
 
 92. Thomas D. West. ''American Foundry Practice." Treat- 
 
 ing of Loam, Dry-Sand and Green-Sand Molding, and 
 containing a Practical Treatise upon the Management 
 of Cupolas and the Melting of Iron. Published in New 
 York. This book is tremendously popular among 
 foundry men and editions appear at frequent intervals. 
 
 93. William J. Keep. "Cast Iron." A Record of Original Re- 
 
 search. New York and London, 1902. 
 
 94. Thomas Turner. "Lectures on Iron-Founding." London, 
 
 1904. 
 
 95. 'Teuton's Foundry List." Cleveland, Ohio, 1906. A 
 
 Directory of the Foundries of the United States and 
 Canada. 
 For an account of small converters for steel casting work see 
 Reference No. 52 and the files of Nos. 91 and 10. 
 
 96. R. Moldenke. "A Suggested Change in Cupola Practice." 
 
 The Foundry, January, 1909. Pp. 225-227. 
 
 97. E. A. Custer, Iron Age, Vol. Ixxxi, 1908, pp. 1227-1234; Iron 
 
 Trade Review, Vol. xliv, pp. 304-315; The Foundry, 
 Vol. xxxii, pp. 137-145; Vol. 33, pp. 10-12; Vol. xxxiv, 
 pp. 203-208. 
 Proceedings of The American Society for Testing Materials, 
 Vol. ix, pp. 442-455. 
 
512 THE METALLURGY OF IRON AND STEEL 
 
 References on the Solution Theory op Iron and Carbon 
 See Nos. 1, 5, 10, 11, 12, and the following: 
 
 100. Reports of the Alloys Research Committee of the Institution 
 
 of Mechanical Engineers. (England). 
 
 101. G. B. Upton. "The Iron-Carbon Equilibrium." The 
 
 Journal of Physical Chemistryj October, 1908, vol. 
 xii, pages 507-549. This is a very valuable study and 
 r^sum^ of the recent researches on the constitution of all 
 the alloys of iron and carbon, and gives the chief known 
 facts of scientific interest in a concise form. 
 
 References on the Constitution of Steel 
 
 110. F. Osmond and G. Cartaud. ''The Crystallography of 
 
 Iron." Journal Iron and Steel Institutej No. Ill, 1906, 
 pages 444-492. 
 
 111. Albert Sauveur. ' 'The Constitution of Iron-Carbon Alloys. 
 
 Same JowrnaZ, No. 4, 1906, pages 493-575. 
 
 112. J. O. Arnold and A. A. Read. "Chemical Relations of 
 
 Carbon and Iron." Journal of the Chemical Society^ 
 vol. Ixv, page 788. 
 
 113. J. 0. Arnold and G. B. Waterhouse. "The Influence of 
 
 Sulphur and Manganese on Steel." Journal Iron and 
 Steel Institute, No. 1, 1903, pp. 136-160. 
 
 1 14. J. 0. Arnold. "The Influence of Elements on Iron." Same 
 
 Journal^ No. 1, 1894, p. 107. 
 
 115. J. E. Stead. B&me Journal, No. 11, 1900, pages 60 et seq., 
 
 and also Cleveland Institution of Engineers, September 
 6, 1906. 
 
 116. Bradley Stoughton. "Notes on the Metallography of 
 
 Steel." International Engineering Congress, 1894. 
 Transactions J American Society of Civil Engineers, 
 vol. liv. Part E, pages 357-421. 
 
 117! Proceedings of the American Society for Testing Materials, 
 vol. i, 1901; vol. vii, 1907. Philadelphia. This society 
 is composed of representative men from the great 
 purchasers and users of engineering and other kinds of 
 materials in America, from the manufacturers and of 
 representative scientific men. All sides of the question 
 are usually represented in the discussions of this society. 
 
 118. Proceedings of the International Association for TesHng 
 Materials, Vienna, Austria. 
 
APPENDIX A 513 
 
 References on the Constitution of Cast Iron 
 
 See Nos. 1, 32, 90, 91, 93, arid 
 
 120. W. G. Scott. ''Effect of Variations in the Constituents of 
 
 Cast Iron." Proceedings of the American Society for 
 Testing MaterialSj vol. ii, 1902, pages 181-206. 
 
 121. Transactions of the American Foundrymen's Association. 
 
 122. P. Longmuir. ''The Influence of Varying Casting Tem- 
 
 perature on the Properties of Alloys," Journal Iron 
 and Steel Institute j 1903, No. 1, pages 457-476. Also 
 Idem 1904, No. 1, pages 420-436. 
 
 123. J. J. Porter. "The Constitution of Cast Iron." Trans. 
 
 American Foundrymen's Association. Vol. xix, 1910, 
 pages 37-200. 
 
 References on Malleable Cast Iron 
 
 See 90, 91, 92, and 101. 
 
 130. Richard Moldenke. "Malleable Castings, Parts I, II, and 
 
 III. Instruction Papers Nos. 547A, 547B, and 547C, 
 of the International Correspondence Schools, Scranton, 
 Pa. This is the most thorough and valuable treatment 
 of this subject that I know of. 
 
 131. Richard Moldenke. "The Production of Malleable Cast- 
 
 ings," Cleveland, Ohio, 1911. The best treatment 
 in English of this subject. 
 
 References on the Heat Treatment of Steel 
 
 140. Report of the Alloys Research Committee of the Institution of 
 
 Mechanical Engineers (England). Minutes of the 
 Proceedings. Report No. 1, October, 1891 ; No. 2, April, 
 1893; No. 4, February, 1897; No. 5, February, 1899; 
 No. 6, January and February, 1904; No. 7, November 
 and December, 1905. 
 
 141. Joseph V. Woodworth. "Hardening, Tempering, Anneal- 
 
 ing, and Forging of Steel." New York, 1903. This 
 book tells much of the practice and manipulation of 
 these operations. 
 
 142. John Lord Bacon. "Forge Practice (Elementary)." New 
 
 York, 1904. This book also deals largely with practice 
 and manipulation. 
 
 33 
 
514 THE METALLURGY OF IRON AND STEEL 
 
 143. Alfred Stansfield. "The Burning and Overhjeating of 
 
 Steel." Journal Iron and Steel Institutej No. 11, 1903, 
 pages 433-468. 
 
 144. Carl Benedicks. "Recherches Physiques et Physico- 
 
 Chimiques sur TAcier au Carbone." Upsala, 1904. 
 With a good bibliography. 
 
 145. Hanns Ereiherr von Jiiptner. '^Siderology, the Science of 
 
 Iron." Translated by Charles Salter, London, 1902. 
 
 146. Hanns Freiherr von Jiiptner. "Grundzuege der Siderologie. 
 
 In three volumes. Vol. i, "The Constitution of Iron 
 Alloys and Slags/' Leipzig, 1900; vol. ii, ''Relation 
 Between the Thermal and Mechanical Changes, Con- 
 stitution and Properties of Iron Alloys," 1902; vol. 
 iii, "The Influence of Impurities; the Manufacture of 
 Iron and Steel," etc. 
 
 147. J. W. Mellor. "The Crystallization of Iron and Steel." 
 
 London, 1905. This is a very clear and readily intel- 
 ligible discussion of the heat treatment of steel, and an 
 introduction to metallography. 
 
 148. George W. Ailing. "Points for Buyers and Users of Tool 
 
 Steel." New York, 1903. Being a general review 
 of the main sources of trouble and how to avoid 
 them. 
 
 149. E. F. Lake. "Composition and Heat Treatment of Steel." 
 
 N.Y. 1911. A good practical book on heat treatment. 
 
 References on Alloy Steels. 
 
 150. G. B. Waterhouse. "The Influence of Nickel and Carbon 
 
 on Iron." Journal of the Iron and Steel Institute, vol. ii, 
 1905, pages 376-407. This includes a bibliography of a 
 part of the subject. 
 
 151. G. B. Waterhouse. "The Burning, Overheating, and 
 
 Restoring of Nickel Steel." Proceedings of the Ameri- 
 can Society for Testing Materials, vol. vi, 1906, pages 
 247^-258. 
 
 152. A. L, Colby. "A Comparison of Certain Physical Proper- 
 
 ties of Nickel Steel and Carbon Steel." Published by 
 the Bethlehem Steel Company, 1903. A very complete 
 compilation. 
 
 153. L. Dumas. "The Reversible and Irreversible Transforma- 
 
APPENDIX A 515 
 
 tions of Nickel Steel." Journal of the Iron and Steel In- 
 stitute, No. 11, 1905, pages 255-300. 
 
 154. D. H. Browne. "Nickel Steel; A Synopsis of Experiment 
 
 and Opinion." Transactions of the American Institute of 
 Mining Engineers, vol. xxix, 1899, pages 569-648. Very 
 complete, with a good bibliography. 
 
 155. E. A. Hadfield. ''Alloys of Iron and Manganese." Pro- 
 
 ceedings of the Institution of Civil Engineers (England) , 
 1888. 
 
 156. R. A. Hadfield. "Manganese Steel." Journal of the Iron 
 
 and Steel Institute, 1888. 
 
 157. R. A. Hadfield. "Alloys of Iron and Silicon." Ibid,, 1889. 
 
 158. R. A. Hadfield, ^'Alloys of Iron and Aluminum." Ihid.j 
 
 1890. 
 
 159. R. A. Hadfield. "Alloys of Iron and Chromium," Ibid., 
 
 1892. 
 
 1510. R. A. Hadfield. "Alloys of Iron and Nickel," Proceed- 
 
 ings of the Institution of Civil Engineers (England), 1899. 
 
 1511. R. A. Hadfield, "Alloys of Iron and Tungsten." Journal 
 
 of the Iron and Steel Institute, No. 11, 1903. 
 
 1512. Henry M. Howe. "Manganese Steel." Proceedings of the 
 
 American Society of Mechanical Engineers, 1891. 
 
 1513. C. E. Guillaume. His articles will be found listed in some 
 
 of the cross-references cited. 
 
 1514. L. Guillet. "Steel Used for Motor Car Construction in 
 
 France." Journal of the Iron and Steel Institute, No. 
 11, 1905, pages 166-203. 
 
 1515. L, Guillet. "The Use of Vanadium in Metallurgy." 
 
 Same journal, pages 118-165, 
 
 For Guillet's other memoirs on alloy steels consult th^ 
 Revue de Metallurgie, and the index of the Journal of 
 the Iron and Steel Institute from 1904 to date, 
 
 1516. J. M. Gledhill. "The Development andUse of High-speed 
 
 Tool Steel." Journal of the Iron and Steel Institute, No. 
 11, 1904, pages 127-182. 
 
 1517. H. C. H. Carpenter. "High-speed Tool Steels." Same 
 
 journal, No. 1, 1905, pages 433-473, and No. Ill, 1906, 
 pages 377-396, 
 
 1518. J. T. Nicolson. "Experiments with Rapid-cutting Steel 
 
 Tools." Manchester Municipal School of Technology, 
 Manchester^ England, 1903. 
 
516 THE METALLURGY OF IRON AND STEEL 
 
 1519. J. Kent Smith. On vanadium steels; see Transactions of 
 
 the American Institute of Mining Engineers^ vol. xxxvii, 
 1907, pages 727-732. Also The Iron Trade Review, 
 November- 7, 1907, page 729. Vol. xli. 
 
 1520. (For Vanadium steels see also the index of the Journal of 
 
 the Iron and Steel Institute j especially 1905 tO' date.) 
 
 1521. L. Guillet. ' 'Quaternary Steels." Journal of the Iron 
 
 and Steel InstitutCj No. 11, 1906, pagfes 1-141. 
 
 1522. For all the alloy steels one should look in the four volumes 
 
 of Revue de Mitallurgie, for which see No. 10, at end of 
 Chapter I, 
 
 1523. De Mozay. "Etude de la Duret6 dans les Aciers k Outil 
 
 de Tour." Revue de Metallurgies vol. iv, 1907, pages 
 885-900. 
 
 1524. L6on Guillet. "Nouvelles Recherches sur les Aciers au 
 
 Vanadium Ternaires et Quart ernaires." Revue de 
 Mitallurgiej vol. iv, 1907, pages 775-783. 
 
 1525. P. Nicolardot. "Le Vanadium."- Paris. About 1906. 
 
 1526. A. J. Rossi. "The Metallurgy of Titanium." Transactions 
 
 of the American Institute of Mining Engineers^ vol. 
 xxxiii, 1903, pages 179-197, with cross-references. 
 
 1527. A. J. Rossi. "Titaniferous Iron Ores." Engineering 
 
 and Mining Journal^ vol, Ixxviii, 1904, pages 501 
 and 502. 
 
 1528. J. Hoerhager. "TJeber titanhaltiges Holzkohlen-Roheisen 
 
 von Turrach in Obersteiermark." Oesterreichische Zeit- 
 schrift fiir Berg- und Huettenwesenj vol. lii, 1904, pages 
 571-577. 
 
 References on Corrosion 
 
 See especially Nos. 117, 118, and 
 
 160. AUerton S. Cushman. ''The Corrosion of Iron." Proceed- 
 
 ings of the American Society for Testing Materials^ vol. 
 vii, 1907, pages 211-228. Same published as Bulletin 
 ■No. 30, Department of Agriculture, Washington, D. C. 
 With many cross-references. 
 
 161. Henry M. Howe. "Relative Corrosion of Wrought Iron 
 
 and Steel." See Proceedings, International Association 
 for Testing Materials, 1900, vol. xi, Part I, pages 229- 
 266. Paris, 1901. Abstracted in No. 162, and in 
 
APPENDIX A 517 
 
 Journal Iron and Steel Institute, No. 11, 1900, pages 567, 
 568. . 
 
 162. Frank N. Speller. "Steel and Iron Wrought Pipe." The 
 
 Iron Age J March 2, 1905, vol. Ixxv, pages 741-745. With 
 several cross-references on relative corrosion. 
 See also the indices of No. 8 for abundant references. 
 
 163. J. Lewkowitsch. "Chemical Analysis of Oils, Fats and 
 
 Waxes." London, 1898. 
 
 164. A. H. Churcli. "The Chemistry of Paints and Painting." 
 
 London, 1901. 
 
 References on the Electro-metallurgy op 
 Iron and Steel 
 
 171. Eugene Haanel, Superintendent of Mines. " Reports of the 
 
 Commission Appointed to Investigate the Different 
 Electro-thermic Processes for the Smelting of Iron Ores 
 and the Making of Steel." Department of the Interior, 
 Ottawa, Canada, 1904 and 1907. 
 
 172. A number of valuable articles and references in vols, i to ix, 
 
 1902 to 1911 inclusive. The Electrochemical and Metal- 
 lurgical Industry. 
 
 173. C. F. Burgess and Carl Hambuechen. "Electrolytic Iron." 
 
 American Electro-chemical Society, April, 1904. Ab- 
 stracted in The Electro-chemical and Metallurgical In- 
 dustry, vol. ii, 1904, pages 183-185. 
 
 174. Geo. P. ScholL "The Manufacture of Ferro-alloys in the 
 
 Electric Furnace." The Electro-chemical and Metal- 
 lurgical Industry, vol. ii, 1904, pages 349-351, 395-396 
 and 449-452. 
 
 175. A number of valuable articles in W. Borcher's Metallurgie, 
 
 vols, i to viii, 1904 to 1911 inclusive. 
 
 176. Several valuable articles and abstracts in Henry Le Chate- 
 
 lier's Revue de Metallurgie, vols, i to viii, 1904 to 1911 
 inclusive. 
 
 177. John B. C. Kershaw. "Electric Furnace Methods of Iron 
 
 and Steel Production," vols, xxxix and xl, The Iron 
 Trade Review, 1906 and 1907. 
 
 178. G. K. Burgess. " Melting-Points of the Iron-group Ele- 
 
 ments," Reprint 62. Bureau of Standards, Washing- 
 ton, D. C. See this Bulletin for best references on 
 pyrometers. 
 
518 THE METALLURGY OF IRON AND STEEL 
 
 References on the Metallography of Iron and Steel 
 
 For preparation of microscopic specimens: 
 
 180. W. CampbelL "Notes on Metallography." Columbia School 
 
 of Mines Quarterly ^ vol. xxv, No. 4 and vol. xxvii, No. 4. 
 
 181. F. Osmond and J. E. Stead. "Microscopic Analysis of 
 
 Metals." London, 1904. 
 
 182. Henry Le Chatelier. "Sur la Metallographie Microsco- 
 
 pique." Bulletin de la SociStS d^ Encouragement pour 
 rindustrie Nationale^ April, 1896. (In English, see) 
 The Metallographistj 1901, vol. iv. Also : Congr^s 
 International de Li^ge, Section de M6tallurgie, vol. i, 
 pages 255-284. 
 
 183. E. Heyn. See especially, Verhandlungen des Vereins zur 
 
 Befoerderung des GewerbfleisseSj 1904, pages 355-397. 
 Also: Stahl und EiseUj vol. xxvi, pages 8-16 and voL 
 xxvi, pages 580-596, 
 
 184. J. E. Stead. Journal of the Iron and Steel Institute^ 1894, 
 
 No. 1, page 292; and the Proceedings of the Cleveland In- 
 stitution of Engineers (England), 1900. 
 
 185. The Metallographist, vols, i to vi, 1898 to 1903. Edited by 
 
 A. Sauveur, Boston, Mass. Succeeded by The Iron and 
 Steel Magazine to 1906. 
 
 186. Paul Goerens. " Einfiihrung in die Metallographie." Halle 
 
 a. S., 1906. 
 For micro-photography: 
 
 187. Edward Bausch. " Manipulation of the Microscope." 
 
 Third edition. Rochester, N. Y., 1897. 
 
 188. Walter Bagshaw. " Elementary Photo-micrography." Lon- 
 
 don, 1902. 
 
 189. Louis Derr. "Photography for Students of Physics and 
 
 Chemistry." New York, 1906. 
 1800. H. Behrens. "Das Mikroskopische Gefuege der Metalle 
 und Legierungen." Hamburg, 1894. 
 See also bibliography given at end of No. 116. 
 
INDEX 
 
 Acetylene, 486 
 
 Acid, compared with basic Bes- 
 semer, 109-110 
 
 compared with basic steel, 48, 
 153, 271 
 
 distinguished from basic steel, 
 155, 156 
 
 furnaces for steel castings, 271- 
 274 
 Acids, definition of, 488 
 
 effect on corrosion, 413 
 Acid steel, oxygen in, (see oxygen,) 
 
 vs, basic furnaces for castings, 
 271-274 
 Adulterants in linseed oil, 422 
 After-blow in basic Bessemer, 107 
 Air, amount of moisture in, 32 
 
 and moisture producing rust, 
 412 
 
 composition of, 21, 32, 477 
 
 weight of, 21 
 Air-furaace, 249, 341, 343, 348 
 
 charging of, 346 
 
 compared with cupola, 347-348 
 
 description of, 343-347 
 
 lining of, 344 
 
 melting in, 343, 346-348 
 
 sizes of, 346-347 
 
 tapping of, 346 
 
 vs. cupola, 347-348 
 Air-hardening steel, 400-405 
 
 necessary to burn coke, 255 
 Akerman, Richard, 183, 509 
 Aktiebolaget electrometall, 430 
 Ailing, George W., 514 
 Alloys, 275-298, 498 
 Alloy steels, 389-411 
 
 definition of, 389 
 
 manufacture of, 390 
 
 quartemary alloys, 390 
 
 ternary alloys, 389-390 
 
 Allotropic modifications of iron, 
 299-302, 370-379 
 theory, 371, 372 
 AUotropists, 372 
 Allotropy, 503 
 
 Alpha iron, 300, 370-373, 312-314 
 Alternate stress, 391, 408-^10, 502 
 Alumina in slags, 31, 42 
 Aluminum, chemistry of, 494 
 
 in steel, 159, 160, 167 
 American iron and steel manufac- 
 ture, scheme of, 46 
 Lancashire process, 54 
 Analysis, 478 
 Anhydrous acids, 488 
 Animal forms, composition of, 477 
 Annealing boxes for malleable cast- 
 ings, 349, 350 
 carbon, 303 
 
 malleable castings, 46, 339-356 
 ovens, 350 
 pots for malleable castings, 
 
 349, 350 
 steel, 165, 166, 211, 214, 357, 
 358-360, 365, 369, 373, 374, 
 379-384, 405 
 Anthracite, 452 
 Anvils, chilling of, 333, 334 
 Armor plate, 213, 400 
 Arnold, J. O., 512 
 Arsenic, effect on steel, 311 
 Atomic weights, 483, 484 
 Atoms, 483, 489 
 Austen. See Roberts-Austen. 
 Austenite, 288, 290, 321, 324, 374- 
 
 379, 399, 401, 405 
 Automobile steels, 177, 400, 408- 
 , 410,411 
 
 Baby Bessemer converters, 270-274 
 Bacon, John Lord, 513 
 
 519 
 
520 
 
 INDKX 
 
 Badly made material and corrosion, 
 
 418 
 Bagshaw, Walter, 518 
 Bakhuis-Roozeboom, H, W., 295, 
 
 297 
 Bale, Geo. R., 511 
 Balling in puddling process, 63 
 Bar iron, 56 
 Bars, forging of, 176 
 
 rolling of, 183 
 Bases, definition of, 488 
 Basic BesseKaer compared with acid, 
 109, 110 
 process, 106-110 
 compared with acid steel, see 
 
 acid 
 distinguished from acid steel, 
 
 see acid 
 furnace lining, 488 
 process, 46, 50, 106-110, 12&- 
 
 138 
 slag, 488 
 
 steel for railroad rails, 152 
 steel in U. S., 153 
 steel, oxygen in, 150, 311 
 steel, strength of, 310 
 vs. acid furnaces for steel cast- 
 ings, 271 
 Bar-relief polishing, 445 
 Bath in Bessemer converter, depth 
 
 of, 85 
 Bauerman, H., 508 
 fiausch, Edward, 518 
 Bed of cupola, 250, 254, 257, 266- 
 
 270 
 Beehive coke oven, 452, 453-456 
 Behrens, H., 518 
 Bell, 209 
 Bell, Sir , I. Lowthian, 52, 505, 
 
 508 
 Bell-Krupp process, 52 
 Bell of blast furnace, 17 
 Bench, 211 
 
 molding, 228 
 Benedicks, Carl, 514 
 Bertrand-Thiel process, 148 
 Bessemer blow, 85, 100-102 
 boil, 100 
 bottoms, 84, 85, 273, 274 
 
 Bessemer, compared with open- 
 hearth steel, 151-153, 271, 
 272 
 converter, 47, 81-85, 272-274, 
 deoxidizing in, 103 
 distinguished from open-hearth 
 
 steel, 156 
 flames, 78, 87, 98, 100-102, 107 
 fume, 102 
 gases, 98, 99 
 ingots, weight of, 90 
 iron, analyses of, 8, 79 
 lining, 82, 83, 273, 274 
 ores defined, 15, 16 > 
 parts, 85 
 pig iron, 8, 79 
 process, 47, 48, 78-110 
 
 recarburizing in, 48, 85, 102- 
 
 104, 108, 151, 170 
 temperature in, 87, 88, 89 
 slags, 79, 83, 84, 93, 96, 99, 100, 
 
 104, 107, 108 
 steel in U. S., 47 
 steel, strength of, 308-311 
 steel, uses of, 103, 152 
 Beta iron, 300-302, 371, 372, 378, 
 
 3J9, 312-314 
 Billets, 183 
 Bi-prefix, 490 
 Bisilicate, 493 
 Biting of piece by rolls, 189 
 Bituminous coals, 452, 453, 454 
 Bi-valent, 490 
 
 Black heart malleable castings, 339, 
 355 
 iron, 435 
 Blast furnace, 8-44 
 blast, 21, 32 
 chemistry of, 27-32 
 cinder, 31 
 
 dimensions and parts, 17-24, 40 
 fuels, 9, 451^63 
 limestone used in, 9 
 lining of, 19-21, 17 
 ore used in, amount of, 9 
 
 varieties, 9-16 
 slags, 31, 34 
 
 V smelting practice, 24-40 
 stoves,21-24 
 
INDEX 
 
 521 
 
 Blast vs. electric furnace, 427, 428 
 Blast, in Bessemer process, 47, 91 
 
 in blast furnace, 21, 32 
 
 pressures in cupola, 254-256, 
 266-269, 270 
 
 volumes in cupola, 256 
 Blister steel, 70 
 Blooming rolls, 189 
 Blooms, 206 
 
 Blower for cupola, 255, 256 
 Blow-holes, 48, 150, 158, 159-161, 
 172, 271, 306, 323, 418, 419 
 Blowing engines for blast furnace, 
 
 20, 21, 25, 
 Blowpipe, 38, 470, 471 
 Blue powder, 424 
 'Body' in bar iron, 56 
 Boilers over heating furnaces, 215 
 Boiling Hnseed oil, 422 
 Boilings, 66 
 
 Bosh of blast furnace, 18, 19, 21 
 Bottom casting, 163 
 Bottoms for heating furnaces, 217- 
 
 219 
 Boyle's law, 499 
 Breeze, defined, 456 
 Brinell, J. A., 159 
 British thermal unit, 473, 482 
 Brittleness, defined, 502 
 
 of Stead, 368 
 
 of steel, 210, 211, 309, 311, 312, 
 372, 373 
 Browne, D. H , 515 
 BuUdog, 58 
 
 Bundled scrap, 47, 68, 154 
 Burdening the cupola, 261-264 
 Burgess, C. F., 440, 441, 517 
 Burgess, G. K., 517 
 Burglar-proof safes, 182, 399 
 Burning of malleable cast iron, 346 
 
 molds in drying, 231 
 
 steel, 365 
 Busheling, 357-370 
 Butterfly reversing valve, 468 
 Butt-welded tubes', 209 
 By-product coke oven. See Retort 
 coke oven. 
 
 Calcining siderite, 10 
 
 Calcium, chemistry of, 495 
 
 chloride used in open hearth, 
 
 134 
 fluoride used in open hearth, 
 134 
 Calculating a blast-furnace charge, 40 
 Calorie defined, 482 
 Calorific equation, in Bessemer proc- 
 ess, 104, 105 
 of basic Bessemer, 108, 109 
 equivalent, 473 
 Campbell open-hearth process, 147, 
 
 148 
 Campbell, H. H., 8, 117, 138, 152, 
 
 155, 169, 326, 327, 505 
 Campbell, William, 377, 518 
 Camera for micro-photography, 448 
 Cannon, 179, 180, 212 
 Carbide of iron. See Cementite. 
 Carbon. See also Graphite and 
 Combined carbon. 
 as a reducing agent, 491 
 chemical afl&nity for iron, 6 
 chemistry of, 492, 493 
 combined, in cast iron, 316, 322, 
 
 323 
 eifect on cast iron (see also 
 
 below), 319-323 
 effect on hardening, 369, 369- 
 
 388 
 control of in pig iron, 323 
 in iron and steel, 3, 6, 45, 75, 
 166, 298, 310, 307-311, 317, 
 318, 319-323, 357, 369, 370, 
 371, 384-388, 400, 407, 497 
 in the blast furnace, 27 
 total, in cast iron, 318 
 Carbonists, 372 
 Carbonizing, see cementation. 
 Carbon theory of hardness, 371, 372 
 Carbiirization of wrought iron, the, 
 
 68-77 
 Car-casting process, 89, 90 
 Carpenter, H. C. H., 514, 404 
 Carr, W. M., 510 
 Cartaud, C, 512 
 
 Case-hardening, see cementation. 
 Caspersson, C. A., 510 
 Casting of ingots, 161-171 
 
522 
 
 INDEX 
 
 Castings, 270-274 
 
 compared with forgings, 172 
 not burning, 367 
 Casting temperature of steel ingots, 
 87, 88, 89, 151, 152, 159- 
 163 
 temperature of cast iron, 337 
 with large end up, 163 
 Cast iron. See also blast furnace, 
 pig iron, 
 checking of, 331 
 chilUng of, 234, 318, 333-336 
 constitution of, 315-337 
 corrosion of, 414-416 
 definition of, 6, 8 
 density of, 329, 330 
 description of, 3 
 fluidity of, 162, 323 
 gray, as impure steel, 317 
 color due to, 316, 317 
 cooling curve of, 320 
 definition of, 7 
 density of, 329, 330 
 photomicrograph of, 335 
 properties of, 316-337 
 mottled, 7 
 oxidation of, 159 
 Cast iron pipe, 256, 416, '423 
 properties of, 315-337 
 rate of cooHng of, 318 
 rolls, 187, 188 
 
 shrinkage of, 162, 247, 248, 249, 
 319-321, 322, 323, 327- 
 329, 331, 352 
 softness of, 331-333, 
 solidification of, 320, 321, 323 
 spongy spots in, 321, 329, 330 
 steel scrap in, 330, 341 
 strength of, 331-333 
 vanadium in, 411 
 vs. steel, 315 
 
 workabiUty of, 322, 331-333 
 white, definition of, 7 
 
 properties of, etc., 187, 238, 
 296, 297, 316, 318, 320, 
 328, 333-337 
 Cast steel, 155, 353, see also crucible 
 
 steel and steel castings 
 Cement manufacture from slag, 34 
 
 Cement, carbon, 303 
 Cementation of iron, 28, 68-70, 384- 
 388 
 
 carbon added, 388 
 
 carbonizing material, 386, 387 
 
 heat treatment of, 388 
 
 influence of time, 386, 387 
 
 steel carbonized, 387, 388 
 
 temperates used, 385, 386 
 
 theory of, 385 
 Cementite, 294, 296, 298, 302-305 
 
 307, 316, 325, 378 
 Centigrade, 482, 504 
 
 converted to Fahrenheit degrees 
 504 
 Chaplets, 234, 235 
 Charcoal, 451^53 
 
 fineries, 53 
 
 hearth, Walloon, 56 
 
 iron, 53-56, 156, 353, 453 
 Charpy, G., 400 
 
 Checking, 158, 233, 241, 247, 261 
 Cheek, 226 
 Chemical aflSnity, 476 
 
 changes, 475 
 
 compounds, 476, 488-491 
 
 energy, 481 
 
 equations, 483-486 
 
 reactions, blast furnace, 27 
 
 reactions, 491-496 
 
 solutions, 496^98 
 
 stability, 490 
 
 symbols, 484 
 Chemistry and physics introductory 
 to metallurgy, 475-504 
 
 of acid open hearth process, 139 
 
 of basic Bessemer process, 107- 
 110 
 
 of basic open hearth process, 
 132-138 
 
 of Bessemer process, 91-110, 
 47, 48 
 
 of crucible process, 75, 76 
 
 of puddling, 63-67 
 Chilling castings, 3lS, 333, 334 
 Chill molds, 234-238 
 Chrome ore in open hearth, 119, 123 
 
 steel, 181, 389, 390, 399, 400, 
 402, 406, 408-410 
 
INDEX 
 
 523 
 
 Chromic acid and corrosion, 414 
 Chromium, effect on crystallization, 
 
 361 
 Church, A. H., 517 
 Cinder, blast furnace, 31 
 See also slags, 
 in puddling process, 61, 66, 67 
 notch of blast furnace, 19, 26 
 Clay, 494 
 
 Cleaning blast furnace gas, 23 
 Cleveland, England district, iron 
 
 ores, 10 
 Cobalt steels, 389 
 Coefficient of friction, 393 
 Cogging mill, 184 
 
 Coke, 9, 73, 452, 454r-464, 491, 492 
 Colby, Albert Ladd, 393, 514 
 Colby, E. A., 436 
 Cold short, 323 
 shut, 163, 240 
 work compared with hot work, 
 
 172, 210, 214-215 
 Collaring, 201 
 Colors, temper, 383 
 Combined carbon in iron, 315-337 
 Combining weights, 483 
 Combustion, 480 
 Comparative cupola practice, 265- 
 
 270 
 Comparison of purification processes, 
 
 148-155 
 of acid and basic steel 48, 109, 
 
 110, 153, 271 
 of Bessemer and open hearth, 
 
 151-153, 271, 272 
 of crucible with others, 153, 154, 
 
 433, 434 
 of wrought iron and steel, 154, 
 
 155 
 Compression of ingots, 163, 164 
 Compressive strength of steel, 307, 
 
 501 
 Concentrator, magnetic, 9 
 Conservation of energy, 500 
 Constitution of the steel, 299-314 
 Constituents of hardened and tem- 
 pered steels, 374-379 
 Continuous heating furnaces, 216- 
 
 218 
 
 Contraction. See Shrinkage. 
 Converting pots, 69 
 Cooling curve of pure iron, 302 
 curves, 283-285 
 rate of. See rate of cooling, 
 strains, see checking, 247 
 table, 206 
 Cope, 223, 226 
 Copper, cooling curve of 320 
 
 effect on steel, 311 
 Cores, 223, 224, 226, 232-238, 240, 
 
 246 
 Corrosion of iron and steel, the, 154, 
 394, 412-426 
 badly-made material, 154, 418 
 cast iron, 415; 416 
 cause and operation of, 412-419 
 manganese, 418 
 pitting, 417 
 
 preservative coatings, 419-426 
 relative, wrough iron vs. steel, 
 
 154, 394, 415-419 
 rust, 412, 414 
 segregation, 414 
 self-protection, 414, 415 
 theories of, 413, 414 
 Cort, Henry, 50 
 Country heat, 70 
 
 Critical range of steel, 291, 296, 312, 
 357-362, 369, 368-388, 395, 
 399 
 Critical temperature in Bessemer 
 
 process, 93, 94 
 Crucible furnaces, 71-76, 270, 390, 
 433, 434 
 steel, 46, 56, 70-76, 153, 154, 
 156, 176, 181, 308, 309 
 strength of, 308-311 
 Crystallization of iron and steel, 173, 
 247, 275, 300, 304, 307, 357, 
 374-379, 503 
 Cupola, 249-270 
 
 blast, 254-256, 266-269, 270 
 blower, 255 
 
 charge, 257, 258, 261-264, 270 
 chemistry of, 258-260 
 compared with air furnace, 
 
 347, 348 
 crucible zone of, 249, 250 
 
524 
 
 INDEX 
 
 Cupola, dimensions, 265-270 
 
 fuel, 257, 265 
 
 gases, 259, 260 
 
 limestone in, 258 
 
 lining of, 253 
 
 loss, 256 
 
 melting in, 257-260, 265-270 
 
 melting zone, 250, 251, 265 
 
 run, or campaign, 256, 257 
 
 stack, 254, 250 
 
 time of melting in, 251, 270, 
 265-270 
 
 tuyere zone, 250, 251, 254 
 
 ■ used for Blessemer process, 80 
 
 zones, 249 
 Cushman, Allerton S., 413, 516 
 Custer, E. A., 240, 511 
 
 Daelen, R. M., 197 
 
 Dalton and Gay Lussac law, 498, 
 
 499 
 Defects in ingots, 158-171 
 Definitions of iron and steel, 66 
 Dellwik-Fleischer gas, 473, 474 
 DeMozay, 516 
 Density, 449 ■ 
 
 of cast iron, 329, 330 
 Depth of chill in cast iron, 333-336 
 
 See also ChiUing. 
 Derr, Louis, 518 
 Dies. See Wire. 
 Direct castings, 249 
 Discharge holes of blast furnace, 19 
 Distinguishing between different 
 
 products, 155-157 
 Dolomite, 122, 495 
 Domnarfret furnace, 428-432 
 Dorman, W. H., 94 
 Double shear heat, 70 
 
 Bteel, 70 
 Draft in rolling, 207 
 
 in wire drawing, 211, 212 
 Drag, 223, 226 
 Drawing, of wire. See wire. 
 
 temper, 373 
 Driers in paint oils, 422 
 Drop-forging,, 176 
 Drying oil, 422 
 Drying ovens, 228 
 
 Dry-sand molds, 222, 228, 230-233 
 Ductility, defined, 173, 501 
 
 of metals, measure of, 173 
 of steel, 6, 307, 308 
 
 as affected by mechanical 
 
 work, 172 
 as affected by oxygen, 311 
 as affected by welding, 365 
 Dumas, L., 514 
 Duplex process, 144, 145 
 
 Effect of work on steel, rationale of, 
 
 173, 174 
 Elastic Hmit, defined, 500 
 
 of carbon and nickel steels, 391 
 of steel, compared with ultimate 
 
 strength, 173 
 of welded pieces, 365 
 when exceeded, 172 
 Elastic hmit, ratio, 391 
 Electric, conductivity of steel, 312, 
 406 
 furnaces, 422-441 
 motors in roUing mills, 198, 199- 
 
 201 
 steel, production of, 153, 427- 
 441 
 Electrolysis and corrosion, 413, 414 
 Electrolytic, iron, 440, 441 
 
 refining, 440, 441 
 Electrometallurgy of iron and steel, 
 
 the, 427-441 
 Electroplating, 423-425 
 Electro-thermic process, 428-441 
 Elements, 477 
 Elongation, percentage of, defined, 
 
 501 
 Enamehng, 426 
 Eutectic alloy. See eutectics. 
 Eutectics, 282, 285, 286 
 Eutectoid, 292, 293, 307 
 Expansibility of steels, 393 
 
 Fahrenheit, 482 
 
 converted to centigrade degrees, 
 
 504 
 Fatigue of steel, 391 
 Feeders on castings, 162. See also 
 
 risers. 
 
INDEX 
 
 525 
 
 Feldspar, 494 
 
 Ferric compounds, 490 
 
 oxide as a pigment^ 423 
 Ferrite, 294, 299-302, 307 
 Ferro-alloys, 8, 157 
 Ferro-chrome, 157 
 Ferromanganese, 8, 157 
 Fertosilicon, 8, 157 
 Ferro-titanium, 157, 169-171 
 Ferrous compounds, 490 
 
 metals, 3 
 Fettling of puddling furnaces, 58 
 
 of open hearth, 122 
 Fin, 201 
 Finery fire, 52 
 
 Finishing temperatures for rolUng, 
 363, 364. See also tem- 
 peratures. 
 
 for forging, 176 
 Flanging, 214 
 Flask, 223 
 Forge iron, 8, 57 
 
 Forging. See also hammering, and 
 pressing. 
 
 compared with roUing, 215 
 
 drop-, 176 
 
 finishing temperatures for, 176 
 of metals, the, 174^180 
 Fojsythe, Robert, 508 
 Foundry irons, analyses of, 8 
 
 ladles, 195 
 Foundry practice, 221 
 Fracture of steel, 357, 450. See also 
 
 crystallization. 
 Freezing. See also soUdification. 
 
 of iron and steel, 286-298 
 Freezing-point curves, 282, 283 
 Fritz, John, 183 
 Fuels, 451-474 
 
 Furnaces for annealing, hardening, 
 and tempering, 379-383 
 
 Gages, Leon, 506 
 
 Galvanizing, 423-425 
 
 Gamma iron, 300-302, 312-314 
 
 Ganister for converter lining, 82 
 
 Gas engines utilizing blast furnace 
 
 gas, 25 
 Gaseous fuels, 462-474 
 
 Gases in steel, 158 
 
 from baby Bessemer, 274 
 
 from cupola, 259, 260 
 Gas mains, 468 
 
 producer grate area, 465 
 
 producers, 464-467 
 Gate, 226, 240-245 
 Gated patterns, 244, 245 
 Gayley's air-dr3dng process, 32-34 
 Gay Lussac and Dalton law, 498 
 Girod furnace, 440 
 Gledhill, J. M., 515 
 Goerens, Paul, 518 
 Gold-silver alloys, 277, 278 
 Grading "cast steel by eye, 75 
 Grain of steel. See also crystalliza- 
 tion of steel. 
 Graphite, 315, 492 
 
 and corrosion, 415 
 
 and expansions. See graphite 
 and shrinkage. 
 
 and manganese, 326 
 
 and phosphorus, 326 
 
 and porosity, 321, 322 
 
 and shrinkage, 319-321, 327, 
 328 
 
 and silicon, 324 
 
 and strength, 322, 332, 333 
 
 and sulphur, 324, 325 
 Graphite and workabiUty, 322 
 
 as a pigment, 423 
 
 distinction under the micro- 
 scope, 446 
 
 for washes, 228 
 
 in cast irqn, 315, 316 
 
 in steel, 315 
 
 precipitation in malleable cast- 
 ings, 339 
 
 properties and structure of, 
 316 
 Green-sand molds, 222, 228, 230- 
 
 232 
 Guillaume, C. E., 515 
 Guillet, L,. 515, 516 
 
 Haanel, Eugene, 517 
 Hall, Joseph, 50 
 
 Hadfield, Sir Robert A., 397, 406, 
 515 
 
526 
 
 INDEX 
 
 Hambuechen, Carl, 517 
 Haminering. See also forging, 
 compared with rolling, 215 
 control of temperatures in, 175, 
 
 176 
 effect of, 174„ 175 
 Hammer refining, 361-365. See also 
 
 restoring. 
 Hanging of blast furnace charge, 37 
 Harbord, F. W., 509 
 Hardened steel, constituents of, 
 374r-379 
 uses of, 372 
 Hardening of steel, 368, 388 
 and magnetism, 312 
 theories of, 370-372 
 Hardenite, 375 
 
 Hardness. See also "workability (for 
 cast iron), 
 defined, 502, 503 
 of oast iron, 329 
 of cast iron and manganese, 323 
 of nickel steel, 392 
 of steel, 308, 311, 312 
 
 as affected by cold work, 172, 
 
 210, 214 
 as affected by forging temper- 
 atures, 174 
 due to carbon alone, 403 
 loss of in tempering, 372 
 loss of on annealing, 369, 379 
 Hardness produced by quenching, 
 
 181, 368-388 
 Harmet's hquid compression, 164 
 Head, or header. See riser. 
 Hearth of blast furnace, 18 
 
 of cupola. See crucible zone. 
 Heat, definition of, 499 
 energy, 488 
 from chemical change, 475, 487 
 
 See thermo-chemistry. 
 in roUing, 206 
 in Bessemer process, 48 
 tinting, 447 
 
 to start chemical change, 476, 
 478, 481-483, 487, 488, 
 491, 492 
 treatment of cast iron, 357 
 treatment of steel, the, 357-388 
 
 Heat, units, 482, 483 
 
 Heating for roUing, etc., 215-220 
 
 furnaces, 215-220 
 
 of steel, improper, 357-368 
 Helve hammers, 174 
 Hematite, 9 
 H^roult process, 432 
 
 steel process, 432, 438, 439, 440 
 Heyl and Patterson pig-casting ma- 
 chine, 36 
 Heyn, E., 518 
 
 High-carbon steel, roUing of, 220 
 High-speed steels, 400-405 
 Hoerhager, J., 516 
 Hoesch process, 148 
 Hollow wire. See tubes. 
 Hopper of blast furnace, 17 
 Horns on manipulator, 190 
 Hot-blast for blast furnace, 21 
 
 stoves, 21, 24 
 Hot spots in cast iron, 330 
 Hot work compared with cold work, 
 
 214 
 Housing cap, 191, 193 
 Housings, 190, 193 
 Howe, Henry M., 7, 162, 167, 505, 
 
 509, 510, 515, 516 
 HydrauUc presses compared with 
 
 hammers, 212, 213 
 Hydro-carbons, 467, 486, 487 
 Hydrogen and corrosion, 413 
 Hydrogen, chemistry of, 486-488 
 
 in iron, 440, 493 
 
 in steel, 158, 307 ' 
 
 Illuminating gas, 486 
 
 Impact, 501 
 
 Impurities in cast iron, effect of, 
 
 323 
 Incomplete combustion, 457 
 IndestructibiHty of matter, 485 
 Induction furnaces, 436-439 
 Ingotismy 158, 164^166, 367 
 Ingot molds, 75, 89, 90 
 Ingots of large size, form of, 179 
 
 casting of, 162 
 
 compression of, 162, 163 
 
 heating of, 179 
 
 solidification of, 161-164 
 
INDEX 
 
 527 
 
 Ingots, taper of, 207 
 
 texture of center of, 179, 330 
 Inorganic chemistry, 491, 492 
 Iodine etching, 446 
 Invar, 393 
 
 Invention of puddling, 50 
 Ionic hydrogen and corrosion, 413 
 Iron, See also cast iron, pig iron, 
 wrought iron, malleable 
 cast iron, malleable iron, 
 abundance of, 4 
 and carbon, 6 
 castings, 316 
 chemistry of, 492 
 cupola. See cupola, 
 density of, 329 
 disposal, 34 
 foundries, use of Bessemer in, 
 
 273 
 hardness of, 370 
 occurrence of in earth, 4 
 ore as pigments, 422 
 ores described, 9 
 distribution of, 9, 11 
 deposits, summary of in U. S., 
 
 10 
 mining of, 14 
 prices, 15, 157 
 transportation of, 12, 14 
 unloading of, 14 
 oxide. See also oxygen, 
 specific gravity of, 316 
 Iron, strength of, 307 
 
 sulphide, 168 
 Irregularities in blast-furnace work- 
 ing, 37-40 
 Irreversible transformations, 395 
 
 Jail bars, 181 
 
 Jobbing foundries, use of scrap in, 
 
 264 
 Johnson, Jr., J. E., 39, 159, 335 
 Jiiptner, Saans Freiherr von, 395 
 
 Keep, William J., 511 
 Kershaw, John B. C, 517 
 Killing in crucible process, 74 
 Kjellin electric furnace, 436 
 Kjellin, F. A., 442 
 
 Knobbled charcoal iron, 53 
 
 iron, 54 
 Knobbling fire, 53 
 Krupp purification process, 52-54 
 
 Ladles, foundry, 252 
 
 Lake, E. F., 514 
 
 Lake Superior iron ores, 12 
 
 Lancashire process, 54 
 
 Lap-welded tubing, making of, 209 
 
 Lathe tools, tempering of, 372 
 
 Layers of iron and coke in cupola, 
 
 251, 255, 266-269 
 Le Chatelier, H., 518 
 Le Chatelier microscope, 448 
 Lead, 480, 481. See also paints. 
 
 and zinc, 497 
 
 coatings, 424 
 
 in terne plate, 425 
 Lead-tin alloys, 278-286 
 Ledebur, A., 505 
 Lewkowitsch, 517 
 Lifters, 227 
 Lifting screw, 223 
 Lignite, 452, 453 
 LiUenberg, N., 510, 511 
 Lihenberg's Hquid compression, 164 
 Lime. See blast furnace, open 
 
 hearth, cupola. 
 Limestone, 9. See also blastfurnace, 
 
 open hearth, cupola. 
 Linseed oil, 422 
 Liquid cast iron, density of, 329 
 Liquid compression of ingots, 163,164 
 Loading a blast furnace, 16 
 
 an ore boat, 14 
 Loam molding, 222-224 
 Longmuir, P., 513 
 Long-tuyere converters, 273, 274 
 Loose texture in center of castings. 
 
 See also Ingots. 
 Loss, in baby Bessemer converters, 
 272 
 
 in Bessemer process, 79, 99, 102, 
 274 
 
 in crucible process, 75, 76 
 
 in cupola, 249, 256 
 
 in puddling process, 66 
 
 in rolling plates, 201 
 
528 
 
 INDEX 
 
 Lorraine iron ore, 9 
 Lothringen iron ore, 9 
 Low-carbon steel in open hearth, 
 making of, 132 
 steels, ferrite in, 299 
 Luminosity of flames, 467, 468 
 Luxemburg iron ore, 9 
 
 Mcllhiney, Parker C, 230 
 McQuillan, W. S., 279. 
 Magnesia, 41, 42, 495 
 Magnesite, 122 
 Magnesium, 41 
 
 chemistry of, 495 
 Magnetism of iron and steel, 4, 312- 
 
 314, 373 
 Magnetite, 9 
 Magnet steels, 405, 406 
 Magroscopic metallography, 450 
 Malleable, cast iron, 3, 338-356 
 
 all-black castings, 340 
 
 annealing of, 339, 349-352 
 
 as steel castings, 353, 354 
 
 contraction of, 352 
 
 definition of, 7 
 
 description of, 4, 338 
 
 expansion of, 352, 353 
 
 manganese in, 342 
 
 melting of, in cupola, 249 
 
 microphotograph of, 339 
 
 molding for, 349 
 
 phosphorus in, 342 
 
 pig iron used, 340, 341 
 
 pouring of, 349 
 
 process, 339 
 
 properties of, 338-340 
 
 shrinkage of, 352 
 
 silicon in, 341 
 
 strength of, 338 
 
 sulphur in, 342 
 
 test lug, 338 
 
 theory of, 355, 356 
 
 total carbon in, 342 
 
 uses of, 338, 339 
 
 See also Air Fiimace. 
 Malleability, 502 
 Mandril, 208 
 Manganese, chemistry of, 494 
 
 as reducing agent, 491 
 
 Manganese, and sulphur, 323, 326 
 in acid open hearth, 133, 139 
 in basic process, 128, 389 
 in Bessemer process, 79, 84, 96, 
 
 389 
 in blast furnace, 32, 44 
 in cast iron, 309, 317, 323, 324, 
 326, 327, 329, 330, 331, 
 332, 333, 334, 350, 361 
 See also in steel, 
 in cupola, 261 
 
 in steel, 47, 57, 159, 166, 303, 
 304, 305, 309, 310, 312, 
 388, 397-399, 418 
 in wrought iron, 50 
 oxides, 490 
 
 salts as paint driers, 422 
 steel, 329, 397-399 
 composition of, 398 
 critical changes, 399 
 treatment, 398, 399 
 uses, 399 
 sulphide, 168, 303, 305, 324, 
 330 
 Manganiferous cementite, 303 
 Manipulators, 190, 196 
 Manufacture. See Pig iron. Wrought 
 
 iron. Steel, etc. 
 Marble, 492 
 
 Martensite, 375, 376, 378 
 Martin furnace. See Open hearth 
 Match for molding, 236 
 Maximum affinity, 481 
 Mechanical mixtures, 476 
 
 pig molding machines, 35 
 puddling furnace, 51, 58, 59, 60, 
 
 61 
 treatment, effect on crystalliza- 
 tion, 358-362 
 treatment of steel, the, 172-220, 
 work, its effect on strength, etc., 
 172, 210, 214, 215 
 Mellor, J. W., 514 
 Melting, fineries, 53 
 heat, 70 
 holes, 71 
 
 iron. See also Cupola, Air fur- 
 nace, Open-hearth furnace 
 for iron. 
 
INDEX 
 
 529 
 
 Melting-point of cast iron. See 
 
 Fluidity of cast iron, 
 steel for castings, 270-274 
 Melting zone of cupola, 249, 251, 253, 
 
 254, 255, 265-269 
 Merchant bar, 52 
 Mesabi iron ore, 14, 39 
 Metallic and non-metallic elements, 
 
 488 
 luster, 488 
 Metallography of iron and steel, the, 
 
 442-450 
 Metallurgy, definition of, 478 
 Metcalf test, 366, 367 
 Metcalf, William, 379, 380 
 Methane, 486 
 Mica schist for converter lining, 
 
 82, 83 
 Micro-constituents of steel, the, 299- 
 
 314 
 Micro-photographic apparatus, 447- 
 
 449 
 Mill iron, 8, 57 
 Mill scale. See also Scale. 
 Mineral oil. See Petroleum and Oil 
 
 for Fuel. 
 Minette iron ore, 9 
 Mining ore at Lake Superior, 12 
 Miscellaneous purification processes, 
 
 52-56 
 Mixed crystals, 277 
 Mixer, 79, 80, 86, 145 
 Mixtures, 279. See Mechanical mix- 
 tures. 
 Modulus of elasticity, 500, 501 
 Moisture producing corrosion, 412, 
 
 419 
 Mold cars for car-casting process, 90, 
 
 91, 92 
 Molderke, Richard, 511, 513 
 Molders' tools, 227 
 Molding machines, 222, 236-245, 
 
 245-247 
 sand, 228 
 Molds, making of, 221-247. See 
 
 also Ingot molds. 
 Molecules, 485 
 Molybdenum steel, 404 
 Mond gas, 474 
 34 
 
 Monell, Ambrose, 146 
 Monell process, 146, 147 
 Monkey of blast furnace, 19 
 Mono-, prefix, 490 
 Monosilicate, 493 
 Mother metals, 284 
 
 liquors, 284 
 Mottled, cast iron, 7, 329 
 
 pig iron, definition of, 6 
 Muck bar, 52 
 
 Mueller, Friedrich, C. G., 510 
 Multiple, molds, 246, 247 
 
 proportions of atomic weights, 
 484 
 Multiple-ply plate, 181 
 Mushet, Robert, 401 
 Mushet steel, 401 
 
 Nailing molds, 225 
 
 Natural gas, 468-470, 486 
 
 Necking of steel, 501 
 
 Net heat of chemical reactions, 
 481, 482 
 
 Ifeutral furnace linings, 494 
 
 Nicolardot, P., 516 
 
 Nicolson, J. T., 515 
 
 Nickel, effect o n crystallization, 
 361 
 effect on electric conductivity, 
 312 
 
 Nickel-plating, 214, 425 
 
 steel, 188, 390-397, 389-411 
 chrome-nickel steels, 400 
 critical changes, 395 
 crystalline structure, 392, 
 
 396-397 
 expansibility, 393 
 hardness, 392, 393 
 irreversible transforma- 
 tions, 395-397 
 modulus of elasticity, 392 
 microphotographs, 396, 397 
 soundness, 393 
 tensile properties, 391, 392 
 uses of, 390, 391 
 vaxadium nickel, 408, 411 
 
 Nitric acid etching for metallog- 
 raphy, 446 
 
 Nitrogen in the blast furnace, 26 
 
530 
 
 INDEX 
 
 Nitrogen, in steel, 171, 307, 367, 
 
 407,411, 419 
 Noble electric furnace, 432 
 Noble, H., 509 
 Non-magnetic iron or steel, 300, 
 
 397, 399, 401, 402 
 Northampton cast iron, cooling 
 
 curve of, 320 
 Norway iron, 300 
 
 iron for magnets, 312 
 Nozzles of steel ladles, 88, 89 
 Number of Bessemer converters in 
 America, 45, 46 
 blast furnaces in America, 45 
 open-hearth furnaces in Amer- 
 ica, 45 
 puddling furnaces in America, 
 45, 50 
 
 Occluded oxidized substances, 
 
 168-171 
 OU for fuel, 470, 472 
 Oolitic hematite, 7 
 Opal, 493 
 Open grain in cast iron, 166, 322, 
 
 329, 330 
 Open-hearth bath, 121 
 
 boil, 134 
 
 bottom, 121-123 
 
 Campbell type, 124-127 
 
 casting pit, 114 
 
 charging, 129, 131 
 
 charging boxes, 113, 123 
 
 charging machines, 111, 113, 
 114, 123 
 
 checkers, see regenerators 
 
 chimney, 119, 120 
 
 cycles. 111 
 
 dirt pockets, 115 
 
 distinguished from others, 155, 
 157 
 
 draught, 119, 120 
 
 fluxes, 132 
 
 for malleable castings, 348, 349 
 
 for melting iron, 348, 349 
 
 fuels', 111, 112, 127, 142, 462- 
 474 
 
 fuels, amount used, 127, 143 
 
 furnace, 49, 114-128, 436-441 
 
 Open-hearth, 121-123 
 
 house, 112, 123, 125, 137 
 
 ingot molds, 112, 141 
 
 ingots, weight of, 183, 201 
 
 life, 121 
 
 lime in, 50, 132, 134 
 
 hning, 50, 121, 122, 124, 126, 148 
 
 melting platform, 111, 113 
 
 molten pig in, 129, 143, 144 
 
 operation of, 49 
 
 plant, 111-114 
 
 ports, 115, 118, 119, 121 
 
 process, 49, 50, 111-157 
 
 regenerators. 111, 114-117, 121, 
 
 125 
 repairs, 123, 125 
 reversals, 127 
 reversing valves, 117, 127 
 roof, 119, 120, 121, 122 
 scrap used in, 47, 129 
 size of, 114, 121, 124, 128 
 slag pockets, 115 
 stationary, 49, 114, 122, 123, 124 
 stock, 114 
 tap-hole, 123, 124 
 temperature, 117, 119, 127, 143 
 tilting, 124-127, 145, 147 
 valves, 117 
 
 Wellman type, 124^127 
 Open-hearth process, the, 49, 111-157 
 acid, chemistry of, 138-143 
 
 loss, 142 
 
 practice, 49, 50, 138-143, 270- 
 272 
 
 recarburizing, 108, 111, 136- 
 139, 170 
 aU-scrap process, 131. See also 
 
 Acid steel 
 basic, chemistry of, 132-138 
 
 fluxes, 132 
 
 loss, 135-136 
 
 practice, 50, 134, 135-138, 
 270-272 
 
 recarburizing, 136-138, 170 
 
 removal of impurities, 132-138 
 See also Basic steel. 
 Bertrand Thiel process, 147, 148 
 boQ in, 143 
 Campbell process, 147-148 
 
INDEX 
 
 531 
 
 Open-hearth, charging, 129, 131 
 Duplex process, 144, 145 
 Hoesch process, 148 
 making low carbon steel, 127, 
 
 135-138, 138-141 
 Monell process, 144, 146, 147 
 ore used in, 134, 135, 143-148 
 oxidation in, 143 
 pig-and-ore process, 128, 129- 
 
 131, 143-148 
 pig-and-scrap process, 128 
 pig iron used, 8, 131 
 recarburizing, 111 
 rephosphorization, 136 
 
 slag, 122, 123, 124, 130, 131, 
 
 132, 133, 134, 135, 136, 138, 
 141, 142, 143, 146, 147 
 
 special processes, 143-148 
 steel distinguished from Bes- 
 semer, 155-157, 271, 172 
 Talbot process, 144, 145, 146 
 teeming, 131, 142, 145, 151, 154 
 Open-hearth steel in U. S., 45 
 
 not often very low carbon, 
 
 134 
 strength of, 308-311 
 teeming, 138 
 Open pass, 186 
 
 Operation of regenerative furnace, 49 
 Ore boats, 14, 15. See also Blast 
 Furnace for ores, 
 handling mechanism at blast 
 
 furnace, 16 
 price of, 15, 157 
 used in open hearth, 50, 132 
 Organic acids and corrosion, 425 
 
 chemistry, 491 
 Osmond, F., 512, 518 
 Osmoiidite, 377, 378 
 Osmond's polish attack, 445, 446, 
 447 
 theory of permanent magnetism, 
 313, 314 
 Osmotic pressure, 497 
 Otto Hoffmann coke oven, 457. See 
 
 Retort coke oven. 
 Overheating of steel, 209, 219, 357- 
 
 368, 376 
 Overman, Frederick, 509 
 
 Overfilling the pass, 186 
 Oxidation by and in paints, 421, 423 
 definition of, 480 
 of cast iron, 159 
 in cupola, 249. See Loss in cu- 
 pola, 
 in puddling, 59 
 Oxidation, relative, of carbon and 
 
 phosphorus, 59 
 Oxide of iron. See Oxygen. 
 Oxidized coatings, 425, 426. See 
 
 Scale. 
 Oxidizing agents, 490, 491 
 Oxygen, chemistry of, 479, 480 
 in blast furnace, 26 
 in cast iron, 159, 323 
 in steel, 47, 158, 159, 168-171, 
 271,306, 311,407,414,419 
 
 Packing for malleable castings, 350- 
 
 352 
 Paint and Painting, 154, 419-423 
 
 kinds of paint used, 421 
 
 linseed oil, 422 
 
 pickHng, 420 
 
 pigments, 422, 423 
 
 preparation of surfaces, 419 
 
 priming coat, 420 
 
 shop vs. field painting, 420, 421 
 Pass in rolling, 181 
 Pass-over mill, 181 
 Pasty condition in freezing, 321 
 Pasty stage of soUdification and 
 
 phosphorus, 321, 326 
 Pattern molding, 223-247 
 Patterns, design of, 247-249 
 Pavlor, M. A., 508, 510 
 Pearhte, 294, 300, 304, 358, 374, 375- 
 
 377, 378, 446, 447 
 Peat, 452, 453 
 Pellets in Bessemer slags, 99, 100 
 
 of iron in foundry practice, 256 
 Percy, John, 73 
 Perfect combustion, 480 
 Permanent molds, 238 
 
 set defined, 500 
 Petroleum. See Oil. 
 Pliillips, J. Arthur, 94 
 PhiUips, W. B., 126 
 
532 
 
 INDEX 
 
 Phlogiston theory, 479 
 
 Phosphide of iron, 304, 305, 306, 
 
 323, 447 
 Phosphorus, chemistry of, 495 
 
 eutectic, 304, 305, 306, 323, 447 
 in basic open-hearth process, 
 50, 51, 131, 132-138, 147, 
 498. 
 in blast furnace, 32, 44 
 in electric smelting, 434, 435 
 in iron and steel, 57, 166-168, 
 305, 306, 309, 310, 311, 312, 
 317, 323, 326, 327, 329, 330, 
 331, 333, 336, 342 
 in ores, 16 
 
 in puddling, 50, 51, 59 
 Photo-micrographic apparatus, 448, 
 Purification of pig iron, 45 
 Physical changes, 475 
 
 properties of metals, 500-504 
 Physics, introductory to metallurgy, 
 498-505 
 principles of, 498-505 
 used in crucible process, 73 
 Pickhng, 214, 420, 421, 424 
 Picric acid etching, 447 
 Pig beds of blast furnace, 37, 38 
 Pigging up in basic open hearth, 132 
 Pig iron, 3, 4, 5, 6, 7, 8-44. See also 
 Cast iron, 
 analyses of, 81 
 ladles, 35 
 price of, 157 
 Pigment in a paint, 422, 423 
 Pig-molding machines, 35, 36 
 Pigs, 35 
 " Pig-washing " process, or Bell- 
 
 Krupp, 52 
 Piling, muck bar, 51, 68 
 
 wrought-iron scrap, 68, 418 
 Pipes in iron and steel, 158, 161-164, 
 
 235, 248, 249 
 Pipe-welding rolls, 207, 208 
 Pitch as a pigment, 423 
 Pitting, 417 
 Platmite, 394 
 Polish attack, 447 
 
 Polishing for metallography, 445- 
 447 
 
 Porosity of oast iron, 321, 322, 329, 
 
 330 
 Porter, J. J., 513 
 Potassium bichromate and corrosion, 
 
 414 
 Power, from blast-fumace gas, 23 
 
 consumed in rolling, 188, 189 
 Precipitation, 497 
 
 Preparation, of samples for metal- 
 lography, 442-i45, 452-459 
 
 of surfaces for coating, 419 
 Preservative coatings. See Paint, 
 
 Galvanizing, etc. 
 Pressing, 212-215 
 
 Pressure on steel, methq/Js of apply- 
 ing, 172, 174 
 
 and volume of gases, 498 
 Prices of ores, 15, 16, 157 
 
 iron, steel, etc., 157 
 Priming coat, 420 
 Prochaska, Ernest, 126 
 Producer-gas, 462-468 
 Production of steel, etc. in U. S., 46 
 Projectiles, 372 
 
 Properties of east iron, the, 327-337 
 Puddle baUs, 51 
 Puddled bar, 52, 65, 68 
 
 iron, 54 
 Puddle rolls, 65 
 Puddlmg, 50-52, 57-68, 143 
 
 operation, 58-63 
 
 boil, 59, 60, 61, 65 
 
 chemistry, 59, 63-65 
 
 cinder, 50, 58, 59, 61, 66, 67 
 
 come to nature, 63 
 
 clearing, 65, 59 
 
 fuel, 57, 65, 66 
 
 furnace, 50, 57, 58, 67, 68 
 
 labor, 66 
 
 loss in, 57, 66 
 
 mechanical furnace, 67-68 
 
 rabbling, 61 
 
 slag, 50, 58, 59, 61, 66-67 
 
 squeezer, 63 
 Pulling in crucible process, 74, 75 
 Pull-over mill, 181 
 Purification of pig iron, 45, 56 
 
 Qualitative chemistry, 478 
 
INDEX 
 
 533 
 
 Quantitative chemistry, 478 
 Quarternary steels, 390, 400-411 
 Quenching steel. See Hardening 
 steel, 
 baths, 382 
 
 Rabbling of puddling process, 60, 61 
 
 Radicals, 488-491 
 
 Ragging. See Roughing. 
 
 Railroad car wheels. See Car wheels. 
 
 Rammers, molders', 227 
 
 Ramming molds, 226-228 
 
 Rate of cooling, effect on iron, 297, 
 318, 327, 333-337 
 cooling of steel, 369, 370, 369- 
 373 
 
 Read, A. A., 332 
 
 Reaumur, 339, 340 
 
 Recalescence, 381 
 
 Recarburizing, 48, 73. See also 
 Bessemer Process ; Open- 
 hearth process, etc. 
 
 Recuperative heating furnaces, 216- 
 218 
 
 Regenerative furnaces, 49, 71, 114- 
 128, 215-220. See Open- 
 hearth. 
 
 Red lead, 423 
 
 Red-shortness of steel, 309 
 
 Reducing agents, 491 
 
 Reduction, definition of, 480 
 of area defined, 501 
 
 of steel Tinder strain, 172 
 of metals in rolls, 180-209 
 
 Reel, 212 
 
 Refinery hearth, 54 
 
 Refining steel, 358-364. , See also 
 Crystallization. 
 
 Refining in electric furnaces, 432—440 
 
 Refractory clays, 494 
 
 Regenerative furnace. See also Cru- 
 cible process; Open-hearth 
 furnace, 
 heating furnaces, 215-217 
 
 Repairs for steel-casting fiu-naces, 
 273 
 
 Rephosphorization of steel, 136-138, 
 150, 151, 502. See Open- 
 hearth. 
 
 Residual silicon in Bessemer process, 
 
 94 
 Resilience, 502 
 
 of malleable cast ironi 338 
 Restoring steel, 358-366. See also 
 
 Crystallization. 
 Retort coke oven, 465-464 
 Return scrap, 341 
 Reverberatory furnaces, for heating, 
 
 215-220 
 Reversing mills, 183, 184, 
 Richards, J. W., 506 
 Risers on castings, 162, 226, 248- 
 
 249, 330 
 Roberts-Austen, Sir William, 312, 
 
 314, 315 
 Roberts-Austen, Roozeboom dia- 
 gram, 295, 297, 293-297, 
 286-297 
 Rock-over molding machine, 241 
 Rod mill, 189 
 
 Rods, wire. See Wire rods. 
 Roechling-Rodenhauser furnace, 
 
 438, 439 
 Roe puddling furnace, 67-68 
 Rolled steel, crystallization of, 368 
 Roll engines, 196-199 
 Roller table engine, 190, 194, 198 
 Rolling, 180-212 
 
 action in, 362 ,363 
 compared "with hammering, 180 
 effect of, 180, 181, 362-365 
 miUs, 180-201 
 delays in, 201 
 parts of, 186-201 
 plates, 201-206 
 practice, 201-209 
 rails, 206-9 
 speed of, 183, 189 
 temperatures of, 361-364 
 Roll scale used for fettling, 58 
 Rolls, 186-191 
 
 chilled, 187, 188, 318, 334, 336 
 Roll tables, 193-196 
 Roozenboom. See Bakhuis. 
 Roozeboom diagram, 286-297, 295, 
 
 297, 293-297 
 Rossi, A. J., 516 
 Rotary squeezer, 63 
 
534 
 
 INDEX 
 
 Rothenberg coke ovens, 461 
 
 Rouge for polishing, preparation of, 
 
 444-445 
 Running-out fire, 52 
 Rust, See Corrosion. 
 Rusting of iron and steel. See also 
 
 Corrosion. 
 Ryland's Directory, 8 
 
 Salts, 488, 489 
 
 Sand-casting of pig iron, 36, 37 
 Saniter, E. H., 149 
 Sappy. See Silky. 
 Sauveur Albert, 379, 512 
 Sauveur method of etching, 446 
 Saws. See Wood saws; Hacksaws. 
 Scabs on castings, 226 
 Scaffolding of blast furnace, 38 
 Scholl, George P., 517 
 Scintillating of steel, 366 
 Scott, W..G., 513 
 Scheme of iron manufacture, 46 
 Scrap produced in forging cannon, 
 179 
 
 used in iron castings, 47, 256, 
 257, 262, 334, 341 
 
 used in steel manufacture, 47, 
 73, 87, 129, 139, 145 
 Seamless tubes, making of, 209 
 Second-class rails, 207-209 
 Segregation, 158, 166-168, 171, 330, 
 
 331, 414 
 Self-fluxing in blast furnace, 41 
 Self-hardening steels, 400-405 
 Semet-Solvay coke oven, 459 
 Semi-steel castings, 353 
 Sesqui-oxide of iron, 491 
 Sesquisilicate, 493 
 Shearing strength, 501, 502 
 Shear steel, 70 
 SheUing of tires, 169 
 Shingling, 63 
 
 Shop vs. field painting, 420 
 Shovel loading soft ore, 13 
 Shrinkage, 247 
 
 cavity. See Pipe. See Grsr- 
 phite. 
 Shrinking of molds in drjdng, 213 
 
 outer cannon tubes, 180 
 
 Siderite, 10 
 
 Siemens, Martin. See Open-hearth. 
 
 Silica, 493 
 
 in slags, 31, 42, 50, 67, 96, 130, 
 
 133, 135, 149 
 wash, 228 
 Silicates, 493, 494 
 Silicide of iron, 306, 324 
 Silicon, chemistry of, 493 
 
 control of, in pig iron, 29-32, 39, 
 
 323" 
 in air furnace, 341 
 in basic open-hearth process, 
 
 131 
 in Bessemer process, 47, 48, 79, 
 
 93-102, 105 
 in cast iron, 41, 295, 317, 323, 
 
 324, 327-336, 497 
 in crucible process, 74 
 in malleable cast iron, 341, 340 
 steel, 48, 156, 159, 309, 312. See 
 
 also Silicon in cast iron, 
 steel, 312, 389, 406, 407 
 Silico-Spiegel, analysis of, 8 
 Silky fracture, 357 
 Silver-gold. See Gold-silver. 
 Single-shear heat, 70 
 
 steel, 70 
 Size of castings and shrinkage, 328 
 of crystals. See Crystalliza- 
 tion. 
 Skin-dried molds, 230 
 Slag and corrosion, 414-418. See 
 also Cinder, 
 distinction under microscope, 
 
 446 
 in steel, 156, 168 
 in wrought iron, 4, 51, 66, 156. 
 See Puddling. 
 Slags, 480, 491, 493-499, 494, 495, 
 498. 
 
 See also Cinder. 
 Bessemer, 79, 84, 93, 96, 99,- 
 
 100 
 heating furnace, 217-219 
 open hearth, 122, 124, 130-138, 
 140-143, 146, 147 
 Slip bands, 173, 392 
 Slips in blast furnace, 37-39 
 
INDEX 
 
 535 
 
 Smelting in U. S., distribution of, 11 
 
 law of. See law of smelting. 
 
 zone of blast furnace, 24 
 Smith, J. Kent, 516 
 Soaking pits, 92, 217, 218 
 Solidification, 164, 286-290 
 Solid solutions, ■275"278, 290-298, 
 374-379 
 
 of iron, 290-293 
 Solubility, 496, 497 
 Solutions, 275-298 
 Sorbite, 375-377, 378 
 Soundness, 231, 309 
 Sows, 37 
 
 Spathic iron ore, 10 
 Specific gravity, 172, 499 
 Spectroscope in Bessemer prpcess, 88 
 Speed of cooling. See Rate of cooling 
 Speller, F. N., 517 
 Spiegeleisen, 8, 85, 86, 136 
 Spitting in Bessemer process, 79, 99, 
 
 102, 274 
 Spongy spots in cast iron. See Cast 
 
 iron. 
 Spring heat, 70 
 Springiness. See Resilience. 
 Sprues, 237 
 Squeezers, 63, 245 
 Stack of blast furnace, 18 
 
 of cupola, 249, 250, 254, 265-270 
 Stansfield, Alfred, 378, 379, 395, 514 
 Stead, J. E., 167, 323, 332, 382, 457, 
 
 512, 518 
 Stead's brittleness, 368 
 Steam hammers, 174-180, 215 
 
 compared with electric motor 
 drives for rolls, 199-201 
 Steel castings, 155, 231, 232, 270- 
 
 274 
 Steel conversion process, 68-70 
 Steel-converting furnace, 69 
 Steel, definition of, 7 
 
 description of, 4 
 
 ladles, 88, 89 
 
 rolls, 187, 188 
 
 shrinkage of, 247 
 Steel, strength of, 67, 172, 214, 215, 
 ^ 299, 307-311, 331, 377 
 
 through heat, 70 
 
 Stools for ingot molds, 89-90 
 Stoppers of steel ladles, 88, 89 
 Stoughton converters, 273, 274 
 Stoves, blast furnace, 21-24 
 Strain, effect of, on steel, 173 
 Strength of welded pieces, 365 
 Strength of steel. See Steel. 
 Stress defined, 500 
 Stripping ingots, 90, 91 
 Structure of eutectics. See Eutec- 
 
 tics. 
 Sub-, prefix, 490 
 Subsilicate, 493 
 Sulphide of iron, 303-305, 324r-326, 
 
 330. See Iron sulphide, 
 of manganese. See Manganese 
 
 sulphide. 
 Sulphur, chemistry of, 494-495 
 elimination in mixer, 80 
 in blast furnace, 29, 30, 32, 
 
 37-39 
 in cast iron, 166, 305, 323, 
 
 324-326, 327, 329, 331, 334 
 in steel, 50, 305, 309. See 
 
 also Sulphur in cast iron. 
 Super-refining of steel, 170, 427, 428, 
 
 433-436 
 Swank, James M., 506 
 Swedish iron, 300, 312. See Nor- 
 way iron. 
 Swedish Lanchashire process, 54, 55 
 Sweeping a mold, 221, 222, 223 
 Swelling of a casting, 226, 
 Synthesis, 478 
 
 Talbot, Benjamin, 167, 510 
 
 Talbot process, 144, 145, 146 
 
 Tap cinder, 67 
 
 Tappings, 67 
 
 Tar as a paint, 423 
 
 Taylor, Fredrick W., 402 
 
 Temper, carbon, 339 
 
 colors, 383, 384 
 Temperature and volume of gases, 
 498 
 annealing malleable castings, 46, 
 339-356 
 steel, 358-360 
 defined, 482 
 
536 
 
 INDEX 
 
 Temperature, drying ovens, 228 
 
 expanding cannon tubes, 180 
 
 hardening steel, 368, 369 
 
 rolling, 361-364 
 
 welding, 364, 365 
 Temperatures, annealing, 358-360 
 
 finishing in rolling and hammer- 
 ing, 361-364 
 
 to produce ingotism, 367 
 
 to produce Stead's brittleness, 
 368 
 
 restoring steel. See Crystal- 
 lization; Restoring. 
 
 tempering, 383-384 
 Tempered steel, constituents of, 
 
 374-379 
 Tempering, 372, 373, 405 
 
 fire clay crucibles, 72 
 
 of steel, 372, 373, 379-384 
 Tensile strength, defined, 500 
 Test lug, 338 
 Ternary alloys, 389 
 Terne plate, 425 
 Texture of center of ingots. See 
 
 Ingots. 
 Thermo-chemistry, 481-483 
 Three-high mill, 183, 
 Tiemann, Hugh P., 506 
 Tin, 425 
 
 lead. See Lead-tin, 
 
 plate, 425 
 Titanium, 9, 169-171 
 Tool steels, 400-405 
 Torsion, .502 
 Toughness, defined, 502 
 Transverse strength, 333, 501 
 Tri-, prefix, 490 
 Trisilicate, 493 
 Tri-valent, 473, 490 
 Troostite, 375, 376, 378 
 Tropenas converters, 272, 273 
 Troubles in rolling, 201 
 Truran, W., 509 
 Tubing, 209 
 Tumbling barrel, 349 
 Tungsten steel, 400-406 
 Turner, Thomas, 291, 338, 347, 508, 
 
 511 
 Tuyere notches, 19 
 
 Tuyeres, of Bessemer converter, 81- 
 85 
 of blast furnace, 18, 19 ^ 
 
 Uehling pig-casting machine, 36 
 United States iron ores, 9-16 
 Uni-valent, 490 
 Universal mill, 186 
 Upton, G. B., 512 
 Upton's diagram, 297, 298 
 Uses of pig iron, 45, 46 
 
 Valence, 489, 490 
 
 Vathaire, A. de, 508 
 
 Vanadium steel, 171, 407-411 
 
 Vehicle, of paints, 422 
 
 Venting molds, 225, 234 
 
 Vertical rolls. See Universal mill. 
 
 Vibration and crystallization of 
 
 steel, 361 
 von Jiiptner, H. F., 514 
 von Maltitz, E., 511 
 
 Walloon charcoal hearth, 52-56 
 "Washed metal," produced by Bell- 
 
 Krupp, 52 
 Washes for molds, 225, 228-231 
 Waste, See also Loss. 
 
 gas from blast furnace, 24 
 Waterhouse, G. B., 171,332,402,403, 
 
 404, 419, 514 
 Water cracks in steel, 381 
 
 gas, 469, 472, 473 
 
 gas-producer, 473 
 
 gates, 248 
 
 ^sealed gas-producers, 466, 
 467 
 
 -sealed reversing valves, 469 
 Webster, W. R., 326 
 Wedding, Hermann, 507, 510 
 Weight, and volume of air compared, 
 
 21 
 Welding, 182, 364, 365 
 West, Thomas D., 511 
 White heart malleable castings, 339, 
 
 340, 356 
 White, Maunsel, 402 
 Whitworth's liquid compression, 164 
 Wind in cupola. See Blast pressure. 
 
INDEX 
 
 537 
 
 Wire, 210-212 
 
 bench, 211 
 
 dies, 211 
 Wood as feul, 451, 452 
 Woodworth, Joseph V., 513 
 Work, effect of, on steel, 172, 174, 210 
 Wrought iron, as malleable iron, 354 
 
 carburization of, 68-77 
 
 compared with low-carbon steel, 
 154, 155 
 
 corrosion of, vs, steel, 154, 416- 
 419 
 
 definition of, 7 
 
 description of, 4 
 
 distinguished from steel, 155, 156 
 
 ferrite in, 300 
 
 from scrap, 68, 154, 418 
 
 Wrought iron, heat treatment of, 
 368 
 manufacture of, 50-52, 57-68 
 micro-photograph of, 66 
 modulus of elasticity of, 500 
 not hardened by quenching, 182 
 properties of, 57 
 slip bands in, 173 
 tensile strength of, 500 
 
 Zinc, 423-425. See also Galvaniz- 
 ing. 
 
 melting point of, 424 
 Zone of combustion in cupola. See 
 
 Tuyere zone. 
 Zones of cupola, 249