CORNELL UNIVERSITY LIBRARY Cornell University Library QC 100.U587 War work of the Bureau of standards.AprI 3 1924 012 325 902 Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924012325902 DEPARTMENT OF COMMERCE BUREAU OF STANDARDS S. W. STRATTON, Dikictob WAR WORK OF THE BUREAU OF STANDARDS APRIL 1, 1921 MISCELLANEOUS PUBLICATIONS OF THE BUREAU OF STANDARDS No. 46 PRICE, 70 CENTS Sold only by the Superintendent of Documents, Government Printing Office, Washington, D. C. WASHINGTON GOVERNMENT PRINTING OFFICE 1921 WAR WORK OF THE BUREAU OF STANDARDS CONTENTS Page Introduction lo Aeronautic instruments ii Information ii Testing and investigation ..\.. . 1 . 13 Development 14 Instrument collection .'. . 16 Magnetic compasses 16 Aeronautic power plants 16 Description of altitude laboratory 16 Operating conditions studied in the laboratory 19 Carbureter performance as affected by altitude ...,!...". 20 Aircraft cooling radiators .\ ........ . 21 Airplane-engine fuels .'.'. ''.. 23 Ignition systems 24 Lubrication of aircraft engines , ... ...^ 30 Miscellaneous investigations 31 Aircraft construction 32 Metal construction for airplanes 32 Design of variable-camber airplane wing 34 Laminated wooden members for airplane construction 35 Various wood species as spruce substitutes ; 36 Artificial drying processes for wood ! 36 Wood protective coatings , . . . 37 Aircraft materials ..... ....... .^."' 37 Airplane dopes '..'..'.'/.'.. 37 Paper as a substitute for linen in airplane construction 37 Development of satisfactory cotton fabric for airplane wings 37 Balloon fabrics 38 Investigation of the thermal expansion of materials used in the aircraft- engine construction 38 Aircraft (miscellaneous) 38 Instruments for measuring tautness of airplane wires and cables 38 Synchronizing devices for airplane guns .............. ..^_. 40 Recovery of helium from natural gas 47 Airplane dopes 48 Introduction 48 Function and history of dopes 50 Composition of dopes 51 Cellulose nitrate dopes 52 Cellulose acetate dopes , ^2 Dope covers 54 Application of dopes 55 Fireproofed dopes 56 Relation of dopes to fabrics 57 Balloon gases ■•■••< 57 Compilation of information 57 Laboratory and field work upon the generation of hydrogen *. . . 58 3 4 WAR WORK OP THE BUREAU OP STANDARDS Balloon gases — Continued. ■P*^' Tests of hydrogen plant at Langley Field 5° Inflammability of jtnixtoes of hydrogen apd, helium and the ignition of balloons by iilcendiary bullets ; . . . . S9 Gas-leak detectors ^^ Means for determining the specific gravity of balloon gas 59 Calibration of testing machines i.'. . J: 59 Comparison bars for calibration : ' ," ' ' ' Preferable calibration methods • • ■ ;• •, Hardness machines and extensometers • • • .■ • "^ Chemical investigations (miscellaneous) • °^ Ferrous materials. • • ^^ Nonferrous materials -,; • ; ■ °^ Platinum metals and platinum substitutes ■ ^4 Prevention of corrosion ••■•■■ ^^ Removing metal fouling from rifle barrels. 68 Soap ■■ ■ 69 Carbon tetrachloride fire-extinguishing liquid ■ 69 Antifreezing solutions for automobile radiators. '. 69 Prevention of foul-gas formation in ammonia-absorption refrigerating ma- chines 7° Development of automatic analytical apparatus for use in nitrate plants . . 70 Hydrogen detectors for use in submarines ; 7^ Determination of carbon dioxide and oxygen in the airof a submarine 71 Tests of gas masks and oxygen-supply apparatus 71 Miscellaneous 72 Chromatic camouflage and chromatically concealed insignia 72 Coke-oven investigations .; 73 Preliminary work of the Bureau 73 Fuel situation in early part of igi8 74 Attention focused on Roberts oven 74 Test conducted by the Bureau 75 Results of test at Dover 76 Conference at Pittsburgh ' . 77 Further work by the Bureau 77 Interest in the use of midcontinent coals increased 78 Test of iCoppers plant at St. T'aul 'I .' ••••,• 7^ Results of test at St. Paul ..[.:..'.'. .', .:.\' 79 Further work a% Doyeir. .'.[" 79 Blast-furnace test on St. Paul coke '. ". . . . .",. . . 80 Additional work carried on by the Bureau V. ........ . 81 Conclusions 81 Cbncrete and cement 82 Concrete ships , 82 Load tests of concrete floors of Arlington building. 82 Laboratory work on cement and concrete 84 Concrete ships 86 The development of a light, strong concrete 86 Corrosion tests of reinforcing steel 87 Bond tests 88 Repeated reversal of load on reinforced-concrete beams 88 Impact tests on concrete slabs and built-up steel plates ; 88 Leakage of water through cracks in reinforced-concrete shell 89 Shear and compression in reinforced-concrete beam^. 89 CONTENTS 5 Gbacrete ships — Continued Page Reinforcement of concrete slabs 91 Length of lap required for reinforcing bars 92 Value of brackets and haunches in flanges 92 Shrinkage of concrete in setting as related to ship construction 92 Electric batteries 93 Development of chemical work on batteries , 93 Standardization and preparation of specifications for batteries 97 Reports on special subjects :.,..'.... 100 Military tests of batteries : loi Electric blasting apparatus 107 Electric generators 108 Primers no Electric tractors and trucks 112 Object of tests 112 Method of testing ....'. I , 112 Results of tests ■. 113 Limitations shown by tests 'j -Vi . . ;■. i ; ; . . . 113 Electrical inductance method for location of metal bodies ;..'.;.;'•.; ... . 114 Gages, precision : j . . . 116 Testing of munitions gages 1 16 Manufacture of precision-gage blocks ..'.... 116 Salvage of obsolete gages 117 School of instruction in gage work 117 Publications on gage work 117 Illuminating engineering : 117 Acetylene generators for field service. .' :.....: 117 Incandescent electric lamps. 120 Protective fence lighting 121 Rockets and illuminating shells ^ 124 Signaling lamps for daylight transmission of messages 126 Inks and ink powders. .....;......... i 128 Liquid inks and ink powders. .. :. .j ; i . ... .i'. ..... 128 Indelible inks ; ...»..'.*... 129 Invisible signaling ;..... .:j>i. 130 Ultra-violet position marks and Ultra-violet signaling ,\ 1 . . . 130 Signaling by invisible radiations. ..........;.. 133 Thermal-radiophonic devices for secret signaling 135 Invisible writing, means for the detection of 139 Leather 141 Investigation of sole leathers 141 Composition soles 143 Investigation of strap and harness leathers w 144 Chrome retan and bark-tanned upper leathers : 144 Transmission belting 145 Leather packing for recoil cylinders 143 Waterproofing materials for use on Army shoes 146 Shoe cements 146 Shoe laces 146 Rubber heels 147 Specifications and routine tests 147 Magnetic investigations : 148 Magnetic compasses 148 M^netic analysis of rifle-barrel steel 149 6 WAR WORK OP tun BUREAU OP STANDARDS Ease Manila rope ^S^ Medical supplies ^53 Dental amalgams ^53 Druggists' rubber supplies • ^54 Dilution pipettes ^54 Metallurgical investigations ^54 Aluminum ^54 Research in connection with the welding of steel ^S^ Copper crusher gages 'S^ Machine-gun erosion '59 Identification tags for the Navy and Army i59 Tests of new alloys '59 Gold-palladium alloys as substitutes for platinum iS9 Nickel spark-plug electrodes i6o Cements for spark-plug electrodes i6o Conservation of manganese ■■ i6o Oil proofing concrete liners ■• i6i Tin conservation ;......... i6i Cadmium solders i6i Bearing metals 162 Metals for aeronautical instruments 162 , Test-bar fractures 163 Hardness of brass cartridge cases 163 Centrifugal-steel castings 164 Magnesium-aluminum alloys for mirrors 164 Steel helmets for the United States Army 165 Effect of prolonged annealing high-silicon cast iron 165 Horseshoe nails 165 Deterioration of nichrome castings upon heating 166 Light armor plate .■ 166 The relative cutting properties of "cooperite"and "stellite" 166 Miscellaneous investigations of iron and steel 167 . Preparation of specifications 169 Foundry research and production of castings 169 Work for the Ordnance Department of the Army. , :. , 170 Miscellaneous metallurgical investigations for the military departments. ... 170 Military inventions 172 Information furnished by correspondence 172 Membership in war boards and military committees 172 Physical tests of metals and metal structures 173 Load tests of 150-ton floating crane for the Navy 173 Strength and efficiency of electric welding. .: 174 Investigation of strength of chains 177 Wheels, investigation, of, artillery, truck, and airplane 178 Metal construction for airplanes lyg Natural-gas investigations jyo Beginning of natural-gas work 170 General natural-gas investigation ; 170 Relation to the United States Fuel Administration 180 Technical results igi Investigation of town-border, natural-gas meters 182 Recovery of helium from natural gas 182 CONTENTS 7 Page Optical glass and optical instruments 183 The production and testing of optical glass 183 Testing and design , of military optical instruments , . , . . 185 Binocular testing , , 189 Mil scales ■ 192 Ordnance 193 Hardness of brass cartridge cases 193 Machine-gun erosion 193 Copper crusher gages , . ,• •,■ 193 Defects of shrapnel steel ••,,'!•> ^94 Light armor plate ■,•■■:• '94 Synchronizing devices for airplane guns 194 Removing metal fouling from rifle barrels 194 Munitions gages 194 Airplane-gun mount 194 Powder-shipping containers 194 Semisteel shells 195 Copper rotating bands for shells 19S Gun barrel with gored rifling 195 French and American trench mortars 196 Railway mount recoil piston rods 196 Wheels, investigation of artillery, truck, and airplane 196 M^netic analysis of rifle-barrel steel 196 Paper 196 Wall and plaster board 196 Paper as a substitute for linen in airplane construction 198 Paper filters for gas masks 199 Paper containers for axle grease and saddle soap 200 Photography ao2 General photography 202 Photography as applied to spectroscopy 203 Photographic spectroscopy at the Bureau of Standards 203 Military problems in photography 204 Photographic method for detecting camouflage 207 Protection of motion-pictture film from heat of lamp 207 Metallic mirrors for cameras 207 Automatic regulation of diaphragm by intensity of light 207 Production of dense barium crown glass for lenses 207 Tests of photographic lenses and camera shutters ..< 208 Protective coatings 208 Electroplating investigations 208 Bituminous materials 211 Varnishes 214 Enamels 216 Slushing oils 217 Fireproofing of fiber board 218 Fire-retarding paints 218 Paint specifications 219 Paint testing 219 Publications and information 220 New scientific publications , 220 Metric literature 220 Soldiers' Manual of the Metric System 221 Technical information activities 221 8 WAR WORK OP THE BUREAU OP STANDARDS Page Radio commimication 222 Radio in warfare ..!........ 222 Relation of Bureau of Standards radio latdratoiy^ to ftiilitaiy services 224 Cooperative activities ■ ^^S Information and reports to the Army and Navy "^^ Radio instruction =226 " Radio Instruments and Measurements' '. . , ^^7 "The Principles Underlying Radio Communication" '. 228 Radio direction finder 229 Coil aerial as transmitting and receiving device -' , 33° Submarine radio signaling 231 Signaling system for airplane landing 232 Electron tubes ' 233 Special military researches on electrdn tubes : 235 Radio measurements and tests 237 Cathode-ray oscillograph ....'.!...'. .'■.' 239 Insulating materials 240 Pertaanent crystal detectiors . i ,.,...!.. ^ , 242 Peace-time value of work ...'!;..... 243 Audiometry .'.'.. . . . ; 245 Researches 245 Radiant power life tests of quartz-mercury vapor lamps 245 Optical properties of balloon fabrics 245 Photoelectric sensitivity of molybdenite aiid viarious other substances 245 Amplification of bolometer current for signaling purposes. 245 Infra-red transmission spectra of various Substances 245 Protection of moving-picture film from heat of lainp. , 245 Glasses for protecting the eyes from injurious radiations 246 Reflecting poWer of stellite and magiialiuin 246 Instruments and methods of radiometry. 247 Lacquers for protecting tamishable metallic sijrfaces. , 247 Specifications and mettods of test 248 Conferences, consultations, etc 249 Publications ^ .■ 251 Thermal-radiophonic devices for secret signaling. , 251 Radium -..■..., ,,,..; 251 Information and recommendations '. :, 252 Investigation of specific problems. . . i . . ,. i 254 . o Equipment for factory inspection of dials 255 Routine precision measurements 255 Rope, manila 2-, Rubber , 255 Solid tires 2-- Pneumatic tires and tubes 2-- Gasoline hose 2cq Oil-pnx)f packing 259 Insulated wire • 260 Mechanical rubber goods 260 Rubber jar rings for canning 261 Druggists' supplies 261 Safety standards for military industrial establishments 261 Searchlights _ , 263 CONTENTS 9 Page Sound-ranging apparatus 265 Introduction , 265 Preliminary work ,....,,.., 268 Brief review of investigations of various phases of the sound-rangiiig prob- lem .:.'... ;..... . .' '. ' 269 Sounds transmitted through the earth. ., 171 Qeophone , 271 Seismomichrophone 272 Stibmarine detection. 273 Means for visual detection. ...:.:... : '. . 273 Telephone problems. . . 274 Study of the telephone situation in the District of Columbia 274 Testing machines, calibration of ........;.. 276 Textiles ..;;..; 276 Cotton fabrics. . , . , ..'.....,.•.■ .^ 276 Woolen and mixed fabrics and felts 279 Waterproofed canvas ..'.'. 281 Airplane fabrics .'.■.'. 282 Cordage. , ;.,.',i;., 282 Military textile equipment, uniforms, blankets, etc. 283 Dyestuff chemistry 285 Timepieces. ; 286 Information and specifications > 286 Chronometers , 286 The beginning of a new industry 287 Testing of timepieces'. ......'. 287 Stop watches 287 ; Airplane clotks^ < 287 Toluol recovery , , 288 Preliminary work, conferences, etc 288 Report on toluol recovery : : : 290 Additional committee work ' .1 290 •Publications on tojuol recovery and gas standards. 291 Previous publications combined in technologic paper 292 Increased production of toluol in the United Stktes. . ; 292 Wheels, inVestigaticin of artillery, truck, and airplane 293 Wheel-strength requirements 293 Types of laboratory tests 294 Value of tests 294 Artillery wheels ; .' 295 Truck iwheels. .'. ;.,.; .li i ;..... , 295 Technologic paper on wheels 295 Unique stress analysis of wheel 296 Analysis aids effifci^t design '....; 296 Airplane wheels 297 , Palmer wire-spoke wheel ,. , 297 Comparison with early American wheel , 297 X-rays '. i. ...'..'.- 298 Protective materials j ,;.;.. 298 Radiography... >. >. , .....,...:. 299 lO WAR WORK OF THE BUREAU OF STANDARDS INTRODUCTION It is the object of this report to collect in one pamphlet short descriptions of some of the more important work carried on by the Bureau of Standards during the war, and which was of direct service to our military forces. It may be stated in this connection that practically all the time and energy of the Bureau's personnel were devoted to military problems during the period of hostilities, and therefore a war-work report becomes, in truth, a report of all the Bureau's activities during that time. Not all the investiga- tions which the Bureau conducted during the war are mentioned in these pages. Unfortunately, some of the most interesting and important of them are of such a nature that for military or other reasons they can not be described. Likewise it has been impossible to cover in a comparatively short pamphlet all the investigations in which the Bureau had a part; but the more important ones (with the exceptions nbted above), and those which seem most likely to result in permanent benefit not only to the military departments, but also to the industries and public, are believed to have been touched upon. Among the most important work which the Burea,u performed dxuing the war was the examination of all sorts of plans and inventions designed to be used by the Army and Navy. The Bureau, through its cooperation with numerous governmental organizations and committees, assisted in selecting from the great mass of ideas submitted, those which appeared to possess actual value. Members of the scientific staff served on these committees, thus maintaining the closest possible contact between the committee and the Bureau. A few particular cases of. this assistance are described in the subsequent paragraphs, but many have necessarily been omitted. A great deal of the assistance which the Bureau, through its experts in all the fields of science, was able to render to our military forces came through oral advice and consultations. The value of this work can not even be estimated, and unfortunately most of it was not of a nature to become a part of the permanent records. Without doubt, a great deal of valuable time and money was often saved by the military service through informal confer- ences with members of the Bureau's staff. It is hoped that those who read the following pages will find some material of assistance to them in their work, as it is the belief of this Bureau that many of the problem^^ — the solution of AERONAUTIC INSTRUMENTS 1 1 which was undertaken as a war measure — are of equal or even greater importance in the arts of peace. Their solution was one of the real benefits resulting from the war. AERONAUTIC INSTRUMENTS Information From the beginning of the war the Bureau of Standards was the best-equipped organization in this country for furnishing information on the subject of aeronautic instruments. Six years of previous experience in the investigation of altitude instru- ments, including tests made for the United States Army and Navy, furnished the basis for a general study of the many instru- ments used in aviation. Much valuable material had been gathered abroad and through conferences with foreign officials who visited this country, and this was augmented by information obtained by a member of the Bureau's staff diu-ing a trip to some of the principal laboratories, factories, and airdlromes in England, France, and Italy during the war. Through such means as this the exchange of information with the air services of foreign countries was facilitated. Routine testing at the Bureau of a specified portion of the instrument production of this country was another source of information. The intimate knowledge thus obtained regarding the construction and performance of the various instruments particularly fitted the menibers of the staff thus engaged for making suggestions and originating iinprovements which many times proved to be important. Early in the war a critical study was made of British-instrument specifications which were placed at the disposal of the Bureau of Standards by the Royal Flying Corps. Data thus obtained were supplemented by those derived from careful tests of all suitable types of instruments which could be collected. Such information, coupled with the results of a survey of the manufacturing facilities of this country, enabled the Bureau to render expert assistance in writing specifications and advising in regard to instrument production. As a result of this prepara- tion, the performance specifications, and in some cases the con- struction specifications also, for all instruments adopted as standard by the Army and Navy during the war were based directly on technical information obtained at this Bureau. The early work of the Bureau of Standards in studying altitude instruments and in assisting to start the production of aeronautic 1 2 WAR WORK OF THE BUREAU OF STANDARDS instruments made it natural that the military and naval author- ities should turn to the Bureau for information whenever questions arose in any phase of the instrument work. Much informatibn and advice was given put in reply to official correspondence and to verbal requests. A' considerable part of this consisted of examining and reporting on inventions relating to aeronautic ij;i?1riiune»ts and cjth^r g.ppliances submitted by the General Staff of the Arniy and by th^ ^[^tio;ig,l ^^dvisory Coi^|nittee for Aeronautics. Members of the staff were called on frequently fqr conferences with visitors. During the summer of 191 8 prodi;ction officials of the A^^y and IvTavy, visited the Bureau practically every day tp foUp)!? the results of investigations or to discuss specifications in epmieqtipn witl;i instruments. To make iaiq>^^t^qn more readily acc^silple and to provide for a simple means -of distribution in the proper channels, a s^es of conftdentjal technical infprmatioh, circulars was issued covering, y.aripijlS^ phases of ^he instrument work. ; These met with immediate apd j?pntjl^ually increasing demand, as they consti- tuted the first attempt of this nature to satisfy a pressing need. Distribution abpad created a demand among the foreign air services, for thi?,,inaterial, iV large number pf the circulars havp ^^pn furnished British, French, and .Italian officials, and it is jmdjerstpod tjiatin many cases translations have been made. A number pf, these circulars contained cpnfidential infprtnation from the,alliedGovernnients which, together with the necessary secrecy surrounding all developments during ^he war, i^tade it imperative to lijni^ tj^eja;, circulation tp pfl[icial circles. T'he clpse of hostilities inakes it prpbalsle^ Jipwever, that they will be macle available fqf more general distribution as fast as permission can be obtained to release the foreign material. ; Altitude tables wliich were used in calibrating the dials of all altitude instrjfment^ were calculate4 at the Bureau of Standards and were furnished to the manufacturers. In addition, tables. of temperature corrections for al|titude instruments and altitude corrections for air; speed indicators were computed. Valuable and much-ijeeded information was given to the various manufactm-ers regarding suitable testing methods and apparatus tp be used in the testing, Al^p from time to time many useful suggestions were mad,e regarding the best methods of manufactitre and the proper materials to use. Another important part of the Bureau's work consisted in furnishing to the Army and Navy information regarding the AERONAUTIC INSTRUMENTS 13 proper use and repair of instruments. At the request of the Navy Department, a school was conducted at the Bureau of Standards for naval aviators. An intensive course covering the theory, testing, and use of aeronautic instruments was thus provided. In addition information was furnished directly to the flying fields and salvage stations of the American Expeditionary Forces regarding the testing and proper use of instrunients. Testing and Investigation The starting of aeronautic instrument production in this country hecessitated immediate preparation for testing the instru- ments. It was necessary to control the quality of the output by frequent tests more thorough than the acceptance tests given by the inspectors at tlie factories. To this end provision was made to have the Bureau of Standards give thorough laboriatory tests to 4 per cent of the total production. For this purpose the design and construction of suitable testing apparatus was rapidly carried out. This apparatus included temperature chambers and vibra- tion stands, so that the instruments might be subjected to the same conditions as would be encountered in actual use. To facilitate testing at the factory the Bureau developed modifications of its laboratory tests for acceptance purposes and instructed the Government inspectors in these methods of testings, In a number of cases the Bureau calibrated manufacturers' stand- ards and even constructed standards and loaned them to manu- facturers. Investigations were made on different occasions to deterjnine the causes for the failure of instruments to function properly. For instance, at one time the Bureau's tests on altimeters showed consistently poor performance. A careful study showed that this was due to the rusting of; the small, steel chain which operates the pointer. This discovery led to the rejection of -5,000 instruments which had already been accepted by the Air Service. Along the same line tests showed that most of the breakdowns which occurred in a certain type of tachometer were due to a faulty type of escapement. The latter was changed to a sturdier type, and the durability of the instrument was much improved. The laboratory tests on large quantities of instruments were nattirally very important as a source of information regarding the average performance of different types., The long-continued investigation of such instruments as altimeters, airspeed indicators, oxygen apparatus, and tachometers has furnished statistics 14 WAR WORK OP THE BUREAU OE STANDARDS which are and will be in the future of considerable value to the scientific and engineering world. Although complete laboratory tests indicate quite acciu-ately the quality of an instrument, the performance of the latter dunng flight tests will often reveal defects which would not be indicated in the laboratory. Consequently, flight tests were made by mem- bers of the Bureau's staff with a view to noting peculiarities in the performance of different instruments. To mention a specific case, laboratory tests on a certain liquid manometer type of rate-of- climb indicator gave most satis;factory results. However, when the instrument was tested in flight the combined effect of accelera- tion and inclination caused such fluctuations of the liquid column as tp make, the readings of no practical value. Improvements were immediately suggested owing to the discovery pf - this fact which might not have been brought out so forcibly had only laboratory tests been made. In such a manner laboratory data were supplemented by flight data, and an extensive knowledge of the performance of the instruments under all conditions was obtained. Development The first aeronautic instruments submitted for test, with the exception possibly of altimeters, were in a more or less elementary stage of development. The data gained from these tests sug- gested improvements which could be made, and thus the instru- ments were gradually developed. As aviation progressed, new instruments were needed; for example, oxygen-control apparatus for furnishing a regulated supply of oxygen to the aviator at high altitudes, rate-of-climb indicators for giving the vertical component of the speed, and an automatic means of taking instrument readings during per- formance tests; The Bureau of Standards took an active part in the develop- ment of oxygen apparatus in the United States by promoting cooperation between the foreign experts and American manu- facturers in order to get production started. Valuable suggestions were made looking toward the elimination of defects and improve- ment in the performance of the instruments after careful tests had been carried on at the Bureau. A new feature designed to avoid the main source of temperature error was developed. The necessity for measuring accurately the speed of ascent in performance tests created an urgent demand for an instrument AERONAUTIC INSTRUMENTS 15 which would read this quantity directly. The inherent errors in the liquid-manometer, rate-of -climb indicator, made it desirable that a new type of instrument be developed, using some means other than a liquid column as a pressure measuring device. The Bureau of Standards consequently undertook the development of a rate-of-climb indicator whose indications depended on the deflection of very sensitive diaphragm boxes. This instrument was nearly completed at the end of the war and gave promise of being an improvement over former types. As high-altitude flights became more common the need of barographs capable of recording such altitudes became more acute. At the request of the flight-test officers, several lo ooo- foot barographs were remodeled to give a 30 ooo-foot range, according to a method developed at this Bureau. For some purposes it is more convenient to have the pressure of the atmosphere indicated instead of the altitude. On this account a number of altimeters were remodeled and fitted with special dials to meet this need. Detail drawings of a gyrostabilizer for instruments had been prepared at the time of the armistice and a working model of about one-fourth size constructed. A bomb sight and double- pivot compass to go on the stabilizer formed a part of the design. A new type of inclinometer or banking indicator for use in determining the position of the aircraft relative to the proper banking angle was developed. Actual tests showed the instru- ment to be practical and to possess advantages over those ordi- narily in use. The instrument differed from the ordinary bubble-type of banking indicator in having, instead of a bubble, a carefully ground glass ball, the inside of which was coated with a self- luminous paint. This ball was moimted inside of a curved glass tube mounted convex downward on an angularly graduated base. The position of the ball in the tube indicated the position of the aircraft relative to the proper banking angle. The instrument could be easily read without artificial illumination. The importance of securing accurate records of aircraft per- formance tests led a member of the Biureau's staff to consider using the principle of the motion-pictvire camera to obtain a con- tinuous photographic record of the instrument board during flight. It was recognized that the development of such a device would be an improvement in many ways over the use of recording 1 6 WAR WORK OF THE BUREAU Op STANDARDS instruments. Later a similar suggestion w^s reqeiyed from Maj.R. W. Schrpeder, and plans werp madie fpi;i4evelopment of such a "dummy observer," as it was called, and the apparatus was constructed subsfque^it to the signing pf the an^iisticp. Instrument Collection The Bureau of Standards possesses one of the most complete collections of aeronautic instruments in the world. Not only samples of practically every kind of airplane instrument made in this coimtry, but all kinds of foreign instruments, as well, were obtained during the war. A collection of typcal instruments which the Bureau prepared for exhibition purposes is shown in Fig. I. The original purpose of the collection was to aid in instrument production by making available for the manufacturer types of foreign instruments with which he had had no experience. In several cases production was facilitated at the beginning of the war by the copying of instruments thus made available by the Bureau of Standards. Later, as it grew, the collection became useful for instruction and investigation purposes. The great variety of American, British, French, Italian,; and German instrur ments included make it one of the most useful collections for reference and comparison in existence. Magnetic Pompasses This subject will be found to be treated under the heading, "Magnetic Investigations." AERONAUTIC POWER PLANTS Description of Altitude Laboratory The part played by aircraft in the war was of the first impor- tance, and at thp time the United States ei^tered the conflict it was apparent that the victory was likely to go to the side haviii^ the largest number and mpst .efficient types oif machines. It therefore became the duty of the Goyeriiment to make a thorough study of the aeronautic situation with a view to improving, wherever possible, the performance of airplanes and their appurtehancesl Since the success of an airplane depends very largely iipon the operation of the power plant, a study of aeronautic-engine per- formance on a larger scale than anything before attempted was decided upon. As a consequence, in cooperation With the National Advisory Committee for Aeronautics, the Bureau cPnstructed what has since become known as the altitiide laboratory for the Miscellaneous Publications. Bureau of Standards, No. 46 Fig. r. — Group of typical aeronautic instru-ments The bureau has assisted the aviator by supplying him with more reliable indicating devices, the sense organs of an airplane Fig. 2. — The ''Liberty 12" airplane engine mounted in the altitude chamber When the doors (one oi which is shown at the right) are closed this chamber is practically airtight. The air pressure and temperature are then lowered to those corresponding to any desired alti- tude, and the engine tested under actual flying conditions Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 3- — The carbureter test plant This plant is specially designed to test airplane carbureters under altitude conditions. Hven the pulsations in the inlet manifold caused by the pumping action of the engine's cyUnders are duplicated in this apparatus AERONAUTIC POWER PLANTS l^ testing of aircraft engines as nearly as possible under service conditions. Since its construction, the laboratory has been in almost continuous operation and results of the greatest possible value, not only in the field of aeronautics, but to the automotive industry in general, have been secured. It is perhaps not generally known that the laboratory testing of aircraft engines presents peculiar difficulties not met with in the case of other t3Tpes of internal-combustion engines. These difficulties are due to the fact that the engine of an aircraft per- forms most of its work not at the earth's surface, but at altitudes ranging from a few thousand to 30 000 feet. Atmospheric con- ditions at the higher altitudes are very different from those at the earth's surface, the pressure of the air at 30 000 feet being less than one-third that encountered at sea level. The average temperatvu-e at such an altitude is in the neighborhood of — 35° F, and both the atmospheric density and temperature have a pro- nounced effect upon the operation of an internal-combustion engine. The anxotmt of explosive charge which the engine is capable of drawing into its cylinders varies with the density of the air, which, of course, is a function of altitude. Temperatiu-e is also a controlling factor, and, besides affecting the quantity of the charge, produces numerous other variations in the operation of the engine, such as those brought about by changes in the temperature of the jacket circulating water, lubricating oil, etc. When it was decided to conduct a thorough investigation of airplane-engine performance, it was apparent that the laboratory must be so designed that the entire engine could be sttrrounded with conditions existing diuing an actual ffight. Time has shown that this decision was a correct one, but it meant the construction of a laboratory unlike anything before attempted. The essential features of the altitude laboratory consist of a concrete chamber, the walls of which are of such thickness and so reinforced that they are capable of withstanding a considerable pressure from without. Within this chamber, the engine is mounted on a stand designed to duplicate as nearly as possible the typical airplane- fuselage mounting. The chamber is provided with doors on opposite sides, so that the interior may be easily reached, but these doors are arranged to close upon rubber gaskets, so that when shut they form an airtight joint. The exhaust from the engine and the air within the chamber are withdrawn by means of a centrifugal exhauster driven by an electric motor so that the 26035°— 21 2 1 8 WAR WORK OF THE BUREAU OE STANDARDS pressure within the chamber may be reduced as low as one-third of an atmosphere, or that corresponding to an altitude of about 35 coo feet. Fig. 2 is a view of the Liberty 12-cylinder engine set up in the chamber. The air supply to the carbureter first passes over a series of refrigerating coils by which its temperature is lowered to a point approximating that met with at the altitude of the test. Similar coils within the chamber keep the air surrounding the engine at about the same temperature, and electric fans cause a circulation somewhat similar to that met with during a flight. These coils for cooling the air form part of an ordinary direct-expansion ammonia-refrigerating plant. Since the temperatiure of the air can not be controlled very readily by means of the refrigerating plant alone, electric-heating grids are placed within the air passage to the carbmreter so that in practice the temperature of the air is first lowered by the ammonia coils and then raised slightly by the electric grids to the exact temperature desired. The engine is connected by a coupling and shaft, passing through an air-tight stuffing box in the chamber wall, to an electric dynamometer and water brake, by means of which its power is absorbed and measured. Thermocouples are placed at all points at which temperature measurements are desired, and manometer tubes are connected to the engine and piping systems to measure the pressures. These are all led to convenient points outside the chamber. The fuel used is weighed by tanks mounted on scales on a platform sufficiently high to insure a flow of fuel to the engine by gravity. The exhaust from the engine is water cooled, the water entering the ekhaiust pipe at some little distance from the engine, so as not to affect the operation in any way. This water is removed from the exhaust gas in what may be termed "settling tanks" placed outside the chamber. From these tanks the gases are removed by the exhauster, mentioned above, while the water is drained by gravity, the drop of the pipe being sufficient to prevent any break- ing of the vacuum. The altitude laboratory as originally constructed and operated during the war consisted of a single chamber placed in a temporary wooden building. This has now been replaced by an installation of two chambers in the permanent dynamometer laboratory of the Bureau. The arrangement is such that the chambers may be used either separately or together; that is, communicating doors ' AERONAUTIC POWER PI^ANTS 1 9 may be thrown open and both sets of pumping equipment used to remove the air and gases from a single engine. In this way the capacity of the plant will be greatly increased. Operating Conditions Studied in the Laboratory The problem first presented by the National Advisory Com- mittee for Aeronautics for solution by the use of the altitude laboratory was that of the performance of different grades of gaso- line at high altitudes in typical aircraft engines. The lubrication division of the Signal Corps requested also the preservation of samples of the lubricating oil to determine the effect of fuel composition and of altitude on the deterioration of such oils. A staff of two or three men was detailed by the lubrication division to assist in securing the desired data. As different grades of fuel (all of comparatively high quality) affect engine power and performance only to a very slight extent, the satisfactory solution of this problem required the highest possible acciuracy in obtaining complete data on engine perform- ance as previously outlined. Thus a practice was established by which all the meastu-ements of power, speed, fuel consumption, barometric pressure, air and water flow, temperature, and pressure provided for by the apparatus are customarily made, no matter what is the immediate prnpose of the test in hand. The result is, in addition to the data directly desired, an accumulation of valuable supplementary data on engine performance. Observations have been made to determine specifically the following relations : I. Horsepower-altitude relation for engines at normal speed. 2.' Horsepower-speed relation at a range of altitudes up to 30 000 feet. 3. Horsepower-compression ratio for normal speed, using com- pression ratios of 4.7, 5.3, and 6.2 to i at a range of altitudes up to 30 000 feet. 4. Horsepower-inlet air temperature at a range of speeds and altitudes. 5. Effect of variation of intake pressiu-e on horsepower at a range of altitudes to simulate the effect of supercharging equip- ment. 6. Effect of exhaust back pressiu-e on horsepower over a limited range of pressures. 7. Mechanical losses at various speeds, altitudes, and engine temperatures. 20 WAR WORK OP THB BUREAU OF STANDARDS 8. Metering characteristics of a number of different types of carbureters with and without altitude compensation or control for the full range of speeds and altitudes. 9. Optimum mixture ratios for maximum power over the range of speeds and altitudes with several different carbureters. 10. The pCTformance of a number of automatic and hand- operated altitude-compensation devices for different carbureters. 11. The total heat distribution for all speeds and air densities at full throttle. 12. The performance of special fuels: "Hector," a combination of cyclo-hexane and benzol; " Alco-gas," a combination of alcohol, benzol, gasoline, and ether at a compression ratio of 7.2 to i. Other relations have been investigated from time to time. Moreover, the detailed records taken for each test include much information bearing on other characteristics of engine perform- ance, such as, for instance, the behavior of spark plugs and igni- tion systems imder conditions of low-air pressure and temperature. Carbureter Performance as Affected by Altitude In addition to the altitude laboratory the Bureau has con- structed a complete plant for the testing of aeronautic carburet- ers, and in cooperation with the National Advisory Committee for Aeronautics has carried on an extensive investigation of these devices. The performance of carbureters under the conditions existing at high altitudes, where air density and temperature are low, is a matter of great importance and affects markedly the operation of airplane engines. The carbureter test plant is made up of a strongly constructed box of wood and metal with a glass door within which the carbureter is mounted. The opening in* the carbureter which would ordinarily be connected to the inlet mani- fold is coupled to a pipe leading to a centrifugal exhauster driven by an electric motor. A shutter is placed in this pipe and con- nected by step pulleys with a small motor, so that it may be driven at any desired speed end thus simulate the conditions of pulsating flow encountered in an actual manifold. Air is admitted to the inclosed chamber through a metering orifice and regulating valve, and its temperatm-e may be raised above that of the svir- rounding air by means of electric heating grids. Fuel is fed to the carbureter from a weighing tank and automatic devices are provided for recording the consumption in a given time. The construction of this plant is clearly shown in Fig. 3. With this plant complete investigations have been carried out on all the AERONAUTIC POWER PIvANTS 21 ordinary styles of airplane carbureters, and their metering char- acteristics have been determined over the range of altitudes likely to be encountered in actual service. The plant was not designed and has not been used to eliminate the engine testing of carbu- reters, which will always be the final determining factor, but the preliminary work may be done on the carbm-eter-test plant much more rapidly and cheaply and with far greater accuracy than on an engine. With the test plant all undesirable variables may be omitted and only those factors changed which it is desired to study. Aircraft Cooling Radiators In connection with the work on aeronautic power plants, the Bureau of Standards has conducted for the past four years an extensive investigation of the performance of airplane radiators. The efficient functioning of the cooling system is an extremely important matter in the case of aircraft engines, since any form of radiator offers some head resistance and must take away a por- tion of the power of the engine which would otherwise be available for driving the plane, while at the same time a radiator of high cooling capacity is needed, owing to the fact that aviation engines are generally nm at their maximum power for long periods of time. The ideal radiator for airplane use is, therefore, one with a large cooling capacity and low head resistance (not necessarily for the radiator alone, but for the combination of plane and radia- tor) , but these two requirements are in general contradictory. This subject became of extreme importance when the United States entered the war, and shortly thereafter the Bureau under- took, in cooperation with the National Advisory Committee for Aeronautics, a complete investigation of the relative merits of existing types of airplane radiators. Since that time, tests have been made on nearly all makes of radiators now in use, and the results have been correlated and issued for the benefit of those interested, through the reports of the above-mentioned committee. Specimens of the various standard radiators were submitted by the manufacturers. These were all constructed to certain given dimensions of height and width, but the sizes of the air and water tubes and the depth of the core were those which would be used in the actual radiators. In this way, all the tests were reduced to a common basis, only those peculiarities which were due to the design and not to the size of the radiator having an effect on the result. 22 WAR WORK OP THE BUREAU OP STANDARDS The investigational work carried on in this connection falls naturally under two heads: (i) The determination of the head resistance of the radiator core, and (2) the determination of the efi&ciency of the radiator as a cooling device. The first property was investigated by mounting the specimen of radiator core upon the vertical arm of an aerodynamic balance in one of the Bureau s wind tunnels. Air streams of various known velocities were caused to impinge on the specimen, and its head resistance could therefore be easily determined. In these tests the bare core without water boxes was used. Experiments were likewise con- ducted to determine the relative head resistances of a model fuselage with stream-line nose and radiator mounted in an open position on the plane and that of the same fuselage with a nose radiator, such as has largely been employed on modem machines. The results of this investigation were of particular interest as showing the very large amount of power absorbed by nose radiators. Although this position possesses certain advantages, particularly in the case of airplanes designed for use in warfare, it would appear to be a very poor position for the radiator from the point of view of head resistance. Two methods were used for determining the efficiency of a radiator as a cooling device. In the simplest of these, the radi- ator was mounted in a timnel through which an air stream was forced. At the same time superheated steam at atmospheric pressure was admitted to the water space of the radiator and the amount of steam condensed in a given time with a given air flow determined. The radiator which would condense the greatest amount of steam in a given time was, of course, the most efficient cooling device. The other method used consisted in moimting the radiator in a tunnel, the whole of which was inclosed in an air-tight tank. The pressure within this tank could be reduced to that corresponding with any desired altitude up to about 35 000 feet, while at the same time the air was circulated through the radiator by a fan driven by belt from an electric motor. The same air constantly passed through the radiator, but was cooled by a honeycomb on the other side of which water was circulated. In these tests water, and not steam, ■v^as used in the radiator. The timnel, just now described, was the one originally used in carrying out the investigation of radiator-cooling capacity, but owmg to the fact that a very long time was required for establish- ing stable conditions within the tank, the steam tunnel was finally AERONAUTIC POWER PLANTS 23 used, as it was found that it gave just as satisfactory comparisons between different types of radiators. As a result of this work it has been possible to state with accu- racy the percentage of the total available horsepower of an airplane engine which any given type of radiator will absorb and also the efficiency of the radiator as a cooUng device. Fundamental data have been collected, and the Bureau has been able to indicate certain general features which should be observed in the design of a successful radiator for airplanes. Airplane-Engine Fuels From the beginning of the war the work of the Btueau of Mines on automotive-fuel conservation was actively supported by the Bureau of Standards through close and constant cooperation. At the meeting of the Interallied Commission on Specifications for Petroleum Products of the InteralUed Petroleum Conference some of the foreign delegates urged a very rigid gasoline specification for fuels for use in combat aircraft, basing their recommenda- tions on their own satisfactory experience with certain gaso- lines, but without full knowledge of the possible performance of other gasoHnes. The adoption of such close specifications would have materially reduced the available supply of combat aviation gasoline and introduced unnecessary expense and serious difficulty, not only in production but in overseas shipment and distribution. A careful comparison of the power-producing quaUties of 10 dif- ferent gasoHnes selected through cooperation with the Bureau of Mines was made by actual engine tests under flight conditions in the altitude laboratory of the Biureau. A report on these tests was the only quantitative data presented to the interallied com- mission on this subject, and it resulted in the adoption of more liberal gasoline specifications, which were amply justified in service and which resulted in a saving to the American people of millions of gallons of oil in their fields and thousands of dollars waste effort in the refining and distribution in France of aviation gasoUne. In connection with the work on aviation fuels, a thorough study was made of the best methods for determining the quality of gasoHne. It is generally recognized that the simplest test, that for gravity, does not give a reliable indication of the value of gasoline as a fuel. On this accotmt tests were made of the vis- cosities of various commercial and aviation gasoUnes, in the hope that this property might prove to be of value, in connection with others, in spedfyiag the quaKty of fuels. It was found that the 24 WAR WORK OP THE BXJREAU OP STANDARDS viscosity has a marked influence on the metering characteristics of carbureter nozzles, and must be taken into account. A thorough discussion of this subject will be found in Technologic Paper No. 125, "The Viscosity of Gasoline." In addition to service tests on common and aviation gasolines, a number of synthetic fuels were tested at the request of the Bureau of Mines and the Navy Department, respectively. Reports were prepared on the performance of " Hector" fuel, a mixture of 70 per cent cyclohexane and 30 per cent benzol developed by the Bureau of Mines, and "Alcogas," a fuel containing a large per- centage of alcohol. The general conclusions from these tests will be compiled for publication. Five proprietary preparations in- tended to be used as gasoline substitutes were tested for the military service. Many of the so-called "gasoline improvers" were tested for the inventions section of the General StaflE of the Army. Two devices for the convenient use of kerosene instead of gasoline in the usual gasoline engine have also been tested for this section. Ignition Systems In November, 1916, the National Advisory Committee for Aero- nautics requested the Bureau of Standards to tmdertake a study of the causes of failure of spark plugs, and to develop, if possible, a new insulating material which would be superior to those then available for use in such devices. This request was the result of reports from abroad which indicated that spark-plug trouble was one of the limiting features in the design of the most recent high- power aviation engines. A program was laid out according to which the Bureau's ceramic laboratory was to prepare a number of samples of porcelain of widely different composition. These were to be tested by the electrical laboratories for their relative merits, and on the basis of these results further porcelains were to be made up until a satisfactory material was found. The early experiments indicated a great need for more informa- tion as to the conditions under which spark plugs were required to operate, and when the more general program for the study of all phases of aeronautic power plants was begun at the Bureau the ignition investigation was enlarged and fitted in as one of the principal divisions of the program. The ignition staff was enlarged, and during the latter part of the work consisted of seven persons; the work included a study of the conditions in aviation engines and the measurements of the various constants of ignition systems, as well as the development of materials. AERONAUTIC POWER PLANTS 25 Throughout the work intimate contact was maintained with ofRcers of the National Advisory Committee for Aeronautics and the Bureau of Aircraft Production, and much of the testing was at the direct request of the inspection section of this latter biu-eau. Owing to the frequent and unavoidable changes in organization of the branches of the War Department connected with aviation during the war, it was found difficult to obtain data as to the actual service performance of the various spark plugs and ignition systems in use at the flying fields. The duty of the ignition system is to produce at the proper time in the interior of the combustion space of gasoline engines a spark which, by igniting the compressed mixture of gasoline, vapor, and air, causes it to burn and deliver its energy to the engine. All of the systems in practical use on high-speed engines are of the jump-spark type. The essential featiu-e of the system is an induction coil, which consists of an iron core upon which is wound a primary coil of few turns of relatively large wire and a secondary coil having a very large ntunber of turns of fine wire. A current is established in the primary winding either by means of a battery connected in series with the coil, as in the case of so-called battery and coil systems or else by inducing ciurent magnetically in the primary coil by the relative motion of the coil and of a permanent magnet, as is done in the case of magnetos. With either type of system when a spark is desired in the engine, the circuit breaker is suddenly opened by means of a cam and the magnetic flux pro- duced by the primary current rapidly decreases. This decrease in flux generates in the many turns of the secondary winding a high voltage which is conducted from the distributor to the spark plug in the proper engine cylinder. The current passes across the spark gap at the plug to the grounded electrode and rettuns through the engine frame to the other end of the secondary coil. There are two essential operating characteristics of this type of apparatus which must be borne in mind in considering its behavior imder various adverse conditions. First, with the secondary cir- cuit open the voltage will build up to a certain maximum value (10 000 to 20 000 volts) which is determined by the inductance and capacity of the circuits, and no spark will be produced xmless the breakdown voltage of the gap between the spark-plug elec- trodes is less than this value. The length of the spark gap usually used in aviation engines is such as to break down at about 6000 volts, thus affording in this respect a factor of safety of two or three. Second, if the spark plug is shtmted by a resistance, the 26 WAR WORK OP THE BUREAU OP STANDARDS maximum voltage obtaineid across its terminals is very consider- ably reduced, the reduction being greater as the shunting resist- ance becomes less. With most ignition systems, the voltage reached becomes less than the 6000 volts required to produce a spark when the shimting resistance is less than 100 000 ohms. It is, therefore, the duty of the spark plug to perform three func- tions: (i) To maintain a gap between its electrodes which shall have a breakdown voltage of approximately 6000 volts; (2) to maintain an insulation resistance between its terminals of not less than 100 000 ohms; (3) to be substantially gas-tight in order that the leakage of heated gas through the plug may not raise it to such a high temperature as to cause preignition of the engine and destruction of the plug. These requirements can be met with comparative ease in spark plugs used in automobile engines, but the conditions of operation in the modem aviation engine, where every effort is made to obtain high powers with small space and weight, are very much more severe. Measurements which have been made at the Bmeau in aviation engines show that the pressures attained may amount to 500 to 600 pounds per square inch, and the difficulty of maintaining gas- tight joints is very great. Also, experience has shown that the mechanical vibration of even well-balanced engines is often suffi- cient to crack the porcelain insulation. During a cycle of opera- tion of the engine the spark plug is alternately exposed to a blast of cold air which during the intake stroke may be at a temperature well below zero, and is immediately thereafter exposed to the flame of compressed and burning gasoline vapor, the temperature of which may be estimated at 2500° C. These severe temperature conditions tend to crack the porcelain, and by the expansion of the parts, to open up cracks for the leakage of the gas. Moreover, refractory materials of the class used for spark-plug insulators (mica, porcelain, glass, etc.) become conductors of electricity if heated to a sufficiently high temperature. In addition to the trouble just mentioned, the flame of burning gasoline deposits a layer of soot on any cold surface with which it may come in contact, and, furthermore, the lubricating oil which is present in the cylinder may be sprayed against the heated por- tion of the spark plug and, decomposing, also leave a layer of carbon. These carbon deposits tend to short-circuit the plug and may reduce its insulation resistance below the limiting value. Another type of failure is produced if the siurface of the electrodes becomes covered with lubricating oil, or more particularly, if a AERONAUTIC POWER PI.ANTS 2? drop of oil bridges the gap, since this oil before being decomposed has a breakdown strength of from lo to 15 times that of air. The voltage required to produce a spark across the gap is much in- creased and may exceed the value which the ignition system can deliver. The aim in spark-plug construction is to satisfy as completely as possible the three requirements of definite spark voltage, high insulation resistance, and gas tightness by suitable modification of design or material, while the aim in ignition-system design is to reduce as much as possible the stringency of these requirements. The ignition staff of the Bureau has endeavored to assist the military departments of the Government in ignition problems: (i) by making such measurements and tests of commercial appa- ratus as were requested by the departments concerned, in order to determine the relative and absolute merits of devices then on the market; (2) by devising new methods of test to show more clearly such relative merits, and preparing specifications based on these tests; (3) by collecting as much information as possible as to the dimensions, constants, and performance of various types of systems in order to be able to give more intelligent advice on the subject: (4) by assisting various manufactures of ignition appara- tus, by tests and consultation, in their development of materials and designs which showed promise of success; (5) by studying the conditions under which such apparatus operates on aviation engines and thus securing data otherwise unobtainable by individ- ual manufacttu-ers; and (6) by developing in the Bureau's laborato- ries improved materials and better arrangements of circuits. The following paragraphs outline in more detail the results of these various lines of work. The program for the development of improved insulating mate- rial mentioned above has been carried out with the result that sev- eral porcelains have been obtained which show very excellent prop- erties as regards electrical resistance while hot, resistance to crack- ing when exposed to sudden temperature changes, and mechanical strength. The composition of these materials has been given to various porcelain manufacturers requesting it, and it has been put into production by the largest spark-plug manufactturer in the cotmtry. The method of measuring resistivity used in this de- velopment work has been adopted by a number of manufactturers for their own laboratories, and others have submitted samples of their product to the Btureau for test. One of these materials, which had hitherto been used only to a slight extent in spark 28 WAR WORK OP THE BUREAU OP STANDARDS plugs, showed very satisfactory results when subjected to the Bureau's tests, and, as a result, has been adopted by sevei'al of the principal spark-plug manufactufers. It was necessary in the work to devise various methods of test- ing spark plugs. In addition to the conductivity measurements developed for the porcelain testing, methods have been devised for testing the resistance of the insulating material to sudden changes of temperature and to mechanical shock, and a quantitative method has been developed for measuring the gas tightness of the plugs while they are heated as in an engine. Methods have also been developed for measuring the heat energy in the spark pro- duced by various types of ignition systems and for determining the ratio of turns and the magnetic leakage between the primary and secondary coils of such systems. A general view of the spark- plug laboratory showing most of the apparatus used in these investigations is given in Fig. 4. The performance of ignition sys- tems as to the maximum voltage obtainable and the minimum re- sistance across which they can produce a spark seems to depend in a very large degree upon the amotmt of eddy current which may be produced in the iron core and upon the electrostatic capacity of the various parts of the circuits. No methods are at present known for conveniently determining these quantities, but the Bureau is still engaged in developing a method which, it is hoped, will provide a satisfactory measurement of these important prop- erties. At the request of the Btureau of Aircraft Production and of the Motor Transport Corps, specifications have been prepared on the basis of which they may ptirchase their spark plugs, and the tests developed at the Bureau have been incorporated in these specifi- cations. A very considerable amount of testing of spark plugs of various designs and of proposed tj^es of ignition systems has been carried out at the request of the inspection section of the Bureau of Air- craft Production. (See Fig. 5.) The results have been reported in detail to this section and to the manufacturers of the particular products tested. Thes6 tests, together with those incident to the development of the Bureau's methods and the study of spark plugs obtained from various foreign sources, have brought up the total number of spark plugs on which records have been kept to over 3 000, among which over 100 distinctly different types are represented. Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 4. — A corner of the spark-plug laboratory At the right are the two electric furnaces used in measuring the electrical resistance of various porcelains, using samples in the form of cups, as shown near the furnaces. The instruments used in the measurements are shown in the center, while at the left are the pressure tanks and classhell jar for measuring gas leakage through plugs. In the rear are sample plugs and porcelain cups submitted for test Miscellaneous Publications, Bureau of Standards, No. 46 ifo Ii7 Ifft H9 Fig. 5. — Cross-sections of typical aviation spark plugs That no one design has yet proved entirely satisfactory is evident from the variety of construc- tions shown here. All these plugs have been tested Sections cf LAMinATED Wood AlRPLAr^E BtAhAS "Tr-is 22G4-0 - Lab fSo.'S:- Cr^ 0206016; 02.18^18- OZeofiB: oazia iS ^^^ **^ But?e./vu OP STA-MBACiDS _(^ Wd.swir-ie.TOM ,13, C - *" Dmwcmm-l F V, U Yvm XE Fig. 6. — Cross-sections of laminated wooden airplane beams This form of construction permits the use of smaller pieces of high-grade wood as well as inferior woods in the building of airplanes. The strength of these laminated beams is just as great as that of solid beams AERONAUTIC POWER PLANTS 29 Measurements have been made of the temperatures existing in various parts of spark plugs when operating in aviation engines, and these results have formed the basis for the other testing work just mentioned. There is still much need for fvirther work to see how these temperatures vary with the type of plug and engine used. At the request of the Bureau of Aircraft Production, extensive oscillographic tests were made on 12 different types of ignition systems. This method of test shows photographically the ciirrent in both the primary and secondary coils at each instant throughout the operation of the system and can be made to show the short- circuit current and open-circuit voltage of the magnetos and the behavior of the circuit breaker. The photographic records ob- tained give a complete accotmt of the performance of the apparatus and the values, of current, voltage, etc., can be read off at leisure. Copies of the records were sent to the various branches of the War Department interested, and the large mass of data contained in them was correlated by the Bureau as time permitted, and the results have been published in the Fourth Annual Report of the National Advisory Conunittee for Aeronautics. In connection with the oscillographic test, other electrical con- stants of the ignition systems, such as the inductance, capacity, resistance, and the heat energy in the spark, were observed on a variety of systems, both to detemime their relative merits and to serve as a basis for possible improvements in the design of such apparatus. It is hoped to continue the testing of a variety of sys- tems in order that the ftmdamental data may be obtained and published for the benefit of American manufacturers. Another distinct part of the ignition work has been the study of the use of subsidiary spark gaps which frequently have been proposed as an improvement over the usual types of ignition systems. These devices have the advantage of enabling a system to produce a spark even though the plug may be shunted by a resistance decidedly less than the limiting value of 100 000 ohms stated above. The manufacturers of these devices, however, frequently have made exaggerated claims as to their merits, and the work at the Bureau was undertaken with a view to determining to what extent these claims are justified and to obtaining definite information as to the possibilities and true performance of this type of device. It has been found that the claims of increased power and efi&dency are unwarranted, but that the device has the very definite merit of causing a fouled plug 30 WAR WORK OP THE BUREAU OP STANDARDS to fire, ance an engine has been made to fire continuously althougn the plugs were fouled with a resistance of only 50O0 ohms. The rubber-insulated (Cables connecting the spark plugs with the distributer occasionally give trouble as a result of the cracking of the insulation. This is a progressive deterioration which occurs where the cable is sharply bent, and after continued ex- posui-e to the high voltage electric stresses incident to the opera- tion of the system. The Bureau of Aircraft Production asked that the precise reason for this cracking be determined, and a series of experiments was carried out which demonstrated con- clusively that the cracking was the result of a chemical reaction between the rubber and the ozone formed by the high-voltage electrical discharges which occurred at the surfaces of the cable. This reaction is localized in cracks whenever the rubber is sub- jected to mechanical tension. The remedies suggested are the use of a thoroughly impregnated braid, the avoidance of sharp bends in the wiring, and the use of such rubber compounds as are least affected by ozone. The results of the various lines of work mentioned above were issued in a number of separate reports which formed a part of a series on aeronautic power plants. These reports were distributed by the Bureau to a limited mailing list made up of those ofl&cers of the War and Navy Departments interested, and, where no data of a confidential nature from a military or commercial point of view were involved, to manufacturers and other persons inter- ested. These reports were all published at a later date in the Annual Reports of the National Advisory Committee for Aeronautics. Lubrication of Aircraft Engines Although the matter of the satisfactory lubrication of aircraft engines was admitted to be one of great importance, it was not until several months after the United States entered the war that a definite investigational program of this subject was commenced. Before the beginning of actual experimental work, representatives of the Bureau visited various aviation fields and obtained more or less complete data on the oils in use. Samples of new and used oils were obtained and were later analyzed in the laboratory. A conference was also held with one of the members of the Council of National Defense, and the subject of lubrication was thoroughly discussed. The results of this preliminary work showed that no one kind of oil was in use at all the fields, and that even at the AERONAUTIC POWER PI.ANTS 31 same field contradictory opinions were held respecting the merits of various brands. In October, 191 7, the Bureau was asked to cooperate with the Signal Corps in testing airplane lubricants. By agreement the Bureau made certain tests of a large number of new oils and of the same oils after they had been used in engine runs made at the Navy Yard. This work led directly to a complete investigation of lubrication problems, which was carried out in cooperation with the Bureau of Aircraft Production. A program was decided upon covering the entire complex situation, and a laboratory fitted up especially for the purpose. Tests in actual engines under service conditions were made in the altitude laboratory, with the object of securing data upon which to base satisfactory specifications for airplane engine oils. Miscellaneous Investigations In addition to the investigations already mentioned, a number of others of fundamental importance have been carried out. Among these may be mentioned experiments to determine the rate of flame propagation in the cyUnder of an internal-combustion engine. For this work a special single-cylinder gasoline engine has been built the cylinder and valves of which are identical with those used on the Liberty aircraft engine. Very successful pre- liminary results have been obtained by an electrical method developed through the cooperation of the aeronautic-power- plants section and the electrical division. The determination of the pressure cycle in the cylinders of high- speed aircraft engines is an extremely difficult matter; in fact, until recently no successful indicator for this purpose was on the market. Indicators entirely suitable for slow-moving stationary gas engines are not at all successful in indicating the pressure ranges on light high-speed aircraft engines, due to the inertia of the indicator's moving parts. The Bureau has been working on this problem for some time, and a successful indicator has been developed which is now being manufactured on a commerical scale. The designing and constructing of a satisfactory indicator for high-speed gasoline engines will greatly aid the work of engineers in problems connected with aircraft power plants. For many years the effect of small amounts of water injected into the intake manifold of gasoUne engines has been discussed, some engineers maintaining that this practice results in the develop- ment of increased power and the removal of a large proportion 32 WAR WORK OP THE BUREAU OP STANDARDS of the carbon from the cyUnder and the piston. In order to inves- tigate the merits of these claims, the Bureau conducted a set of tests in which water in varying amounts was used. The results of these experiments have been made public through one of the reports of the National Advisory Committee for Aeronautics. Many special designs of engines and accessories were submitted for test to the aeronautic-power-plants section during the war. Among these may be mentioned a small direct-injection Marburg heavy-fuel engine, a "barrel type" of engine, and a special Ford automobile engine having i6 instead of the usual 8 valves, as well as two direct-connected gasoline engines and generators designed for battery charging work for the Signal Corps. Among the gas- engine appurtenances may be mentioned the testing of a very successful mercury-cooled exhaust valve for airplane engines and the investigation of numerous carbureters with special devices for altitude compensation. The materials-testing laboratory of the Bureau conducted a thorough study of a broken crank shaft from a small-sized Liberty engine. Tensile, hardness, impact, photomicrographic, and chemi- cal examinations were made on this specimen. The tensile test covered those portions of the shaft forming the crank pin and cheek near the break. The only one of these various lines of study which showed any defect in the crank shaft was that for impact resistance. The Izod impact results showed an average of 50 foot-potmds of energy absorbed, while chrome-vanadium steel corresponding approximately to the analysis of this shaft usually tests about 72 foot-poimds. This lack of resistance to shock might have been due to f atdty heat treatment, especially in the final drawing temperatture, and those concerned with the con- struction and use of these engines were cautioned in that regard AIRCRAFT CONSTRUCTION Metal Construction for Airplanes The tremendous increase in the use of airplanes brought about by the war resulted in an unusually heavy demand for spruce, the wood considered most suitable for constructing such machines. Even before this country entered the war in 1917, however, spruce of the required grade was becoming scarce, and this, together with the desirability of rendering the airplane noninflammable, directed the attention of engineers to the possibiUty of using metal for at least a portion of the airplane. Wood has some AIRCRAFT CONSTRUCTION 33 advantages over metal, particularly where but a few machines are to be built, as the initial cost of the plant necessary for turning out the planes is much lower than for any designed to construct metal parts. Nevertheless, when the last-mentioned plant is once completed and equipped with suitable dies, tools, jigs, etc., the cost per plane is comparatively low. Therefore, metal con- struction appeared to offer mmierous advantages where airplanes were to be produced in large quantities for miUtary ptuposes. The Bureau made a very careful study of the possibility of substituting metal for spruce for certain parts of airplanes, and undertook complete computations of the probable strength of various types of metal beams suggested for this class of con- struction. Many of the proposed sections were found to be im- practicable and were, therefore, discarded. One type of box beam was worked out which appeared to possess the desired strength, and test beams of this design were made from sheet aliuninum, and when subjected to transverse loading showed a strength well above that of Sitka spruce of standard design and equal weight. About this time several commercial organizations engaged in fabricating sheet metal became interested in the possibility of producing a satisfactory all-metal airplane. The above-mentioned company actually constructed a complete machine similar to the Curtiss JN-4, except that it had no engine or wing covering. This was tested by the Btireau, but the wing under sand load was shown to have a factor of safety of only 2 as compared with 7 or 8 for standard wood construction. Since this result was con- sidered unsatisfactory, a member of the Bureau's staff went to the factory of this company and aided them in producing a better design of all-metal plane. A representative of the Aircraft Board was also assigned to this work to assist in securing material, and the results accomplished were largely due to his skill and energy. Many attempts showed that sheet metal having the necessary tensile strength could not be obtained in the nonheat-treated form, either in a carbon or chrome-vanadium alloy steel. Efforts to seciu-e added strength by cold-rolling also failed, and for this reason the use of steel for airplane beams was given up and atten- tion devoted to aluminvun alloy. As these aluminum-alloy beams made a good showing in average strength, two complete sets of wings using this type of beam and steel ribs were constructed 26035°— 21 3 34 WAR WORK OP THE BUREAU OP STANDARDS by the manufacturer and tested on an airplane at McCook Field, Dayton, Ohio. During the tests, which covered about 58 hours of flying, this plane behaved in all respects in as satis- factory a manner as those made of wood. Sand-loaded tests of one of these sets of wings showed that this method of building was in every way as strong as the best wood construction. The possi- bility of the successful manufacture of airplanes with metal-wing frames was thus demonstrated, and the problem of insurmg fire- proofness of the plane now hinges on the discovery of a satisfactory noninflantmable wing covering to replace the fabric ordinarily used. Although hostilities ceased before the solution of this problem was reached, considerable progress had been made and the investigation is being continued with fair prospects of success. It is hoped to develop a covering which will be not only fireproof, but transparent as well. Many other tests of metal parts besides wing structures were conducted in this laboratory, such as hydrostatic, tensile, and compression tests of electric arc-welded longitudinal joints in steel pipes, and routine physical tests of inventors' models of ribs, beams, struts, cables, tumbuckles, etc. An investigation was carried out on triangular girders for dirigibles, and as a result it was shown that those first constructed were relatively too strong in the diagonal lattice elements and too weak in the longitudinal members. A redistribution of metal made as a result of this dis- covery resulted in a very satisfactory form of girder manufac- tured either from the French aluminum alloy " Duralumin " or its American equivalent 17-S. Design of Variable-Camber Airplane Wing The problem of increasing the speed range of airplanes is one of the most important in military aeronautics. The maximum speed has greatly increased within recent years and doubtless will continue to do so; but with this increase in maximum speed has come an equal increase in landing speed, since the latter has been and still remains about one-half that of the former. Various devices may be used to increase the speed range. Among these the most promising from all points of view, including that of construction, appears to be the variable-camber wing. Such a wing, designed by a former member of the Royal Flying Corps, was thoroughly investigated by this Bureau. This wing is so constructed that at low speeds it assumes a deeply cambered or lifting form, while at high speeds it takes a stream- AIRCRAFT CONSTRUCTION 35 line shape. In combmation with wings of the usual construction in triplanes or biplanes, it allows a vastly greater speed range. The design of the wing hinges on the construction of a satisfactory variable-camber rib. This ^ivas accomplished in the wing con- structed at the Bureau, and wind-tunnel tests proved it to be thoroughly practical. Laminated Wooden Members for Airplane Construction As mentioned in the opening paragraph of this article, the tremendous demand for airplanes during the war resulted iri a scarcity of airplane spruce. Numerous attempts were made to meet this condition. One of these, that of constructing metal planes, has already been mentioned. Other possibilities were: (i) The use of some form of laminated members which would allow the utilization of spruce with less waste than in the cus- tomary construction, and (2) the discdvery of suitable substitutes for spruce. The Bureau early undertook an investigation of the first- mentioned solution and has tested many different forms of lami- nated construction as used in ribs, beams, struts, spars, tailbooms, and girders, and has also assisted in the development of wood veneer for wing atid fuselage coverings. As a result of this work, it has been proved that laminated beams made iip in a proper way are just as satisfactory as solid beams, and although the cost of manufacturing is somewhat greater, the reduced cost of the raw material utilized results in a saving. Cross-sections of some of the laminated wooden beams which the Bureau tested are given in Fig. 6. In addition to the tests conducted on soUd and lami- nated beams of the same kind of wood, this series furnished com- parative data on the strengths of such woods as spruce, fir, and cypress, and the relative advantages of rectangular and oblique I-sections, and on the strengths of various types of joints and dififerent brands of glues. It was shown that beams of fir can be produced which will have as high a specific strength as those made from spruce, although such beams will not be quite so stiff. Cypress can not be considered as a satisfactory wood for this construction. Beams made up of pieces glued together were proved to be as strong in shear as the solid wood, and no weakening of the bond with glued joints was observed even when subjected to moisture and vibration. In addition to the test on beams, research work was conducted on many other \vood members. An interesting test was that of 36 WAR WORK OP THE BTJREAU OP STANDARDS an experimental three-ply mahogany veneer-covered wing.;, This wing was constructed with glued joints throughout^ no nails or screws being used. Two different designs of rib were submitted with the wing, one of which proved to be considerably superior to the other. When subjected to drifting and sand-load tests, the wing showed a satisfactory strength; but since its weight was approximately 30 per cent in excess of the ordinary type of fabric- covered wing, it cotild not be considered altogether satisfactory. Other miscellaneous tests included those on so-called "pigeon hollow" spars, Navy model— T-spars, etc. In th^ investigation and development of various types of airplane stmts, the value of the Bureau's work is very forcibly shown. Tests of veneer struts were under way on the ^ day America declared war upon Germany. Many different styles were investigated and certain types of veneer construction were shown to be satisfactory. The compressive strength of such members was investigated. Many types of square-end and round-end struts of various slendemess ratios were tested, and as a conse- quence of this work rational mathematical expressions were formulated and constants evaluated which ^ve results corre- sponding closely to the values determined experimentally. Prof. Boyd likewise developed a formula, of more general application, which is suitable for use in connection with the tapered struts used in airplane wing construction. Various Wood Species as Spruce Substitutes As another possible means of assisting in the rapid production of rnilitary airplanes, the Bureau, in cooperation with the Aircraft Production Board, undertook an investigation of the possibility of using woods other than spruce. The investigation of the different types of woods was one of the most complete carried out by this laboratory. Woods of all promising species were investigated, and many special tests were developed for use in this connection, some of which required the use of novel designs of testing appa- ratus. A special hammer for a beam-impact testing machine and a rotating bar machine for fatigue testing were designed par- ticularly for this investigation. On the basis of these tests, a table and data were furnished showing the relative values of various woods for airplane work. Artificial Drying Processes for Wood The successful substitution of other species of wood for spruce in airplane construction def)ended very largely upon the discovery AIRCRAFT MATERIALS 3 ^ of a satisfactory quick-drying process which would not check the woods. Air drying is, of course, the ideal method for preparing lumber for the mill, but in the case of certain species of wood it involves months or even years of delay. The Bureau's participa- tion in the development of suitable artificial drying processes con- sisted chiefly in the inspection of the plants of operators using various patented processes, in proposing desirable improvements, and in testing the products. In nearly all of the plants inspected it was found that no particular engineering skill had been used in the kiln or plant construction, and the operators did not appear to have much knowledge of wood technology. The Btireau's sugges- tions were thankfully received and promptly put into operation in the majority of instances with marked gains in plant efficiency. Some of the processes investigated indicated that it was possible to produce wood equal in strength to that dried in air by artificial drying methods with a great saving in time. Wood Protective Coatings • For certain parts of the airplane structure, such as the propeller, the protection of a metal coating or a composition impregnation has proved highly desirable. Initial tests on electrodeposited sheet copper for such a piu-pose led to the conclusion that a satis- factory grade of this material could be produced. The covering of the entire propeller involved too great an increase in weight, however, to be practical, but the plating of the tip of the blades has been successfully carried out. Bakelite-impregnated woods were also investigated with the idea of using them for propeller construction and also for the knanufacture of ttio tor- truck steering wheels. Bakelite-impregnated maple and walnut specimens shoWed increased strength over ordinary wood, but the weight increases were so great that such wood was riot deemed suitable for aircraft use; although it has been successfully employed in automobile work. AIRCRAFT MATERIALS Airplane Dopes A complete description of the Bureau's work in this field will be found in the article xmder the above heading. Paper as a Substitute for Linen in Airplane Construction This work is described in the article on "Paper." Development of Satisfactory Cotton Fabric for Airplane Wings This work is treated under the heading "Textiles." 38 WAR WORK OF THE BTJREAU OF STANDARDS Balloon Fabrics . ^ Radiant-power life tests of quartz-mercury vapor lamps: This work was undertaken in connection with the investigation of the deterioration of balloon fabrics when exposed to sunlight and ultra- violet light. The ultra-violet as distinguished from the infra-red rays appear to have a marked effect in accelerating chemical action; as, for example, in tiie fading of dyes. There has arisen among manu- facturers of paper, dyes, cloth, rubber goods, paints, etc., a need for. information concerning the sotirces and their constancy.^ of emission of ultra-violet radiations, for use in testing the lasting quality of their products. It is well known that the intensity of the radiation (especiallj the ultra-violet component) from quartz-mercvuy vapor lamps, decreases greatiy with usage. In response to the demand for exact data, methods were devised for determining quantitatively the jlecrease in intensity of emis- sion jcjrith usage, and measturements were made on radiant-power life tests of a number of quartz-merciuy vapor lamps, using the radiometric instruments devised at this Bureau, and described in Scientific Papers Nos. 229 and 282. The data obtained indicate a marked decrease in the total radia- tion emitted, as well as the ultra-violet component, in the course of 1000 tiours' usage of the quartz-mercury vapor lamps now obtain- able on the market. See Scientific Papers Nos. 196 and 244. In connection with the general investigation, (chemical, mechan- ical, etc.) of this subject, tbe transmissive and reflective properties of numerous samples of balloon fal^rics were determined, using a hemispherical mirror and thermopile. Measurements were ma,de also of the rise in temperature when the fabrics were ejqjoseil to solar radiation, using a modification of the fine thermocouples used in measuring heat from stars. Investigation of the Thermal Expansion of Materials Used in Air- craft-Engine Construction The thermal expansion of the materials used in the building of aircraft engines was determined by the Bureau, to aid in the selec- tion of proper metals and alloys for the various parts. AIRCRAFT (MISCELLANEOUS) Instruments for Measuring Tautness of Airplane Wires and Cables In an attempt to reduce the weight of airplanes to an absolute minimum the factors of safety employed in the design of many AIRCRAFT (MISCEIylvANEOUS) 39 of the members were frequently lowered during the war. Since in many cases the stresses set up in the members by the initial tension in the stay wires is a large percentage of the total stress to which they are subjected, it therefore becomes important to know the magnitude of these loads. No satisfactory instru- ments were available to manufacturers or inspectors in the early stages of the war for determining the initial load on the stay wires. A "tuning" method was used in some instances with satisfactory results, but this method required an observer with a highly trained musical ear. The Biureau attempted the solution of this problem, which was successfully completed by a member of the Bureau's staff. An instrument known as a tensiometer was designed and built in the latter part of 191 7. (See Fig. 7.) The instrument is now being successfully used by both the Army and Navy in their air- plane-manufacturing plants. It reqiiires no special skill or ex- perience in its operation and Jias been found to give results suffi- ciently accurate for all practical piurposes. The design is simple and is based on the fundamental prin- ciples of mechanics. If any balanced system of forces acts at a point, the resultant force is equal to zero by definition. If a wire is supported at two points and deflected at the middle of the span thus formed, we have a system of forces in equilibrium. Knowing the length of span, the amount of deflection at the center, and the force required to produce this deflection, it is possible to calctdate the tension in the wire. The tensiometer provides a convenient method of obtaining the necessary data for this solution. The wire is supported on two pins at a known distance apart and loaded at the center of the span thus formed. Load is applied through a yoke, spring, and plimger by the closing of hand levers. The required deflec- tion of the wire is indicated by an Ames dial and the corresponding lateral forces exerted on the wire are shown by another dial. We know that the tension in the wire is proportional to the load on the spring for a given amount of deflection and it is therefore possible to make the second dial indicate the tension directly by always deflecting the wire a constant distance. The design of the instru- ment is such that estimated divisions on the second dial indicate pounds of tension on the wire for the standard deflection of one- tenth inch (or one revolution of the first dial). The instrument is portable, quick reading, and does not have to be attached integrally to the wire at any time. The instrument 40 WAR WORK OP THE BUREAU OF STANDARDS can be constructed so as to give direct load readings for the type of wire most commonly encoimtered in the construction of the planes on which it is designed to be used. For shapes and sizes of wires diEfering materially from those for which the instrument was designed, it is necessary to apply correction factors. Instru- ments have been built for use on cables of various sizes and on wires of circtdar and stream-line shapes. The Bureau was subsequently called on to conduct calibrations of other types of instruments designed for this purpose, but these were found to be more dependent than the tensiometer on the skill of manipulation and the position in which the instrument was applied. They also suffered from unsatisfactory mechanical construction and in some cases were based on incorrect principles. Most of them used varying deflections of the wire for different ten- sions on the wire. This prevented the instruments from being self-indicating, and calibration curves or tables always had to be used to enable the observer to determine the tension. Synchronizing Devices for Airplane Gtms Tests were made upon a hydraulic synchronizing system such as is used in synchronizing machine-gtm fire from airplanes. These tests were made in order to determine the effect of varying the size of different parts on the operation of the mechanism. The operation of the gear may be described as follows: The necessary pressure in the system is effected by means of oil in a high-pressure plunger tube and is maintained by a reservoir plunger spring. A needle valve operated from the airplane control stock closes the opening to a low-pressure tube and at the same time opens a ball valve, which in turn releases the pressure to the auxiliary pipe lines. This pressure forces a hardened steel ball attached to the end of a generator plunger against a cam which is mounted on a rotating axle. When the cam is rotated so as to move the pltmger, the oil in the system is forced through two pipe lines — ^an auxiliary and a main — against the plunger in the reser- voir, and against the base of the trigger motor. The pressure per unit area exerted by the reservoir plunger is not Sufficient to produce motion of the trigger-motor plunger. Hence, if the cam is rotated slowly enough to allow the oil to flow through the auxiliary pipe line, the trigger motor will not operate. How- ever, as the tube connecting the generator and reservoir is very small, the viscous forces are so large that considerable resistance is offered to the rapid movement of liquid through this tube. Miscellaneous Publications, Bureau of Standards, No. 45 ..J_ Fig. 7. — Tensiometer for measuring the tension of airplane wires There is a correct tension for every wire used in the "standing rigging" of an airplane. The tension on a wire is shown in a few seconds by this instrument Miscellaneous Publications, Bureau of Standards, No. 46 Fig. S.—Two views of the Dover By-Product Coke Co. 's plant at Canal Dover, Ohio The bureau, in cooperaUon with the Bureau of Mines, carried on extensive tests on the coking of low-grade midcontment coal in the Roberts ovens installed at this plant AIRCRAFT (miscellaneous) 41 Hence, when the cam is rotated rapidly, the resistance in the auxiliary pipe due to viscosity is so great that the liquid is forced through the main pipe against the base of the trigger motor with sufficient force to operate the plunger of the trigger motor. The minimum speed at which the trigger motor will function will depend not only on the relative sizes and lengths of the two pipe lines, but also upon the strength of the plunger spring and the trigger-motor spring. When the speed is sufficiently high to make the operation satisfactory, there is comparatively little motion of oil in the auxiliary pipe line. Attention may then be concentrated on the main pipe line. The movement of the generator plunger during the rotation of the cam causes an increase of pressure in the main pipe line. This pressure is not immediately communicated to the trigger-motor end of the pipe line, but passes down as a pressure wave. The three factors which affect to some extent the velocity with which this wave is propagated are: 1. The expansion of the copper tube of the main pipe line imder the increased pressure. Since the copper walls are thick (one- eighth inch) , this expansion relative to the compression of the oil is small and may be neglected. 2. The compression of the oil in the main pipe line due to its retardation by viscous forces which are developed when a liquid is forced at a rapid rate through a small tube. Computations of the pressure thus developed at the generator plunger by a speed of 1200 rpm show that the pressure at this end of the main pipe line for a 6-inch pipe is not in excess of 10 pounds per square inch over that of the motor end. 3. The compression of the liquid due to its inertia. Computa- tions of the pressure due to the inertia of the liquid show that at the same speed the pressxure at the generator end of the main pipe Irae exceeds that of the motor end by 4000 pounds per square inch for a 6-foot pipe, and 6600 pounds per square inch for a lo-foot pipe. The inertja effect may then be considered as the important factor in preventing the instantaneous equalization of pressiure in the main pipe Une. This pressure wave is propagated in the same manner as a sound wave; therefore at approximately the same velocity. This velocity determines the time lag between the generator and the motor, but has no relation to the velocity of the liquid in the pipe. 42 WAR WORK OP THE BUREAU OF STANDARDS The effect of changing the various elements of the system is as follows: I , , ! , n r. Increasing the strength of the spring in the reservou- increases the pressure in the system, and, therefore, puts a greater initial pressure on the spring of the trigger motor. ' It also has the ad- vantage of compressing any small air bubble. 2. Changing the size and length of the auxiliary pipe line. De- creasing the size and increasing the length increases the resistance offered to the flow of the liquid, hence the slower the speed at Which the apparatus will operate. If the auxiliary pipe line be- comes too short, the oil system may have very little damping, so that chattering of the plunger of the trigger motor may occur 3. Size of the main pipe line. Increasing the main pipe lines does not materially affect the rate of propagation of the wave; but, since it increases the mass of the liquid, the strength of the trigger-motor spring must be increased in order that the plunger of the generator may at all times be in contact with the cam. 4. Length of the main pipe line. Increasing the length of the main pipe line increases the time for transmission of the pressure wave and also increases the mass of the liquid. In case the pipe is too short, chattering of the trigger-motor plunger occurs, due to the very small amotuit of damping. In the results obtained some chattering has occurred in most cases. This, however, is not excessive, and it is not believed to be sufficient to be the cause of stray shots, since the extrusions are too small to cause firing of the gtm. From the discussion above it will be seen that the factors enumerated will determine a suitable design for a given installa- tion. The experiments performed from which the above results were obtained are described below. Two methods of measurement were used — the rotating drum method and the stroboscopic method. In the rotating-drum method the movements of the pltmger of the trigger motor were recorded on a rotatmg drum operated by the same motor as the generator cam. A stylus attached to the trigger motor recorded its movements on a paper carried by the rotating drum. In the stroboscopic method two disks were used, one being mounted on the end of the motor shaft and the other near the first on a shaft coaxial with the motor shaft, which was rotated by AIRCRAFT (MISCEI.I.ANEOUS) 43 hand. On the latter disk was a scale graduated in degrees and read by means of a stationary pointer. To make an observation the position of the pointer was read when the highest part of the cam was pressing against the ball of the generator plunger. The motor was started, the speed of the motor determined, and the fixed disk rotated until the trigger-motor plunger was at its maxi- mum extrusion, when the position of the pointer was again read. The time required for the rotating disk to turn through the ob- served angle of rotation of the fixed disk is the time for the trans- mission of the pressure wave from the generator plunger to the trigger-motor plunger and for the trigger-motor plunger to act. Measurements taken in this manner give values for the time lag of from 0.002 to 0.004 second. The observations taken were insufficient to determine accurately the velocity of propagation. However, values ranging from 1 500 to 4000 feet per second were obtained. The effects of removing the damping valve from the T-piece upon chattering of the trigger motor were also observed. The experiments were made with the Constantinesco system set up in the regular manner. This is the hydraulic system in most general use and was used in making the preceding tests. It was found that when strong springs were used in both the trigger motor and reservoir a little more chattering occurred with the damping valve removed. In order to determine the effect of using different lengths of pipe line, tests were made with lengths of 10, 6, /^%, and i>^ feet. For all lengths of 4K feet or more, no differences were noted in the trigger-motor plunger. This was the case for various pressures in the reservoir, various speeds of the generator, and for the removal of the damping valve. For short lengths of pipe, the system was found to operate better tmder low pressures, the extrusions of the trigger-motor plunger being exceedingly large for high pressures. With the reservoir plunger spring removed tests were made upon the pressure required to operate the system. This depends upon the strength of the trigger-motor spring and is independent of the length of pipe Une. With a 25-poimd trigger-motor spring, 70 to 90 pounds pressure was necessary to operate the system, increasing for a 50-pound spring to 100 to 125 pounds. The effect upon the trigger-motor action of varying the speed was also studied. It was found that the initial parts of the im- pulses were very much alike, also that they occurred at approxi- 44 WAR WORK OF THE BUREAU OP STANDARDS mately the same place on the drum. However, at 1200 rpm the plimger remained at its maximum extrusion for a longer time. This is caused by the fact that the trigger-motor plunger is re- turned by the force of the spring. Some general observations were made upon the working of the system as to the number of bursts, effect Of air, and kinks in the system. With the Constantinesco system of gears, the average of several trials was 24 bursts. No direct efifect could be noted as a result of kinking the pipe. However, kinks in the pipe make it hard to free the sytem from air. The washers used during the tests were of English composition and were found to be very satisfactory. Both Enghsh and American made reservours were used. The valve fittings on the EngHsh reservoir gave less leakage trouble. An English generator was used in making the tests. It worked very satisfactorily and showed no signs of wear. In order to determine something of the behavior of the oil when the system was being operated, the top half of the pipe line was cut away for a distance of about 3 inches, and a piece of glass was slipped over the hole. The motion of a rider placed on a fine copper wire stretched along the slit was then observed. The observations showed that in the transmission of the generator- pluaiger action to the trigger motor, there was a movement of the oil as well as a compression wave. The oil used in the system during the experiments consisted of about 9 parts kerosene and 1 part machine oil. A test was also made as to the effect of using Liberty-motor oil in the system. No difference could be noted in the operation when this oil was used. A. ^eat number of tests were made with the Marlin trigger motor. , For an indicator a small brass strip was soldered directly to the trigger-motor plunger. For pressure tests, variation of speed, and lengths of pipe Une, the same results were obtained as with the Constantinesco system. However, the construction of the trigger motor permitted a greater motion of the trigger-motof plunger, and for rather great pressures, very great extrusions of the trigger motor were obtained. The tests made with the system are not, however, conclusive, due to the fact that the complete Mariin system was not operated together. The generator used belonged to the Constantinesco system. The difference in this generator from the Marlin, quite likely, would influence the results. AIRCRAFT (miscellaneous) 45 On June 17, 1918, an officer who had had extensive experi- ence in flying on the European battlefields visited the Bureau and pointed out the need of a synchronizing gear or gun-control system, which would permit of the gun being fired through the propeller in any direction at the will of the operator. He was of the opinion that an electrical type of gear would most easily meet the requirements imposed by the movable-gun featture. On July 24 a communication from the office of the director of miUtary aeronautics requested that the work on this electrical synchronizing gear be carried to completion. After several conferences and some preliminary work the problem was taken up actively about the middle of August, 191 7. At that time, a modification had been made in one of the heavy Brpwning guns under the direction of two officers of the Ordnance Department. The modification replaced the regular sear and trigger by a sear opa-ated from above and back of the bolt by a link which moved in a direction par^jllel with the barrel of the gun. The purpose of this modification was to get a more direct action between the sear and an electromagnet or solenoid with which it was proposed to fire the gun. If the gun is to be fired through the propeller in any direction at will, the phase of the control mechanism must be changed as the position of the gun is changed, otherwise adjustments which are correct for one position of the gun would not be correct for other positions, and the propeller blades be struck by projectiles. As the position of the gun is changed, the control apparatus should change the phase of firing by the same amount that the line of fire is changed with respect to the axis of the propeller. : With most synchronizing systems, adjustments are made so that the gun fires only a safe distance back of the propeller blades with the propeller at rest, or turning with a fairly low speed; then as the speed is increased the gim fires further^and further back of the blades, and at the highest permissible speeds may be firing just ahead of the next blade. With a propeller speed of 1800 rpm the blades rotate through an angle of about 11° in o.ooi second and with the two-blade propeller the safety zone between blades may be considered to be about 145". Consequently, if the gun is to be fired at any speed from o to 1800, rpm it is necessary either to advance the phase of the control mechanism as the speed increases or make the over-all time lag less than o.oi 3 second. Any arrange- ment which would automatically change the phase of firing with 46 WAR WORK OP THE BUREAU OF STANDARDS changes in speed Would add to the complication of the apparatus, and thus should be avoided if possible; We may, therefore, con- sider that we must keep the over-all time lag less than o.oi 3 second. The overfall time lag depends tlpon the time required for the separate operations which must occur in succession between the beginning of the actuation of the control mechanism and the passage of thfe projectile throiigh the plane of the propeller. It includes the time required for the action of the control motor (the timing apparatus connected directly or indirectly to the propeller or crank shaft), the time required for this action to be transferred to the trigger motor (the apparatus which actuates the trigger, or sear, and releases the firing mechanism of the gun); the operation of the trigger motor and sear; the operation of the firing mechanism or firing pin of the gun, and the explosion of the charge and the passage of the projectile through the barrel of the gun and through the plane to the propeller. The time requited for the last-mentioned operations depends upon the ammimition and position of the gun on the plane. With the regular ammunition and the gun at a distance of 6 feet from the propeller, it amoimts to about 0.0035 second. It can not be made much less than this, since usually the gun can not con- veniently be placed much closer to the propellet, and it is not practicable to obtain quicker action ammunition. The time required for operation of the firing pin of the Browning heavy field gim is about 0.006 second. With this gun set at a distance of 6 feet from the plane of the propeller, we would have onlj* 0.004 second as the maximum possible time for the action of the entire control system, unless the phase of the control apparatus were to be advanced as the speed of the propeller is increased. With the Browning aircraft gim, the time required for the opera- tion of the firing pin is probably about 0.00 1 5 second; consequently the permissible time for the operation of the control system is about 0.009 second. A preliminary study of the problem showed that it would not be practicable to fire the heavy Browning gun as modified with any type of solenoid or electromagnet. The difiSculty with apparatus of this type is that it would be entirely too slow in action to meet the conditions of the problem, since the inertiat of the sear and sear trigger together with that of the armature or plunger would necessarily be large, and the force which could be produced Would be small. This preliminary^ study showed fiu-ther that if an electrical control system were to be considered, AIRCRAFT (MISCEI.I small quantities of cobalt and nickel is required in the analysis of some of the new alloys, and methods are being studied to this end. 64 WAR WORK OP THE BXIREAU OP STANDARDS Platinum Metals and Platinum Substitutes A chemical laboratory for the study of platinum and the metals associated with it was organized in this Bureau in July, 191 7. In conjunction with other laboratories of the Bureau its purpose was to prepare the pure platinum metals and their alloys, study their chemical and physical properties, and develop the analytical methods for the group. Opportunities for rendering assistance in the prosecution of the war soon became a.pparent. Work for the nitrate division of the Ordnance Department of the Army came first in order and soon occupied a position of first importance. Platinum wire woven into gauze was used in large quantities as a catalyzer in the oxidation of ammonia to nitric acid. Much experimental work as to the size of wire, mesh of gauze, and purity of the platinum necessary to obtain the highest efficiency had to be carried out. The first two of these items were handled principally by the nitrate division, but several of the gauzes used were examined by the division of weights and measures of this Bureau for size of mesh and regularity of weave. There was considerable uncertainty as to the effect of small amoimts of rhodium, iridium, palladium, iron, and other metals alloyed with platinum on the efficiency of the latter for catalytic purposes. It was claimed by some authorities that the platinum must be specially free from iridium. As this eleinent, together with rhodium and iron, is almost always present in commercial platinum, it was necessary to be able to obtain accurate analyses of the platinum used. The methods used were based for the most part upon those employed by Deville and Stas in the analysis of the platinum-iridium alloy used in the international standard mete. Their method gave very accurate results for iridium. The separation of rhodium and platinum by potassium pyro- sulphate fusion proved extremely tedious, and a primary separa- tion by distillation in a current of carbon monoxide and chlorine, as recommended by the German Reichsanstalt, was later adopted. Iron was best determined colorimetrically on a separate sample. Complete analyses were made on several gauzes the efl&ciency of which had been previously determined. Some iridium deter- minations were made on material refined for the nitrate division by the New York Assay Office. An attempt was made by the optical division of the Bureau to determine small amounts of impurities in some of the samples submitted by the nitrate division by spectroscopic methods. CHEMICAL INVESTIGATIONS (MISCELLANEOUS) 65 Owing to the lack of standard alloys this proved unsatisfactory. Occasionally use was made of a rapid methbd for an approximate determination of the total impurities which was developed by the metallurgical division of the Bureau (Scientific Papers Nos. 254 and 280). This method depends upon the emf developed at a hot jtmdtion of a platinum alloy with pure platinum. Some of the determinatiohs were made by the metallurgical division. Another section of the last-named division made metallographic examinations of some of the gauze before and after being used in the oxidation of ammonia. The results obtained were helpful in determining the phenomena which occurred in the catalytic process. In addition to the work already described the chief chemist of the Bureau served the nitrate division of the Ordnance Depart- ment in an advisory capacity on numerous questions relating to the sources and supply of platinum and the assay of platinum Ore received from Russia. The nitrate division in turn cooperated with the Bureau by detailing a member of their chemical staff to work in the platinum laboratory and furnishing fimds for part of the necessary equipment. ^ In the latter pajrt of 19 18 problems of a similar natvire to those already outlined arose in the Bureau of Aircraft Production. One conlplete analysis was made on platinum-iridium contact points for airplane magnetos. Later the electrical division of the Btu-eau undertook a study of the relative eflSciency in service of magnetos equipped with contact points of varying iridium content. This was done in order to determine the possibility of econolinizing on iridium, the available supply of which was very low. Magnetos for several different types of engines were tested. Determinations of the iridium and iron content were made on all tbe types pf contact points used. It was found that lthe iridium content fell somewhat over 50 per cent short of specifications in each type. , This rendered the results pf the whole series of tests only partially tfseful. Tl^e , manufacturer of the contact points proposed to, make up a new set of alloys, but the second series of service tests was not xmd^rtaken owing to the termination pf hostilities. The shortage of platinum a short time before and dining the war aroused the interest both of commercial prganizations and the public in general. Quite a number of "platinum substitutes" were developed, and many samples of supposed platinum ore were 26035°— 21 5 66 WAR WORK OP THE BUREAU Olf STANDARDS submitted to various Gbvernnlent laboratories. A, few of the latter were examined by this Bureau. None of the samples sub- mitted showed any platinum. The most interesting case was of a mineral from the Grand. Canyon, which was repeatedly brought forward as a source of platinum in spite of the statement from the Geological Survey that it contained no precious metal. New sam- ples were taken by the Survey and assayed in their own laboratory; that of the Mint, and of this Bureau. No platinum could be found. A more detailed report of this investigation is given in Mining and Scientific Press (118, p. 185; 1919). The substitutes proposed for platinum were of two kinds, those composed of base metals and alloys of the noble metals. Of the latter, two were examined, known as rhotanium and palau. Both are alloys of gold with relatively small amotmts of palladiuifi. These were tested in conjunction with the metallurgical division of the Bureau and were foimd suitable substitutes for platinum in numerous operations in the cheriiical laboratory. A detailed report of these tests was published in the Journal of Industrial and Engineering Chemistry. - In the case of the base-metal substitutes none was found which approached platinum in general usefulness. In one case steps were taken to suppress extravagant advertising. Two alloys were examined which were found suitable for the purposes for which they were designed. One was an alloy very resistant to nitric add and the other could be used for electrical contact points under a limited range of conditions. The former was submitted by the nitrate division of the Ordnance Department, but the greater part of the work on platinum substitutes was done for the platinum section of the War Industries Board. Prevention of Corrosion The work of the Bureau has included a study of the corrosion of metals and its prevention. The tests were generally made by the salt-spray method, originlally described by J. A. Capp (Proceed- ings of the American Society for Testing Materials, 14, part 2, p. 475) and greatly impt'oved in the Bureau's laboratory, as described in Proceedings of the American Society for Testing Materials (18, part I, p. 237, I 918). It is thought that this test indicates, in an accelerated wa:y, the relative resistance to corrosion of metals, especially when exposed to saline conditions. Tests oh corrosion of aluminum, imdertaken chiefly in the interest of aeronautics, were made on alloys of aluminum With copper, nickel, matiganesfe, and CHEMIGAI. INVESTIGATIONS (MISCELLANEOUS) 67 magnesium prepared by the metallurgical division of the Bureau. The following conclusions were drawn : (i) A decided difference in resistance to corrosion may be pro- duced by quenching some aluminum alloys ; (2) a less marked dif- ference is produced by annealing ; (3) with some alloys no apparent difference is produced; (4) if any change is produced by quenching, it improves the quality of the metal; (5) the magnesium, nickel, and nickel-magnesium alloys have about the same resistance to corrosion regardless of heat treatment; (6) annealing improves somewhat the quality of the copper-magnesium alloys and reduces the quality of the manganese-magnesium alloys; (7) quenching produces the most desirable effects in the copper, copper-magne- sium, and manganese alloys; (8) commercial aluminum, as rolled, does not resist corrosion satisfactorily, and the sample tested was almost completely disintegrated at the end of the test; annealing or quenching materially improves it, but it is not equal to some of the alloys. ' . In developing the data on which these conclusions were based and in other tests, it is noteworthy that ordinary sheet aluminum as rolled frequently exfoliates, showing a laminated structure in- stead of a homogeneous mass. The various types of zinc coatings were tested and no evidence was obtained to indicate superiority of hot galvanizing, sherardiz- ing, or plating. The use of zinc was recommended for protection of iron or steel against corrosion wherever its application was possi- ble, the value of the coating (or protection afforded) being judged by its life in the salt spray or amount of zinc applied per unit area, depending on conditions of exposure anticipated in service. This report is covered in more detail in the report on " Protective Coat- ings," appearing elsewhere. Corrosion produced by the use of soldering fluxes was also con- sidered. For general purposes the use of zinc chloride fluxes seems to be almost universal, and although they will undoubtedly accel- erate corrosion, their use seems to be unavoidable. Rosin was recommended whenever it could be used, and if zinc chloride was used, thorough washing with hot water was advised. Examination of hydroplane radiators indicated that leaks which develop in storage are caused by corrosion accelerated by the zinc chloride flux which was left on the metal. Washing inside and out with hot water and sodium carbonate solution was recommended as a probable means of correcting this difficulty. 68 WAR WORK OF THE BUREAU OF STANDARDS Removing Metal Fouling from Rifle Barrels Prior to the war, a solution made according to the following formula was used for removing the cupro-nickel fouling from rifle and machine-gim barrels : Ammonium persulphate ounces. . i Ammonium carbonate: i grains. . 200 Water.. ounces. . 4 Ammonitmi water (28 per cent) "°- ■ ■ ■ , ' In September, 191 7; the office of the Chief of Ordnance of the Army found thiat not enough ammonium persulphate was pro- duced, in the enture country to supply the needs for cleaning rifles, and requested this Bureau to recommend a substitute. At this time there appeared to be only a shortage of ammonium persulphate, and numerous experiments were made in trying to find a satisfactory substitute for this chemical. Nothing was found which at all met the requirements. While these experi- ments were still in progress it developed that the supply of am- monia water was becoming depleted and that no further sup- plies of this substance could be relied upon. At this stage of the investigation^ an electrol5rtic method for cleaning the barrels was successfully developed. This was quite simple and consisted in using an electrolyte made of a mixtmre of sodium carbonate and ammonium sulphate dissolved in water, a steel rod being used as the anode and the rifle barrel forming the cathode. The experiments with-this method demonstrated that in using a current of about 0.5, ampere and from 4 to 5 volts, cleaning was more rapid by the electrolyiic method than by the use of the old persulphate solution. It is believed that the ordnance officials developed an appar ratus for generating the required electric current for this work-, and it was submitted for approval. The electrolytic method had the objection Of requiring a source of electric cturent, either from a central station or from batteries, and hence could not be satisfactorily used except at base stations. The method would, however, probably have, been introduced had it not developed — middle of Decembetj 1917^-that a sufficient supply of ammonium persulphate would be available. The ordnance officers deemed iti advisable, however, to elimi- nate the use of liquids which wodld have to be transported, thus necessitating avoiding the use of ammonia water. After making numerous experiments, it was found in December, 1917, that a mixture of 3 grams of ammonium persulphate, 5 grams of am- CHEMICAI, INVESTIGATIONS (MISCELLANEOUS) 69 monium sulphate, and 2 grams of caustic soda dissolved in water to make 35 cc was quite efficient in dissolving cupro nickel and had no appreciable injurious effects on the steel. Such a mixture had distinct advantages; the reagents could all be kept in a solid form, and it was suggested that the ammonium perstxlphate and ammonium sulphate cotdd be prepared in solid tablets which could be kept in glass, metal, or cardboard containers. The sodium hydroxide could also be kept in metal or glass containers. It would then be very simple to make up the solution by dis- solving the requisite number of each kind of tablet in water and mixing the solutions for use. Ordnance officers took up the question of: making contracts for tablets and found that it would be necessary to use a small amount of, binding material, such as sugar, and a lub^cant shnilar to stearic acid or paraffin, in the manufacture of the tablets. The amount of these additions was small, and specifipaiions were finally drawn up for these tablets and have beeji issueci by the War Department. Soap ' A number of different brands of shaving soaps (including stick and cream) were examined by the Bureau for the purchase, storage, and traffic division of the War Department during; May and Jime, 1918. Following this examination, specifications for stick shaving soap and shaving cream were drawn up by the Bureau and submitted to the War Department. Numerous analyses of: laundry soap have been made for the same department during the same period. Carbon Tetrachloride Fire-Eztinguishing Liquid At the request of the War Industries Board and the Quarter- master General of the Army experiments were made to find a substitute for chloroform as freezing-point depressant in carbon tetrachloride to be. used as a fire-extinguishing liquid. It was found that gasoline and turpentine were both efficient freezing- point depressants of carbon tetrachloride, and that the resulting mixtures were fire-extinguishing liquids. The signing of the armistice removed the demand for conserving chloroform, and hence the mixtures suggested were apparently not used. Antifreezing Solutions for Automobile Radiators The question of a satisfactory antifreezing solution for radia- tors used in connection with gasoline motors was studied. The chief constituent of all antifreezing compounds tested was cal- 70 WAR WORK OF THE BUREAU OF STANDARDS Gium chloride, and numerous tests showed that the corrosion produced bytheir use was decidedly rapid, particularly at soldered joints in radiators and on all-aluminum parts which are used in some engines. Whenever solutions of calciiun chloride get on the ignition system, either from leaks or overflow, electrical leaks or short circuits are produced which interfere with engine performance, and such electrical leaks are very difficult to correct. The use of alcohol and water was recommended for trucks, automobiles, etc., but such solutions have a pectdiar limitation when used on aircraft, that of a low-boiling liquid for high altitudes. Tests were undertaken and are still in progress to determine if the addition of glycerin to such a solution will overcome this objection- able feattire. Prevention of Foul-Gas Formation in Ammonia- Absorption Refrig- erating Machines A problem started before the war, but whiph assumed added significance shortly after the beginning of hostilities and which was, therefore, pushed to corapletion, was the causes and means Of prevention of foul-gas formation in ammonia-absorption refrig- erating machines. On account of the scarcit]?"' of ammonia during the war and the large losses necessitated in purging a plant of foul gas, the solution of this problem was of considerable impor- tance, as it helped toward the -conservation of ammonia.' It was found that the main causes of this trouble were leaks of gas into the system and the gas produced by the corroding of the metal in the plant by impure aqua ammonia. This latter cause was en- tirely prevented by the addition of sodium dichromate to the charg« of aqua ammonia. The resj^t^,of t;tu$ wprk have^been publish^cl in. Technologic Paper No. i8o, .''Causes and Prevention of the Formation of Noncondensibl^ Gas^, .in Amin,o|da,, Absorption Refrigeration Machin,es." .. •■, Development of Automatic' Analytical Apparatus iot Use in Mtrate Plants During the summer and auttunn of 19 18 a force of about 15 men Was engaged upon the problem of designing and building an automatic apparatus for the continuous analysis of the gas mix- tures of widely different character occurring in the contact process for the manufactvure of ammonia. , Thi^, apparatus was regarded as essential to the efficient operating control of the large nitrate plant projected at Miissei Shoals. At the signing of the armistice one imit ojE this apparatus capable of automatically making and CHEMIGAI, INVESTIGATIONS (MISCELLANEOUS) 7 1 recording in series 12 different analyses was nearly ready for installation. A smaller unit for the same purpose was provided for the laboratory engaged upon the fixed-nitrogen research. While this work did not reach the stage of application to the purpose for which it was intaided, the methods developed, which are largely new, shovdd find extensive application in many industries. Hydrogen Detectors for Use in Submarines The problem of designing a suitable hydrogen detector to prevent the accumulation of a dangerous amount of gas in submarines was presented to the Btu-eau as the result of two serious explosions prior to the war. Ten or more devices for the same piu^jose, designed and constructed outside the Bureau, were submitted to the Bureau for tests and criticisms. An apparatus believed to be satisfactory for the purpose was designed at the Bureau and, after extensive laboratory tests, was submitted to the Navy Depart- ment. The tests upon the instrument made in the New York Navy Yard were reported to be entirely satisfactory, and 60 of the instruments were constructed and calibrated. The severity of the mechanical and electrical conditions to which they were to be subjected was underestimated, however, and they did not render satisfactory service. Another type of instrument was then designed especially to meet the most severe conditions of service, even at a sacrifice of several of the desirable features of the earlier device. This instrument has been tested by the Navy Depart- ment and reported upon favorably, but has not yet been placed in production for use aboard ship. The earlier work along this line is reported in Scientific Paper No. 334 of this Bureau. Determination of Carbon Dioxide and Oxygen in the Air of a Submarine Much time was devoted to the problem, initiated by the Navy and transferred to this Biu:6au by the Sanitary Corps, of auto- matically indicating the percentage of carbon dioxide and oxygen in the air of a submarine. This problem appeared to be well on the ■#ay to solution when the Bureau was informed that it was no longer regarded as of much importance, and it was abandoned for more urgent work. Tests of Gas Masks and Oxygen-Supply Apparatus Tests were made and recommendations reported regarding the use of gas-supply apparatus for aviators when more than one man was to be supplied, from the same soiurce. Prior to the organ- 72 WAR WORK OF THE BUREAU OP STANDARDS ization of the Chemical Warfare Service this Bureau also did the preliminary work of examining the gas masks used by European armies. This work included the identification of the materials used as absorbents and testing the resistance to respiration and other mechanical features of the masks. Miscellaneous -^ A great many gasolines and gasoline substitutes were examined in connection with the work on £aij?pliiBerengines. 'Castpr oiis were investigated' quite frequently for the Signal Corps* and mis- cellaneous materials, including neat's-foot oil, saddle soap, and white floating soap, were examined for the Ordnance Department; CHROMATIC CAMOUFLAGE iiND CHROMATICjALLY CON- CEALED) insignla; " Owing to, the fact iiiat different spectral distributions of light may give the same color, it is possible to have samples' accturately color-matched as examined by the unaided eye in day%ht, and still show glaring chromatic contrast; when examined tiirough swtable light filteas (colored glasses, etc.) . This circumstance was a source of danger to the camouflage artist. For example, a fabric might be dyed green to closely match green foliage, and through a properly chosen filtfer it might appear bright red on a dark green or black backgroiind of normally^ green foliage. Indeed, it might be about as conspicuous as a fire at night. On the other hand, it might, through another filter, appear dark on a bright red back- ground of normally green foliage. At the request of the: War Department, the Buneau made several tests of materials in this respect and was able to recom- mend dyes ?Hid select fabrics to overcome this- difficiilty. The Bureau itself did not undertake a comprehensive and unified investigation of the subject of "camouflage." However, a great deal of work bearing on it was done at the. request and instigation of others working on these subjects. A great deal of oral information on ray filters, spectral transmission, aad reflection and chromatic camouflage, pft^j illuslxated by experimental denionstration, was gjlven to officers of the ^nay and Navy and representatives of the National Research Council. Ray filters were lent from stock, and others were made to order of prescribed properties and supplied to the applicants for them. ' The Bureau was also asked to investigate the practical possi- bilities of chromatically concealed insignia ; thatis, insignia woven; painted, oi- made in such a way that they would be invisible to the COKE-OVEN INVESTIGATIONS 73 unaided eye in daylight and be rendered visible by examination through properly selected filters. Many combinations of dyed fab- rics were prepared which were nearly or quite indistinguishable to the imaided eye, but showed remarkable color differences when examined through suitable filters. These were exhibited and ex- plained to numerous representatives of the War and Navy Depart- ments. An exhibit of some of these was made in connection with the American Physical Society exhibit at the Bureau, April 25 to 26, 1919. COKE-OVEN INVESTIGATIONS During the latter part of 191 7 the attention of the Department of Conimerce was directed to the Roberts coke arid gas oven through the efforts made by the American Coal & By-Prbducts Coke Co., the owners and operators of the Roberts pi'ocess, to secure Government Siipport for this enterprise; This matter was referred to the Bureau of Standards for action. Preliminary Work of the Bureau Since there were 24 Roberts ovens in operation in Canal Dover, Ohio, three representatives of the gas engineering section of the Bureau were detailed to inspect the plant at this point and carried on their work on December 12 and 13. TWo views of this plant,' showing the general la,y-out, are presented in Fig. 8. The promoters of this process claimeid a number of advantages for, this oven ^hich it was said were not embodied Jn other t)^^. Among these were the following: (i) Ability to coke so-called noncoking coals and produce a satisfactory grade of metallurgical coke from high-volatile, miidcontinent coals; (2) larger yields of by-products recovered than from other types of ovens; (3) more substantial construction than possible with existing t5T)es of ovens; (4) greater degree of flexibility of heat contrpl than with existing t)rpes of ovens; and (5) increased earning power per oven over the Other types of ovens. At this investigation it appeared that the ovens were operating in a fairly satisfactory maimer, and although the Bureau was not able to verify these cl^ms to any great extent in such a short praiod, the indications were that the oven could be developed iiito a commercial success. Therefore, the Bureau of Standards re- ported to the Assistant Secretary of Commerce on December 15, as follows : We do not hesitate to recommend favorably this type of oven for further considera- tion for we feel sUre that benefit to the Government and the'people may be expected from the commercial development of the process. 74 WAR WORK OI^ THE BUREAU OE STANDARDS Fuel Situation in Early Part of 1918 Early in this country's participation in the war it was realized that a successful end would be possible onlythrbugh the intensive driving of our industries. This became particularly evident dur- ing the early part of 1918, and a great expansion of the steel, industries of the country; resulted. Likewise, it was realized- that high explosives had to foran a large part of the equipment of the American Expeditionary Forces. The rate of production of these necessary materials was limited to a great extent by the rate of the recovery of by-products of coal. The output of steel was cur- tailed because of the lack of iron, which in turn was not produced in sufficient quantities due largdy to the lack of coke. The pro- duction of high explosives was not, keeping pace with the growth of the armed forces of the Nation, because of the lack of funda- mental materials — toluol, benzol, ammonia, etc. In addition to the lack of coke-oven plants, climatic conditions during the winter months of 191 7-1 8 gr?:atly :redijped the coal supplies of the existing plants. Transportation of large quanti- ties of coal became increasingly difficult as the severe winter weather continued and as the railroads became choked with ma- terials moving toward the seaboard. This particularly interfered with the inaintenance of the production at full capacity on the part of the coke ovens in the central section of the coimtry, since most of the e;asting plants in this section were dependent for their coal supplies upon so-called eastern coal. Likewise^ with the increase of traffic more and more coal was used by the rail- roads and the use of raw coal in boilei: plants and in other indus- trial activities, and in the heating of dwellings, ifactories, etci, due to the extreme cold and lack of other fuels, depleted the supply of fuels for these oven?. Accordingly the' utilization of much of the undeveloped fuel resources of the country became a question of paramount impor- tance. Attention was forcibly directed to the large quantities of midcontinental coal available, particularly in Iddiana and Illi- nois, but which was not being Used. Inasmucli as the steel pro- duction of large plants, such as those at Gary, lAd , and Joliet, 111., was sometimes reduped to 30 per cent through lack of coke, it was evident that some relief was essential. Attention Focused on Roberts Oven Siace it was qlaimed for the Roberts oven that it was able to produce a first-class grade of metallurgical coke from these mid- continental coals, it was natiiral that this type of oven should be COKE-OVEN INVESTIGATIONS 75 looked upon as a proper agency to aid in the forestalling of the crias which seemed imminent, as well as to assist in the develop- ment of these hitherto miused coal resources of the country. Therefore, using the Bureau's report of December 15, Mr. Roberts endeavored to interest the Government to aid in the further devel- opment of his process and to construct two plants, one at East St. Louis, 111., and the other at Chicago, 111., which would use coal from the Illinois (midconttnent) field, thus avoiding the long raih-oad haul and accelerating the production of coke, toluol, etc. Test Conducted by the Bureau Since the December 15, 191 7, report of the Bureau was only a recommendation for further consideration, early in March, 191 8, the Secretary of Commerce, acting by direct instructions from the President, ordered the Bureau to conduct an operating test of the Roberts oven installation at Canal Dover and to associate with itself the Bmreau of Mines and the Geological Survey in this work. Accordingly, plajis were immediately made for such a test, and representatives of the Bureau proceeded to Dover on March 18, biit found that it was not feasible to begin the test at the time specified owing to a lack of boiler capacity. The ovens at Dover were operating at this time on Cariibridge (Ohio) coal, which, alth,ough somewhat similar to the midcontinent coals, bears a striking resemblance to the eastern coking coals. It was decided, however, that the test should be condutced with the same coals which would be used at East Sit. Louis and Chicago; therefore the Bm-eau was instructed to assist in obtaining the necessary coal for this test, together with the required boiler parts. This the Bureau proceeded to do, but hindered by numer^ous construction and transportation delays, it was not xmtil April 26 at 7' a. m. that the test was dfficially started. Quantitative observations of all parts 6f the plant operation were undertaken and were continued 24 hours each day, including Sundays, imtil 7 a. m., May ii. During this test 4800 tons of Illinois coal were used in the ovens and all of the usual by-products, including light oil, ammonia, and tar, were recovered and either carefully weighed or measured; The coal used was shipped froni the Royalt©n, Sesser, and Orient mines in Franklin County, 111., and the Ayrshire mine. Pike County, Ittd. This work was conducted by 22 repre- sentatives • of the Bureau of Standards and 9 members of the regular staff of the Bureau of Mines; 5 consultiiig engineers were also present at the request of the Bureau of Mines to advise con- cerning the work. 76 WAR WORK OF THE BtJREAU OP STANDARDS The Bureau's representatives were in charge of the test and respon3ible for observatio^s of all work done about the coke oven, the by-product department, theiUght-pil department, and other accessory operations. Exact records were made of the charging and discharging of the ovens, the quantities of. tar, ammonia liquor, light oil (containing benzol, toluol, etc.),.aaid gas produeed during the test period. Samples of all these materials were regularly taken,, and analyses or t^ts of these gamples made; at appropriate intervals. , Likewise, high-temperature measurements at various parts of the battery were carried on during the test. The Bureau of . Mines was responsible for the samp%g of the coal at the mines, at which time record w:as made pftbe car num- bers and weights of the coal shipped for t^fe test. . The Bur^u of iSiaes also weighed and sampled for analysis the coal as it was iiniqaded in i)oyer for the test, thus identifying by car numbers and car weights the origin of the fu^l used.^ The Bureau of Mines also was responsibly i(^r the samplmg, weighing, and analysis pf the coke.prqduced from the test coal. Because of mechanical difficulties tlie regenerators were liot tfsed during the test; therefore the ovens were not operated upon preheated air, but upon cold air. This precluded the making of any surplus gas, as it was necessary to use all that was qbtained in heating the pvens. Difficulty was also expjerienced in the uniformity of the heating of the oven walls, the ends particu^rly being npticeably cold- Results of Test at Dover The coke produced frpm three of the test cpals was satisfactoiy as to quantity and, quality, the yield beng substantial^ the same as that obtained from other by-product ovens. On the basis of dry coal, the yield of dry coke was 69.5 per cent, and the percentage of the sizes of dryqpke to tptal coke were furnace. size 75.1, nut 16.8, and breeze 8.1 per cent. A sufficient amount of coke was not produced by the ovens to operate ;the 500-ton blast furnace adjacent to the plant without the use of additional coke from other sources. Ssace the furnace was not able to operate at more than three-fpurths capacity, and, since thje quajatity of coke had to be increased by the addition of outside qoke to the extent of from 30 to 50 per cent, it was impossible to determine the real metallttr- gical value of the coke. However, the opinion of the superinten- dent of the blast furnace was that when 100 per cent of this Illinois coke was used it would make a satisfactory blast-furnace fuel with the furnace at full capacity. COKE-OVEN INVESTIGATIONS 77 The yield of toluol both as to quantity and quality was good and amounted to about 0.43 gallon per ton of dry coal, and the yield of benzol amounted to 1.55 gallons per ton of dry coal. The yield of other by-products per ton of dry coal were as follows: Ammonium sulphate, 26 pounds; tar, 10.7 gallons; and gas about ID 000 cubic feet. However, on the basis of the quantity of these materials actually in the gas, the yields were somewhat higher than those quoted. Conference at Pittsburgh After the test at Dover was Completed, a conference, which was arranged by the Bureau of Mines, was held in Pittsburgh on May 2 2 , and the observations and results of the tests were discussed at length. Substantial agreement was reached as to what had been demonstrated by the test and what the results signified, and a report of this conference was made to the Secretary of Commerce on May 25, which was signed jointly by the Directors of the Bureau of Standards and the Bureau of Mines. The report was as follows: The Roberts oven produced a satisfactory grade of metallurgical coke from three of the four coals used. These coals have not heretofore been used to make metal- lurgical coke in commercial quantities. The gross yield was substantially the same as that obtained by other by-product ovens. The coke was mixed with about an equal amount of coke from oth?r sources, because of the limited supply of Dover coke, and was used in a blast furnace running at about three-fourths capacity. A satis- factory grade of -pig iron was produced. The superintendent of the fiimace is of the Opinion that he could operate satisfactorily with this coke alone and at full capacity. The quantity and quality of the gas produced was good. After removing the by- products all the gas was used in the process of coke rilaking. An excess of gas will be available when regenerators are in successful operation. The quantity of by-products.in the gas was good. The amount of tar, toluol, benzpl, etc., iri the gas was about the same as is obtained in other ovens when high- volatile coals are used and was greater than the average yield from low- volatile coal, or the iJSual mixture. , T^he amount of ammonia in the gas was very good. , ,. The construction p£ th?; ovens js substantial. The heating system of the, Dover plant Ifi^ves something to be desired in the matter of uniform heating of walls, and the. regenerators for preheating air have not been in operative condition, E;xperience with the plant shows it can be improved in some particulars, and the company believes successful preheating of, air and uniform heating can be demonstrated in a short time. It was necessary to run the test without regenerators, and, as their use is essential for most efficient op^ation under usual conditions and in normal times, this further detribnstration is believed to be necessary. The. Roberts Co. immediately proceeded to carry out the agreed program for a further demonstration, but great mechanical diffi- culty was experienced in putting the ovens into shape for this purpose, and it was not imtil the Iktter part of September that it was possible to complete the work. Further Work by the Bureau The Bureau wished further to study the effect of temperature distribution in the various types of ovens, so that the feature of 78 WAR WORK OF THE BUREAU OF STANDARDS greater umformity and flexibiUty of heating, which Mr. Robais claimed for his oven, could be either verified or disproved. Th^e- fore a separate series of high-temperattare measurements were con- ducted at Canal Dover during June, and similar tests were likewise carried out in the Koppers oven installation of the Laclede Gas Light Co., St. Louis. The Bureau has also made an investigation of the existing patent literature and other literature on the subject of coke ovens, and prepared a report showing the development of the art of by- product coking of coal. This report has been given to several persons for criticism. Interest in the Use of Midcontinent Coals Increased During the period from May to October, 191 8, other coke-oved companies had taken vinder serious consideration the use of these midcontinent coals in their particiilar types of ovens and had car- ried on experimental work. As a result, the Bureau recdved a request from the Koppers Co., of Pittsburgh, to conduct an operat- ing test of the Koppers oven installation in the plant of the Minne- sota By-Product Coke Co., St. Paul, Minn,, which would be similar to that conducted at Dover in May, At, this time the Bureau was advised that 600 000 tons of In- diana coal had been used in the Gary, Ind., coke-oven plant, and received a report which referred to this occassion from G. C. Lowell, of Chicago, who mentioned it in the following manner: During the latter part of the year 191:2, when there was a serious depression in the steel business, it was decided to charge a portion of the ovens at the Gary plant with 100 per cent of western coal. This coal was what was known as Clinton coal, coming from the Universal mines, located near Clinton, Ind. There was also a number of ovens charged with ioo per cent of coal from the Kelly mines located near Westville and known as Westville coal. This coke, made from 100 per cent of Indiana coal, was first produced with an idea of supplying a cheap coke to the blast furnaces at Gary that were isiipplying gas to the electric generating station, and no particular attention was paid to the yield of iron, it being attempted to hold the iroa ptoduction to as low a figure as possible and produce a maximum quantity of ^as. The coke was used in this manner for a few months, and when the iron production was again increased by the blast furnaces one battery of ovens was kept on the Clin- ton coal, and the coke produced from this battery was mixed with the regular coke produced at the Gary plant and charged into the blast furnace as mixed coke, Using up to about 15 per cent of the Clinton coke in the blast-furnace charges. Fairly good results were obtained from the blast furnaces. Test of Koppers Plant at St. Paul Since such a long delay was experienced in placing the Roberts plant at Canal Dover in operating condition, the bureau acted upon the request of the Koppers Co. and secured permission from the President, through the Secretary of Commerce, to conduct an COKE-OVEN INVESTIGATIONS 79 extensive operating test of the plant at St. Paul. The Bureau of Mines likewise cooperated in this test, which was officially started at 7 a. m,, September 27, 1918, and continued until 7 a. m., Odtober 5, 1918. The test of this plant, consisting of 65 Koppers coke ovens and all accessory equipment and apparatus for the recovery of all byr products, was conducted by 37 engineers and chemists, and in addition a considerable number, of expert engineers visited the plant while the test was' xmder way. ,The Bureau of Standards was responsiblefoi: the general planning and supervision of the test yvprk,. Its representatives made all obsarvations of battery operation, high-temperature measurements, by-products recovery, and chemical laboratory work on gas and by-products. The Bu- reau of Mines was responsible for the sampling of the coal at the time it was loaded at the mines and for the weighing, sampling, and analysis of the coal as used at the plant, and of all sizes of coke produced. Its representatives also made general observa- tions on the character of the coke and the operation of the ovens. For this test 7600 tons of coal were used from the Orient mine, Franklin County, 111. The quantity of all products was carefully weighed or jneasured at regular intervals, samples of each of the products were taken, and complete analyses of the coal and these products have been made. Results of Test at St. Paul The yield of dry coke was 68.4 per cent, and the percentage of the sizes of the coke on the basis of total coke were: Furnace, 56; domestic, 37.6; and breeze, 6.4 per cent. The yield of toluol amotmted to 0.54 gallon per ton of dry coal, and the yield of braizol to 2.29 gallons per ton of dry coal. The 3deld of other by-products per ton of dry. coal were as follows: Ammonium sulphate, 30.3 pounds; tar, 8,5 gallons; and gas, about 1 1 000 cubic fept. Further Work at Dover Immediately upon the completion of the St. Paul test, the members of the Bureau's party proceeded to Dover for the final demonstration of the Roberts process, with the regenerators operating. Since the War Industries Board had requested that a complete operating test be made with mid-continental coal, 1500 tons of Orient and 1000 tons of Ayrshire coal were secured for this test, and arrangements made whereby the coke would be stored until a sufficient amount had been accumulated, so that the blast 80 WAR WORK OP THE BUREAU OP STANDARDS furnace could run for several days at full capacity on loo pet cefit Illinois coke. . The test began on October 12, with the same methods pursuM as were carried on in the May test, but before evening on October 14 the representative of the Roberts Go. reqtitested that the demonstration be stopped, since the quality of the coke and the condition of the plant showed that it was not possible to make a successful demonstration at this time. The Bureau acceded to the request, and after conference with the War Industries Board, decided that until the company could demonstrate that the Roberts ovens could cokfe 480 tons of cdal per day, the Bureau would not consider another test. Blast-Furaace Test on St. Paul Coke During the latter part of the work at St. Paul, arrangements were made to ship 1500 tons of the fiu^ace coke produced to the blast-furnace plant of the Mississippi Valley Iron Co., St. Louis, Mo. Soon after begiiining the use of the test coke in the blast furnace an additional quantity of about 360 tons was ordered. Of the total tonnage ordered, 1746 tons were actually used in the blast-furnace test. A representative of the Bureau of Standards was present during the period of the test, which started with the first charge of Illinois cOke into the blast fiurnace at i p. m.., Octo- ber 19, and continued until 4.30 a. m., October 28. This furnace was operating at a capacity of about 190 tons of iron per day and at the beginning of the test was making malleable iron. However, it was recognized that the most con- clusive test of the merits of the coke would be obtained in making a run on basic iron. . . , ' From' this test the following general conclusions were drawn by F. W. Sperr, jr., chief chemist of the Koppers Co., w^ho was present throughout the test: 1. The coke was fiist-class blast-furnace fuel, which ciauld be substituted directiy for the regular coke (Laclede-Elkhom) without creating injurious irregularity in operation. 2. The coke carried the normal basic burden well, with every indication that it would make any grade of foundry iron with the normal burden. 3. It was indicated that with the Illinois coke the co&sumption of coke per- ton of pig iron is normal, and possibly even less than normal. This fact can not, however, be absolutely established in So short a test. Great caution should be used in applying the results of this test and in predicting what Illinois coke might do if used on an older blast fvunace of larger capacity or of different construction. The coal from which the coke used in this test was made represented Missing Page EI^ECTlilC BATTERIES 97 basis of the results obtained on this trip, the Bureau recommended the use of this make of cell in Ba-2 batteries, provided that they were shipped unassembled to France to be assembled into batteries when desired. The chemical work on storage batteries for the War Department was devoted principally to the study of methods of tests and com- parison of specifications for sulphuric add used in storage batteries. The need for this information was emphasized when the Signal Corps requested such data for making specifications to be used by a large number of manufacturers and users of batteries, and it was found that the specified amoimts of impurities varied as much as 1000 per cent. Some of the methods in common use for testing battery add were foimd to be unreliable, this bdng particularly true in the methods for determining the amounts of nitrates and nitrites. Much labor has been spent upon the comparison of methods of testing preliminary to the main problem of determining the limit- ing amounts of impurities permissible, and eventually of drawing up reasonable spedfications for battery add. In connection with the comparative testing of truck storage batteries for the construction division of the Army, the electro- l)rtes of the batteries were analyzed. The results of analyses showed that the batteries giving the best electrical performance contained the purest electrolyte, while the batteries giving the poorest electrical performance contained the most impure add. These chemical tests were, therefore, confirmatory of the quality of the different brands of batteries tested. In addition to the work upon battery add some analyses were made upon spedal storage-battery plates and fillers for dry stor- age batteries. At the request of the Signal Corps analyses were made upon a large number of alkaline electrolytes taken from storage bat- teries supplied by the Edison Co. to determine whether they con- tained the spedfied amounts of potassium and lithium hydroxides. Standardization and Preparation of Specifications for Batteries Before the United States entered the war, questions relating to dry cells were brought to the attention of this Bureau by several of the military departments. The first problem presented was to find a suitable cell of American manufacture which could be used to replace cells purchased by the British Navy from manufacturers in Scandinavian comitries. This and other problems emphasized 26035°— 21 T 98 WAR WORK OP THE BUREAU OP STANDARDS the need for the Btireau undertaking a systematic study of battery problems. Information has been secured on the performance and specifications of dry cells from pubHcations on the subject, by consulation with the mantifacturers wherever possible and by experiments in the laboratory. The collection of information with regard to storage batteries has been undertaken in a similar way. The information which the Bureau has gathered has been made available to military departments (i) in the form of special reports when these have been asked for; (2) in reports of tests on batteries which have been submitted; (3) in a circular of informa- tion about dry cells; (4) by cooperation with the Standardization Committee of the War Department. The Bureau has prepared and published a circular of information on dry cells (Circtdar No. 79) which contains a brief historical re- view of the development of the dry cell and the theory of the reactions which take place in it. The various kinds of construc- tion are described in detail, since this has an important bearing on the utilization of the cells. The two most common types are the so-called " paper-lined " cell and the "bag-type " cell. The paper- lined cell is the one most commonly made in this country, and it is well adapted to some piuposes, besides being easily and cheaply manufactured. It does not possess, however, as desirable charac- teristics as the bag-type cell for many of the uses which resulted from the war. This is because the cells deteriorate when standing on open circuit. If long periods of time must elapse between the manufacture of the cell and its use in a distant field of operation, the deterioration which takes place greatly reduces the service which the cell can render. The bag-type cells of the larger sizes are less subject to this deterioration and are more commonly used in Europe than in this coxmtry. However, the small flashlight cells which are made in the United States in very large nimibers are nearly all of the bag-t5T)e construction, this being necessary since the deterioration of these small cells is more rapid than in the larger sizes. The circular describes sizes and kinds of cells which are avail- able on the American market and the uses to which they are best adapted. The electrical characteristics are discussed with a view to showing how the best service may be obtained from the cell, an important matter, since it often happens that dry cells are used in such a way that only a small part of the service which they might perform is obtained. ELECTRIC BATTERIES 99 The tests which are commonly made on dry cells have grown up with the industry; many are not considered to be entirely satis- factory, and some are not easily carried out. Nevertheless, these tests are discussed in detail, since in the absence of simpler and more reliable tests, they are the only ones that can give relative information as to the performance of the various brands now on the market. Specifications covering all of the various sizes and kinds of dry cells in common use have been prepared by the Bureau and in- corporated in the circular as an appendix. These specifications were prepared as a possible guide for the various Government departments in the piurchase of dry cells, and they have been formally adopted by the War Department. Before the circular on dry cells was published, the War Depart- ment undertook the standardization of electrical supplies in gen- eral, and a subcommittee on batteries, consisting of officers repre- senting a number of different bvureaus of the War Department and two representatives of the Bureau of Standards, was appointed. When this committee began its work, the question of specifications for dry cells was immediately taken up. The specifications which the Bureau had already prepared served as a basis for the discus- sion which took place in the committee, with the War Industries Board, and with the War Service Committee of the Manufacturers. Only minor changes in these specifications were made, and an agreement was reached with all the parties concerned. The Stan- dardization Conmiittee then adopted these specifications. Two important results of these specifications were, first, the limiting of the number of sizes which are considered as standard, and, second, the specification of the minimum service required of the different sizes. In limiting the ntunber of sizes, the Bureau had the cordial cooperation of the manufacturers who had been more or less compelled in the past, for trade reasons, to manufacture a large variety of shapes and sizes for which there was compara- tively little use. The nvunber of the large-size cells was cut from ID to 3, and of the small or flashlight cells from an indefinite num- ber, about 40, to 6. (See Fig. 10.) The Biureau also had the cordial cooperation of the War Industries Board in thus limiting the number of standard sizes, since part of their conservation work was directed toward the ehmination of useless sizes. It is probable that the Bureau's recommendation for limiting the number of sizes would have been made mandatory by the War Industries Board lOO WAR WORK OF THE BUREAU OE STANDARDS but for the signing of the armistice. It is anticipated, however, that the Bureau's recognition of these as standard sizes will lead the manufacturers to eliminate gradually those which are less desirable. This subcommittee on batteries also considered the question of storage batteries. The Bureau cooperated by gathering informa- tion from various sources about the characteristics and perform- ance of different types of storage batteries. The Bureau also carried out extensive tests on these batteries as described elsewhere. Specifications covering storage batteries were prepared jointly by representatives of the construction division of the War Depart- ment and by the Bureau of Standards in consultation with the engineering committee of the Manufacturers' Association. These specifications were formally adopted by the standardization committee. Reports on Special Subjects The Bureau has been called on in a number of instances for information dealing with battery problems by different bureaus of the Government departments. Only three of the more im- portant of these will be mentioned here. A controversy arose between the Food Administration and a large manufacturer of batteries concerning the use of wheat flour in the manufacture of dry cells. This flour is used for making the electrolytic paste which separates the zinc container from the depolarizing mixture in the cloth bag. A few manufacturers in this country have been using cornstarch instead of the wheat flour, but this process of manufacture of the cell is so entirely different from the more common or wheat flotir process that a manufac- turer using one process could not change to the other without alteration of the methods and machines employed. For this reason the proposal of the Food Administration to prohibit the use of wheat flour in the manufactiu-e of dry cells was a serious one to the industry, and particularly so in the case of the above- mentioned company, whose product was largely for the War Department. At the suggestion of this company the Food Ad- ministration referred the matter to the Bureau of Standards, and conferences were held with both the manufacturer and the Food Administration. A satisfactory basis of agreement was found, since it is possible to manufacttu-e dry cells with wheat flour which is unfit for human consumption. Accordingly, the Bureau recom- mended to the Food Administration that the battery maker be ELECTRIC BATTERIES lOI permitted to use such flour, and apparently this recommendation was entirely satisfactory to both parties. A request was received from the Bureau of Ordnance of the Navy Department for information on silver chloride cells with a view to their use in submarine mines. In addition to reporting on the structure and electrical characteristics of these cells, the Bureau made tests on sample batteries which were submitted by the Navy Department. In connection with the testing of storage batteries for indus- trial trucks and tractors, it was requested that additional infor- mation be furnished on the relative characteristics of add and alkaline storage batteries as applied to tractor work. The Bureau gathered information on this subject from various reliable sources, and the report was made as a lecture before the truck and tractor committee of the War Department and other officers whom they invited to be present. Military Tests of Batteries About 50 tests on batteries of various kinds have been made for the military departments, many of which have involved an extended series of measurements. The purpose of some of these tests has been to determine the fitness of various kinds and makes of batteries for particular purposes, while in other cases the object was to determine the quality of material submitted by manufac- turers. In still other cases the work has taken the form of experimental development of new t3T)es and kinds of batteries to meet imusual service conditions. A brief description of some of the more interesting tests is given below. A test of over six himdredis-cell, batteries of the type Ba-2 has been made for the Signal Corps. These batteries were intended for use in the amplifiers of radio sets such as are used in airplane service, but had not given satisfactory service, due to deterioration when standing idle, as in storage or in transit. This had been so large that many of the batteries had been found unfit for use when wanted. The Btureau was asked to make a study and test of these batteries to find what could be expected of them and what improve- ments could be made. Nearly 700 of these batteries, representing the product of several manufacturers, have been sent to the Bureau for test. In addition to testing these batteries at the Bureau, measurements were also made on much greater numbers in the military storehouses during June and October of 19 18, the restdts of the latter measurements showing that very large I02 WAR WORK OP THE BUREAU OF STANDARDS percentages of the batteries in storage were unfit for service. The deterioration of dry cells is more rapid in the smaller than in the larger sizes, and for this reason it was suggested to the Bureau of Steam Engineering of the Navy Department that similar batteries for Navy use should be made from larger cells. This recommenda- tion was adopted and shortly afterward the Navy bought batteries containing cells of the recommended size, which they designated as CBG-3535. The Signal Corps, however, preferred not to use batteries of a larger size, since they would not fit in the compart- ments of their apparatus. The Bureau suggested that batteries might be made in such form that water could be added to the individual cells at the time they were required for use, thus making them similar to the so-called "reserve" or "desiccated" cells which are common in the larger sizes. It was also suggested that these batteries might be assembled in France, since in this way those for use abroad could be obtained without the delay which otherwise occurred between the time of manufacture and their utilization. This work was undertaken by the Signal Corps in the latter part of the year. Tests which the Bureau made showed that, of the batteries manufactured in this country, some were distinctly superior to others, and it was therefore suggested that a large number of indi- vidual cells of the best makes, complete but not assembled into batteries, should be shipped abroad so that each cell might be tested individually at the time they were to be assembled. In this way those which were defective might be discarded. Most of the batteries manufactured in this coimtry were assembled without any such tests being made on the individual cells, and the Bureau considers that this is partly the reason for the large number of failtires which were experienced. As the failure of any one cell is sufficient to cause the failtire of the entire battery, it is apparent that the chances of the battery failing are 15 times as great as that of the failure of any one cell. In a number of cases the Bureau was able to show conclusively that the failure of the battery was of this type. One result of the Bureau's experiments was to show that a very simple empirical test of these batteries could be made in the miUtary storehouses or in the field, which would give reliable information as to whether they were fit for service. This test consisted in measuring the voltage of the batteries with an ordinary voltmeter and discarding any batteries showing less than 20 volts. This test depends on the fact that the ordinary voltmeter of 100 ELECTRIC BATTERIES 103 ohms per volt of the scale draws enough current from the battery to indicate whether the latter is in fit condition, in addition to indicating what may be termed its semiopen circuit voltage. The batteries tested at the Bureau were discharged under many different conditions. In one of these experiments it was found that when they were discharged at temperatvu-es below 0° C or 32° F the service rendered was very small even though the batteries were in good condition. This is a matter of importance, since the service in airplanes, for which they were intended, may expose them to very low temperatures. In addition to the electrical tests a careful chemical examination of the contents of some of the batteries was made. The results indicate clearly that those giving the most satisfactory service were the most uniform in their chemical composition and con- tained sufficient manganese dioxide and other ingredients. Batteries similar to the above, but designated by the Signal Corps as type Ba-5, have also been tested in large numbers. These batteries are of the so-called "reserve " type; that is to say, the battery is made up dry and is made active by adding water at the time it is required for service. Nmnerous mechanical difficvdties developed in connection with these batteries owing to the difficulty of providing sufficiently large vents for adding water. Since these batteries were assembled dry, there was no means of testing them until such time as the water was added when they were required for use. For this reason, large numbers failed almost immediately after being filled with water because of the leakage of electrolyte from the cells into the battery, causing short circuits. It was fotmd that batteries which did not show this initial defect gave reasonably good service. The results of the Bureau's tests on these batteries indicated that the percentage of failures due to leakage was rather too high to consider them satisfactory. Another type, of battery intended for the same purpose, and designated by the Signal Corps as the type Bb-12, is made up of 12 small storage cells. They were made as an experiment by two different manufacturers and were of the so-called "dry" or non- spillable type, in one case the battery electrolyte being thickened with gelatin. The service which these batteries gave was un- satisfactory, and it also developed that when they were placed in a closed compartment containing other apparatus the metal parts of the latter became corroded. The Bureau's experiments indicated that these batteries should not be used. 104 WAR WORK OF THE BUREAU OF STANDARDS The Bureau undertook, at the request of the Signal Corps, a study of the galvanic pile to ascertain its suitability for furnishing the small currents required of the batteries mentioned above. In the course of this investigation a form of disk battery was developed which is rugged and can be made active by the addition of water as in the case of the reserve cells. They are now being made at the Bureau and assembled for further tests. These batteries have certain advantages over those of the Ba-2 and Ba-5 type, although the actual service which they will render is slightly less than the others when the latter are in the best condition. The units of this disk battery are usually small dry cells, but the Bureau has also made experiments on a similar construction, in which the elements were to be those of a storage battery. In the case of the disks assembled as a battery the cells are assembled in a glass tube, so that the battery can be watched during the time it is in use. Such an arrangement reduces the maximum difference of potential between adjacent cells to i X volts, a distinct advantage in diminishing lieakage trouble. Experiments on this form of battery are being continued. An tmusual test of dry cells intended for use with a trench signal projector was made at the request of the Biu-eau of Aircraft Production. Its object was to ascertain how long the batteries would operate the signal lamp satisfactorily. In imitation of the service anticipated, the lamp was caused to flash twice in each second with a duration of one-fourth second for each flash, and measturements were made on the batteries until the light given by the lamp was no longer serviceable. On this test various t)T)es and sizes of dry cells were used, some of them being ordinary cells, while others were of the bag-type construction. Competitive tests of various brands of flash-iight batteries were made for the Depot Quartermaster, the Field Medical Supply Depot, the General Engineer Depot, and the Ordnance Depart- ment, to show the relative merits of the different brands submitted, as a basis for large purchases. The experiments showed conclu- sively that in the case of a few brands, on which an unusually low price had been quoted, the cells were of very inferior quality. After these tests were completed and contracts awarded, the General Engineer Depot asked the Bureau to test samples of the batteries furnished by the manufacturer, to determine whether they conformed with the samples which he had submitted. A large number of batteries of various kinds were submitted for ELECTRIC BATTERIES 105 test by the Signal Corps, in addition to those which have been previously mentioned. At the request of the Bureau of Ordnance a long-diu-ation test was conducted on the shelf Hfe of special batteries designed for use in submarine mines. Part of this test consisted in keeping the batteries tmder temperature conditions comparable to those of sea water with the idea that this Bureau should notify the Navy Department when they might expect failures of the batteries in planted mines. They also requested information and tests on silver chloride batteries, and a number of experiments were made. This cell differs materially from the ordinary dry cell, as the de- polarizing agent is chloride of silver instead of manganese dioxide, and the cells are less subject to deterioration on open circuit than the ordinary dry cells. An examination was made for the Navy Department of six small dry cells which had been taken from a German submarine mine. The extensive use which has been made of the dry battery in the present war has been emphasized by the number of different military organizations which have requested tests by this Bureau and by the number of diflFerent kinds of batteries which they have submitted for test. Altogether, more than 1 600 dry cells or bat- teries have been delivered to the Bureau in connection with purely military tests. In many cases these tests have been of a special character because of the use for which the batteries were intended. The results of the Bureau's experiments and tests have been not only to benefit those requesting the information, but in several cases have led to important changes on the part of manufacturers. This is particularly true in one case where a manufacturer has carried out a partial reorganization as a result of the Bureau's work in showing the poor quality of the cells which were being supplied. The Bureau has carried out a test of large storage batteries such as are used for the propulsion of electric trucks and tractors. These batteries were representative of seven different manu- factiurers, and they were tested for capacity, efficiency, and ability to maintain their voltage over widely varying conditions. I/ike- wise, some measurements were made at low temperatures, and a number of experiments were carried out to determine the velocity and acceleration of an electric tractor when equipped with different types of batteries. The request for this test came from the chairman of the Committee on Standardization of Electrical Equipment and Supplies. I06 WAR WORK OF THE BUREAU OP STANDARDS The results showed that some of these batteries gave a very satisfactory performance in the laboratory. Others, however, failed to equal their rated capacities, and these for the most part gave a relatively poor performance on the other tests to which they were subjected. The batteries may be divided into two main classes: (i) The lead-add type of storage batteries, including the ordinary flat-plate type of 15 or 17 plates per cell and some similar styles with thinner plates, and the so-called "ironclad" or pencil- plate batteries; and (2) the nickel-iron storage batteries, contain- ing alkaline electrolyte. Measurements of the ampere-hour capacity of these batteries for continuous and intermittent discharge were made, and their efficiencies were determined at the normal rate of 45 amperes and at 90 amperes discharge. There was no great difference between the various makes of lead-acid batteries, but both the ampere- hour efficiency and the watt-hour efficiency of the alkaline bat- teries were distinctly lower than in those of the lead type. The watt-hour efficiency of the different types of batteries has an im- portant influence on the cost of charging them. The averagevolt- age of the different batteries at the normal rate of discharge was approximately the same for all the different types tested, the differences being less than 2 per cent, but at the highest rate of discharge — that is, 350 amperes — the average voltage of the alkaline batteries was found to be only 60 per cent of the average for the lead batteries. Measiurements on these different types of batteries at low temperatures indicated that the lead-acid batteries gradually decrease in capacity as the temperature of the cell is lowered, but a different phenomenon was observed in the case of the alkaline batteries. It was found that for these a certain critical temperature, dependent on the rate of discharge, exists, below which the battery can give but a very small output. The experiments made with the electric tractor equipped with the lead-acid batteries and the alkaline batteries consisted in making exact measurements of the velocity and acceleration of the tractor operating under different load conditions. For this purpose a series of 39 electrical contacts was prepared, spaced at definite intervals along the path of the truck and arranged to be broken as the truck passed along. The time at which each contact was broken was recorded on a smoked paper record which also bore a time scale. This latter apparatus was developed by the section of the Bureau devoted to work on sound ranging. From the records obtained the time at which each contact was broken ELECTRIC BLASTING APPARATUS IO7 was determined to within o.oi second, and from the time observed and the spacing of the contacts it was possible to calculate the time-distance diagram from which the curves of velocity and accel- eration were plotted. The results of this test showed that when the truck was running without load or with only a Ught load, its speed after the attaining of a unifrom velocity was greater when the truck was equipped with the alkaline batteries, but when the load was increased beyond 2 tons a greater velocity was obtained with the lead-acid batteries. This observation confirms the labora- tory tests which showed that, at the higher rates of discharge, the alkaline batteries could not maintain as high an average voltage as the lead type, due to their higher internal resistance. It was also fotmd ia testing these alkaline batteries that when the tractor was heavily loaded and nmning up a grade of 3 per cent, the accel- eration of the tractor at the end of the run was negative; that is to say, the tractor had attained a maximum velocity which the alkaline batteries were imable to maintain. A number of other tests of storage batteries have been made, but these will not be described in detail. The Signal Corps sub- mitted to the Bureau some miHtary batteries of French manufac- ture, which were constructed with celluloid containers. A con- siderable fire hazard is encountered when celluloid is used, as a bad contact between two cells may ignite the celluloid. In one case examined the complete destruction of the battery was pre- vented by an outer container also of celluloid which inclosed the individual cells. This did not bum through before the oxygen within was exhausted and the fire extinguished. For testing portable storage batteries submitted by the Signal Corps a rack was constructed and installed at the Bureau. The Bureau provided facilities and cooperated in the supervision of this test, but the personnel for carrying it out was provided entirely by the Signal Corps. An interesting test of small storage batteries was made for the Chemical Warfare Service. These were intended for use with indi- vidual smoke precipitators to prevent the clogging of gas masks. ELECTRIC BLASTING APPARATUS About June i, 191 8, the Bureau of Standards was requested to undertake an investigation of equipment used by the American Expeditionary Forces in the electrical ignition of explosives. Two branches of the Army were concerned with this apparatus — the Ordnance Department, the officers of which were developing I08 WAR WORK Olf THE BUREAU OP STANDARDS equipment for firing Livens gas projectors, and the Engineer Corps, which was charged with the problems of demolition. The Livens gas projectors are designed to be hurled into the enemy's territory from points near the front-line trenches. Short shafts are sunk in the ground and the projector placed therein above a charge of powder. A small primer (or squib), made of a capsule containing fulminate of mercury and guncotton surround^ ing a fine metal wire, is placed in the mass of powder. This is so connected to external wires that it can be fused by sending through, it an electric current of the proper intensity. This fusion ignites the material of the primer, which in turn fiires the entire charge under the projector. A similar primer was used by the Engineers to fixe charges of dynamite in demolishing bridges and other structures of value to the enemy. Electric Generators The apparatus used to generate electric current for this purpose should have at least three characteristics, especially when used for military service. Eirst, it should be reliable; second, it should be portable; and, third,, it should have ability to supply a suflScient amount of energy to insure firing the primers connected to it. It is important that the operator shall be absolutely cer- tain that when his generator and circuits are properly arranged he can depend upon them to fire, for most of this work involves life hazard for the men detailed to it. Closely alUed with this is the abihty to supply sufficient energy. The complete demolition of a bridge often requires from 30 to 60 separate charges of dynamite, each sometimes containing two primers. The electric generator should then be capable of firing all of these simultaneously, which means that it should generate sufficient voltage to force through all the primers in series a current of the proper intensity. The necessity for compactness scarcely needs any comment. The conditions imder which these generators are used prohibit an elaborate arrangement of permanent apparatus. The source of power should be capable of being carried about by one man with ease. At the beginning of the war there were available in this country several small generators which partly met the above requirements. They were being built for use in mines and quarries, and were called by the trade name "electric blasting machines." A typical machine of this sort, with the number of primers which it would Miscellaneous Publications, Bureau of Standards, No. 45 Fig. II. — The old type of blasting machine used for firing primers when the United States entered the war The total number of primers which this machine is capable ol firing are piled at its base Fig. 12. — The improved blasting machine developed for the use of our overseas forces by the Bureau It is interesting to compare the number of primers which this machine will fire (shown at its base) with those shown at the base of the old machine in Fig. ii. The machine likewise proved to be much more reliable, and a great deal easier to manufacture Miscellaneous Publications, BLireau of Standards, No. 46 Fig. 13. — Set of gage blocks constructed at the Bureau Prior to the war all such blocks were marie iti Europe. As these are used as reference standards iu munitions plants, lar^e numbers were required during the war. The Bureau aided hy pru- during nian>' sets iu its own shojjs Fk;. 14. ^Elaine Uuiipsfor daylight lra)ismissio)i of inessaijes The type nt' laniii ui.ed, ausiiiK nasliiiu,' sik'u.ils, and utlier phases of llie problem were studied ELECTRIC BLASTING APPARATUS IO9 ordinarily fire, is shown in Fig. 11. They were small, portable, series-wound, two-pole, direct-current generators, in appearance somewhat like the early Edison dynamos. The armatures were of the Siemens type, and both armature and field were of soUd iron. A long rack with a handle is meshed with a pinion on the shaft of the generator. In operation the external circuit of primers was connected in parallel with a switch, short-circuiting the generator. As the rack was pressed down, current flowed through the gener- ator and short-circuiting switch. At the end of the stroke, this switch was opened by the rack, thereby making available a com- paratively high voltage to force current through the circuit of primers. These machines were portable and were probably reliable enough for the purposes for which they had been used. However, soon after the American Army began using them, it became clear that their capacity was very limited, and that even with only a few primers connected to their circuits, they often failed to function. Upon examination of the constructional details, some members of the Bureau's staff recognized that they were precisely similar to generators designed for the same purpose more than 25 years ago. In view of the great development in the electrical field dvuing that time, the Bureau was asked to undertake an investigation to determine whether a generator could be designed of approximately the same size and weight, but of greater relia- bility and capacity. It was apparent that with solid-iron construction the output of these generators was much below that of machines of similar size constructed along modem lines. Accordingly, the experimental work was begun by seeming several standard direct-current motors, such as are used for driving electric fans, and arranging them to be driven as generators. After testing the capacities of several such machines and making a number of changes in the electrical design (such as size of wire and number of tiims on armatture and field), it appeared that a machine could be built with considerably greater capacity and with practically no increase in weight. Moreover, iron castings and stampings, which were being used in large quantities in electrical factories in the manufacture of small motors were suitable for this purpose. As the Army had immediate need for these generators, this was a decided advantage. Some trouble developed in these experimental machines in the form of armature breakdowns. The new generators, wound with no WAR WORK OP THE BUREAU OP STANDARDS fine wire, generated voltages which were too high for the insulation of the wire. Using heavier insulation, however, at the expense of a smaller number of turns, solved this difficulty without seriously detracting from the capacity. These results were communicated to the General Engineer Depot about one month after the investigation began. Mean- while, tirgent messages had been received from the commander of the American Expeditionary Forces in France, telling of numer- ous failures of the blasting machines and setting forth the neces- sity for a more reliable source of current. With the results of these tests, the General Engineer Depot took up with two of the large manuf actxurers of electrical apparatus the question of adapt- ing stock material to the building of a new generator. By Octo- ber I, 1918, the details had all been worked out, contracts had been let (with these two companies) for 10 000 generators, and the machines were already in production. Since the imreliable blasting machines in use constituted a real Ufe hazard, arrange- ments were made by which large orders for them already placed were canceled. As a result of this work there was made available for the Army a portable generator which was capable of suppl3ruig sufficient energy to fire 100 blasting primers connected in series without danger of failure. Its weight was no greater than that of the early blasting machine, which was rated at 20 primers and accom- plished this only occasionally. The improved machine developed by the Bureau is shown in Fig. 12. The -large number of primers which it will fire should be contrasted with Fig. 1 1 . Incidentally, the cost of the new machine was only about 90 per cent of that of the earlier one, due to the fact that its construc- tion was similar to that of other machines ah-eady being manu- factured in large quantities, so that while far greater reliability and capacity were secured, several himdred thousands of dollars were saved to the Army in the machines then contracted for. Primers In order to determine the electrical output required of the blast- ing generators, a study was made of the current required to fire the primers and of the time between the application of the current and the explosion of the primer. This latter information is of importance in cases where a number of caps are connected in series, as it is essential that the slowest cap be heated to the ignition point before the explosion of the fastest cap interrupts the circuit. EI , ,, Among the activities of importan,ce may be mentioned partic- ularly that of the; chief of the metalliurgipal division as the rep- resentative of the Department of Commerce on the reqtiirements division of the War Ind^stries Bojard. Thi?, enabled the Depart- ment to take an active interest in and to seciure valuable assistance on many exceedingly important matters coming before the Govern- ment.. Four months were spent abroad by this member of the committee in the spring and summer of 191 7, while serving on the scientific mission sent to obtain information concerning applica- tions of science to warfare and the part to be played by scientific men in the war. The information thus gained proved of inesti- mable value in planning investigations in the United States. Another member of the staff was very helpful as adviser in ceramics to the War Industries Board and assisted in framing specifications for enameled ware as a member of the Army com- mittee on standardization. Another rendei;ed valuable service as a member of the Interdepartmental Committee on Minerals and Their Derivatives, and also aided in the manganese-conservation program. The committee on light alloys of the National Advisory Committee for Aeronautics parried out njost of its experimental work through the Bureau of Standards, and the results of this have appeared in a, series of pubUcatipns., fhe National Research PHYSICAL TESTS OF METALS AND METAL STRUCTURES 1 73 Council has actively cooperated with the Bureau in several military problems, and the Bureau's representatives have been active on 1 1 of its committees, among these one of the most important being a committee to make a survey and recommend practice as to the ingot and finishing practices of American steel mills. This work originated in the Ordnance Department of the Army in connection with the works practice as influencing the output of steel. Various members of the metallurgical division were called in as technical advisers by the several branches of the War Industries Board, particularly on questions relating to tin, steel, and platinum, and the Bureau of Standards in consequence has undertaken con- siderable experimental investigation, especially on tin conservation and substitutes. In addition to formal committees with regular programs there have been held a great many conferences at which the Bureau of Standards' representatives have been present, called by the various miUtary departments to discuss with manufacturers technical questions bearing on military materials and specifications. PHYSICAL TESTS OF METALS AND METAL STRUCTURES Load Tests of 150-Ton Floating Crane for the Navy At the request of the Bureau of Yards and Docks of the Navy Department, engineers of the Bureau of Standards conducted a loading test and strain-gage analysis of the 1 50-ton (336 000 pound), revolving floating crane built for use in the navy yard at Norfolk, Va. (Fig. 19.) The test was made to determine the stress distri- bution in the various meinbers of the structxu-e while imder load and to obtain a more complete knowledge of the actual amotmt of the stresses in some of the members of such a staticaUy indeterminate structure. This Norfolk crane represents the most advanced type of re- volving floating crane of large capacity. The jib is a tapering Pratt truss, and the balance of the superstructure consists of two sub- divided triangular trusses rigidly connected by cross-bracing. The entire revolving load is carried on a pintle which transmits the load to a thrust bearing on the deck of the pontoon. The pintle is sup- ported laterally by a hexagonal tower which is constructed in- tegrally with the framework of the pontoon. The method of stress analysis used in this investigation parallels that used in the load tests of the Arlington building (described in the article on cement and concrete) and is the same method which has been so successfully used in the study of bridges, steel frame- 174 WAR WORK OP THE BUREAU OF STANDARDS work of buildings, and similar engineering structures. A Berry- strain gage was employed to determine changes of length of estab- lished gage lines by comparison with standard reference bars. With known moduli of elasticity for the materials in question, the unit deformations may be converted to the corresponding stresses^. In the hands of trained observers such an investigation yields results of great value for fMure designing work. Readings were taken at various points of the deck to determine the magnitude, and, if possible, the position of maximum stress in the upper deck plate; on the tower members to determine whether the tower acted as a unit; around the manhole in the tower legs to find the effect of such an opening; on the pintle to determine the magnitude of the stresses in these members; and on various members of the superstructure to determine the manner in which the loads were carried down into the pintle. The strain gage measurements on this structiu-e showed that all stresses in the various members were within safe limits. Such knowledge is an important factor in the design of structulres such as this, which are statically indetierminate. This detarminatipn was of especial importance at the time of conducting the investiga- tion, inasmuch as another large floating crane of somewhat like construction had only recently collapsed under load while in use on the Panama Canal. Strength and Efficiency of Electric Welding The tu-geht need for ships to transport our men and materials to Europe during the war made it necessary to adopt every means of seciuring maximum output from the shipyards of this cpuntry. The " fabricated " ship, constructed from plate material on which all cutting and drilling had been completed at remote mills before shipment to the yard, saved time at the shipyard under the conditions existing diu^ing the war. A radical suggestion which was never used to the extent which its merits appeared to warrant was the electric welding of the frames and plates of the hull instead of riveting. Autogenous, or fusion, welding has been in use for many years. Most of thi? work has been done by the use of the gas torch to which oxygen and acetylene are supplied. The weld is formed by fusing together the adjoining edges of the pieces of metal. Material is sometimes added at the same time from a metal rod. Due to the prohibitive cost of the gases required for oxyacetylene welding and to nonacceptance of welding for certain work by the PHYSICAL TESTS OP METALS AND METAL STRUCTURES 1 75 insurance companies, the use of welding ship hulls was greatly restricted. The possibility of so developing electric welding as to permit its being accepted to replace riveting was carefully considered. In the electric-arc welding process the fusion of the metal is effected by an electric arc formed between a metal rod (or elec- trodie) and the pieces to be joined. Metal from the rod is de- posited in the weld. The welding committee of the Emergency Fleet Corporation was formed to study the matter carefully and to conduct any needed investigations. This Bureau had representatives on the committee and performed much of the laboratory work for deter- mining the properties of the welds, particularly the strength. In order to obtain definite information upon welds which were being made commercially, a large number of pieces of one-haJf inch steel jplate Were welded at different shops using all available t3^es of apparatus. These were known as the "Wirt- Jones" tests and attracted a great deal of attention from engineers and shipbuilders. The welding data included for each test weld, besides the identification of the operator, the position in which the weld was made, the type of weld, the rate of welding, the type of electrode used (with its diameter and manufacturer), the arc and open- circuit voltage, and the amperage. These data were obtained ia the shops where the welds were made. The specimens were machined there and elsewhere. The physical tests were all made at the Bureau of Standards and consisted of tensile, torsional, fatigue, and cold-bending tests. It was found that all of these except the fatigue tests were consistent. A specimen showing high strength in the tensile test showed high strength in the torsional also. The specimens showing relatively high ductility in the tensile tests showed equiv- alent ductility in the cold-bend tests. The fatigue tests were made in three Upton-I^ewis machines and by three different engineers. Unfortunately, the results do not appear as consistent as could be wished. Many fatigue exper- iments should be conducted if reUable data are to be obtained upon this important property of welds. It is believed by many engineers that the failure of welds in service is usually by fatigue and that this property of welds is relatively more important than the fatigue resistance of the material in which the weld is made. .176 WAR WORK OP THE BUREAU OP STANDARDS The total number of specimens in this series was over two hundred, and the testing work required most of the time of two engineers for nearly a week; ; 1 1 - In addition to the Wirt- Jones tests, this Bureau was at one time making progress reports to the welding committee on 14 investigations planned to give information upon some of the many problems constantly arising in connection with welding work. The niunber of specimens involved in these investigations varied from 5 to 42. Due to the satisfactory results obtained from these tests, it was decided that the electric-arc welding process was suitable for ship construction, and the committee formally recommended tp the Emergency Fleet Corporation that a merchant ship be de- signed to utilize welding wherever it could be employed to advan- tage and that the ship be completed as soon as possable.. The signing of the armistice, took place befoi;e much progress was made, but some commercial shipyards are now energetically preparing to construct ships in which nearly all the frames and plates for the hull are to be welded into place. It is estimiated that the cost of welding will approximately equal the cost of riveting plates and frames after they have been "laid off" and the rivet holes punched. For welding, however, all this preliminary work is unnecessary, and, the plates and angles do not need to be as accurately cut to dimensions. The saving if welding is used is therefore very large, and it is chiefly: a saying in labor cost, which under present conditions constitutes a large portion of the total cost. . One result of these extensive investigations was the important conclusion drawn from the results of the Wirt- Jones tests by the chairman of the researcii subcommittee of the welding com- mittee, that the best physical properties were obtained from those specimens welded with the highest current density in the eleG1trod#- Apparently, imder this conditioiij thorough fusion occurred, in the Weld. The importance of this discovery, which is generally accepted at present, is very great. The fact that little properly conducted investigational work has been done upon welding processes, and that there are many possible lines for improvement, leads to the conclusion that such work offers great possibilities if properly carried out. The American Welding Society,: an outgrowth of the welding committee, has been formed to encourage welding research work PHYSICAL TESTS QP METALS AND METAL STRUCTURES ^^^ by qoqrdinating the efforts of industrial, educational, and govern- mental laboratories. Thris: Piu-eau is assisting the work of this society in so far as possible. , , In an extraision: of the exp^iujiental work to determine the fatigue resistance of welded joints such as would be actually uspd in ship construction large specimens were made from plates one- half inch thick, 8 inches wide, and 40 inches long. A fatigue machine driven by a .15-horsepower electric motor was designed and built to subject these specimens to repeated bending stresses, wit,h provision not only for recording the number of cycles of stress required to cause failure, but also for making an autographic record of the stress produced in the welded mate- rial for each application. Investigation of Strength di Chains The Bureau was called on to make two extensive series of tests of chains in connection with its war activities. The eariiei" series comprised numerous samples of links for anchor chains and was carried out in cooperation with the chain committee of the Emer- gency Fleet Corporation. The research was intended especially to determine the value 6f electric welding in chain manuf actiiring and the relative efficiencies of cast and fbrged links. The anchor-chain links tested included wrought-iron, machine- steel, rolled-shafting steel, and electric-cast steel specimens. Some were cast integral, others were drop-forged, and still others were welded. Certain of the links were prepared with studs and others without. The tests were not extensive enough to warrant general conclusions on the relative values of these different' types of links, but they sufficed to furnish considerable comparative data. ' ' : ' Six electric-welded basic open-hearth steel links of 2}^-mch. bars showed an average strength of 350000 pounds. Three links, each drop-forged from 2-inch bars of steels of various manu* facture, averaged as follows: Rolled shafting steel, 212600 pounds; machine steel, 345 000 pounds; and electric steel, 380 OOP pounds. Electric-welded links of 2 X-inchistock aver- aged 260 000 pounds strength for chain iron and 280 000 poimds (one test only) for steel. Fracture occtuxed in ^^^ cases a,t tiie welds. An interesting comparison is afforded by the study of tiie results of tests of six links of cast steel, all of 2-inch diameter stock. Three of these links which had been heat-treated after 26035°— 21 ^12 178 WAR WORK OF THB BUREAU OP STANDARDS casting failed at loads of 326 000, 400 000, and 365000 pounds', respectively. A microscopic study of the fractures showed the steel to possess a very good structure. IV0 cast electric-steiel links averaged 440000 pounds strength, and an accompanying link made from basic open-hearth steel failed at 370000 pounds. This last open-hearth had also been heat-treated after casting, and showed six tiines as great elongation as did the two electric- steel links. AH anchor-<:hain links tested were about 12 inches long over all. This series of static laboratory tests demonstrated the relative superiority of cast-steel links. Dytiatnic tests made elsewhere with the cooperation of the Bureau also served to establish the relatively greater desirability of a steel Unk than of a wrought- iron link. Service tests by the Navy Department also contributed to the establishing of the efficiency: of cast chain which received the approval of Lloyd's Register of Shipping. In the process of reorganization of the Army, by which the purchase of various supplies was concentrated in suitable Army branches under the general direction of the Piurchase, Storage, and Traffic Division the procurement of chain was assigned to the Engineer Corps. The Bureau was represented, together with other Government departments on the committee on chain. The Biureau made extensive tests of harness, vehicle, towing, sash, anchor, and other types of chain with straight links, twisted links, plain stud, and otiier links, in its cooperation with the War Department standardization committee. It also tested harness, clip-grab, box, and barrel hooks in the course of this investigation. The test data obtained here, combined with that obtained in com- mercial tests, furnished the basis for the formulation of chain specifications. Standard; War Department specifications for chain, together with much other data, including a list of manu- facturers, have been published in the War Department Catalogue No. 5. Wheels, Investigation of, Artillery, Truck, and Aiiplane This subject will be foimd under the above heading in another part of this report. Metal Construction for Airplanes A description of the work carried out on this subject will be found under "Aircraft Construction." NATURAIv-GAS INVESTIGATIONS 1 79 NATURAL-GAS INVESTIGATIONS Among the war activities of the Bureau iliay be mentioned the work on the various problems presented by the natural-gas industry. The increased demands upon all industries brought about through the war and the unusually severe winter of 191 7-18 combined to cause a complete breakdown of this industry in par- ticular localities. Calls for assistance came from several such places and were given the attention of the gas engineering section of the Biweau. Beginning of Natural-Gas Work The first of these calls came in the early winter from Cleveland, Ohio, where a shortage of natural gas was threatening to cause considerable sufifering. The Bureau had in 1916 made a study of the city's testing facilities, and in October, 191 7, was asked for further assistance along that line, and also for help in dealing with the natural-gas problem in that locality. The survey of local conditions made by the Biureau's representative in January, 1918, showed that the situation was being handled as well as could be expected locally, but that the problera demanded more than merely local treatment. Recognition of this fact led the Bureau to consider an extensive investigation of the needs of the industry. On December 31, 1917, Louisville, Ky.,, requested the assist- ance of the Bureau in dealing with its shortage of natxural gas, since the Bureau in 191 3 had assisted in framing the Louisville gas ordinance. As a result of this request a representative of the Bureau was sent to Louisville in the laitter part of January, 19x8, and a brief preliminary report on the situation was rendered on January 28 with the recommendation, similar to that made in the Cleveland report, that the available supply of gas be rationed so as to seciure its most equitable distribution. The need for an ex- tensive study of the natural-gas situation was reemphasized by the Bureau's experience in Louisville. General Natural-Gas Investigatioa Late in February, 1918, the Bvu-eau definitely decided to make the extensive investigation that was so mrgently needed. It was proposed to make a stu-vey of the production, transmission, and consumption of natiu-al gas with a view to eliminating waste in its production and consumption, to securing a fairer distribution between communities of the available supplies, and to securing a more equitable division of the supply between members of the same l80 WAR WORK OP THE BUREAU OP STANDARDS community. The wasteful consumption of natural gas in the manufacture pf carbon black, brick, etc., has shortened the Ufe of the industry by many years, and- the use of natural gas for pther industrial purposes, such as the manufacture of steel, is undoubtedly a case of putting a natural resource to one of its least important uses. The interstate transportation of natural gas gives rise to many complications, not the least of which is the in- stinctive desire of a producing State to keep its gas at home. Thus West Virginia dislikes to see two-thirds of its production trans- ported to Pennsylvania, Ohio, Kentucky, even to Indiana, while it| own citizens experience a shortagfe. In one locality 8 per cent of the consumers use half of the available gas, and thereby cause others who need it to go without. These are some of the problems to the solution of which the Bureau was intended to contribute. Particular attention was to be given to the Louisville situation as presenting in concentrated fofm the various aspects of the natural-gas problem, with the expectation of later applying to other localities the principles developed here. The Bureau was interested principally in the public utility or service aspects of the work, while it asked for and secured the cooperation of the United States Geological Siu^ey and of the Bureau of Mines in working up the scientific and engineering aspects of the problem. One of the Bureau's consulting engineers was placed in charge of the field investigation. ' An elaborate report entitled "Fimdamental Principles of Natural-Gas Production, Service, and Conservation, with Special Reference to the Natural^Gas Situation at Louisville, Ky.," was prepared and presented on August 20. As indicated by the title, this report explained the methods of natural-gais production, transmission, and Ibcal distribution and utilization, with a useful discussion of the vital problem of conservation. A fifth part con- tained mUch useful information on the Louisville situation. The Bureau recognized the considerable merit of this work, but did not think it advisable to publish it without certain modifications of the treatment. It later appeared as No. 7 of the " Mineral Indus- tries of the United States " series of the National Musetmi, and no doubt is finding a wide circulation among those interested in the naturalTgas industry and in the conservation of oiu: natural resources. Relation to the United States Fuel Administration However, the Louisville case still remained imsettled and de- manded attention. In September, 1918, a representative of the t^ATURAI,-GAS INVESTIGATIONS l8l Bureau went to Kentucky and, with the assistance of an esqpert, made a field study of certain gas territory in eastern Kentucky which might be used to add to Louisville's supply. The report rendered by this expert has been much in demand. The Bureau's representative then visited Louisville in hopes of effecting a solu- tion of the problem, which was becoming daily more aggravated. The local situation rendered the Bureau's recommendations as to rationing and the manufactiure of gas impracticable. Appeal was made by the local company to the United States Fuel Administration, and a rationing order similar to those issued for several other locaUties was issued. Although the Bureau did not actively participate in the hearing of this case before the Fuel Administration, it nevertheless is a source of gratification to know that the action of the Fuel Administration was exactly in Une with the Bureau's recommendations to the city. The increasing activity of the natmral-gas division of the Fuel Administration rendered it unnecessary for the P : Mil Scales The mil scale consists of an auxiliary range-finding reticle which is placed in the eyepiece of the military binocular. Fine vertical and horizontal lines are ruled on the glass reticle at definite intervals so that it is possible to make comparisons arid range determinations at the same instant one may happen to ORDNANCE 193 discover an object or target. The scale posesses the further advantage of enabling the artillery to point at an invisible or indefinite target by aiming at some conspicuous object which may be taken as a reference point on the mil scale. The Signal Corps requested the Navy Department to construct in its optical shop large quantities of binoculars equipped with these scales. Difficulty was experienced in producing them, and the Bureau was called upon to assist, and in a few days succeeded in correcting the apparatus so that quantity production became possible. Early in 1918 the Bureau discovered that the scales supplied by one of the largest commercial manufacturers contained serious errors. This discovery was promptly communicated to the Sig- nal Corps and to the manufacturer. The Bureau was authorized by the Signal Corps to confer with tl\g mantifacturer on this point and to see that proper corrections were made. After the matter had been gone over thoroughly, the corrections were made and no further troubles were experienced. ORDNANCE Hardness of Brass Cartridge Cases This subject is treated under the heading " Metalltu-gical Investigations." Machine-Gun Erosion A systematic study of this subject is being made from the metallurgical point of view, including the preparation of a con- siderable number of special tests, determination of the interesting physical properties, and the ballistic tests, which last are being conducted by the Ordnance Department itself. Copper Crusher Gages At the request of the Society of American Manufacturers of Small Arms and Ammtmition and of the Ordnance Department of the Army, the Bureau has been actively engaged in the pre- paration of specifications for copper crusher gages for testing ammunition and standardizing the method of use. A considerable amount of work has also been done on the characteristics of the copper crusher cylinders, particularly as related to their properties under various conditions of annealing and as dependent upon their conditions of precompression, and includes a study of the resulting errors. Several conferences have been held, and it is expected shortly to close this subject. 26035°— 21 13 194 WAR WORK OF THE BUREAU OF STANDARDS Defects of Shrapnel Steel This subject is treated under " Metallurgical Investigations." Light Armot Plate This subject is treated under the heading "Metallurgical Investigations." Synchronizing Devices for Airplane Guns This subject is treated under " Aircraft, Miscellaneous." Removing Metal Fouling from Rifle Barrels This subject is treated under "Chemical Investigations, Miscellaneous." Munitions Gages A complete description of this work is found under the heading ' ' Precision Gages. ' ' Airplane-Gun Mount Airplanes designed for fighting were provided with flexible mounts for the machine guns and a safety belt for the gunner. It was important to determine whether both the moimt and belt were sufficiently strong for the purpose intended. In order to subject these to tests approximating actual conditions, spe;cial equipment had to be provided for attachment to one of the Bureau's testing machines. Through the use of this special equip- ment the mount could be tested in an incUned position, thus approximating the probable average flying condition. The Bureau carried out quite extensive development work along this Une and after a satisfactory form of construction was established performed a number of acceptance tests. Powder-Shipping Containers Some of the powder containers tested were of fiber and others were of sheet metal. The fiber containers were subjected to the following tests: Internal air pressure, submergence, Mullen or bursting strength, and tensible strength tests of the cardboard of the wall. The load required to pull off the crimped-on container ends was also determined in several instances, as well as the effect of the submergence test on the metal of the ends. All metal containers were tested by submergence, air pressure, and in impact by tipping over a container loaded with sand onto the floor and also by dropping a weight on it from increasing heights until failure occiured. As was the case very frequently with all materials, the Bureau was here called on to pass judgment upon ORDNANCE 195 radical modifications of current design in powder containers. Such advice, which was more frequently furnished orally than by correspondence, served as the basis for the Ordnance Department's action in the standardization of their powder-container design. Semisteel Shells Extensive tests were made in an investigation of the possibility of using semisteel or processed cast iron in Heu of steel for shells. Initial experiments of American manufacturers with the use of semisteel were tmsuccessful, although French factories were then turning out satisfactory shells of this material. Careful compara- tive studies were made of samples of American and French manufacture, respectively, which included physical, chemical, and microscopic examinations. The physical tests conducted included tensile, compressive, and transverse strengths, hardness and impact determinations. Suitable criterions were established which served as acceptance bases for the purchase of shells. Quality production of satisfactory semisteel shells had been begun before the signing of the armistice. Copper Rotating Bands for Shells Considerable trouble was experienced by the Ordnance Depart- ment with the copper rotating bands on shells. It was found that some of the bands when formed to the required size and shape were so brittle that they failed in service by breaking. Other bands were so soft that they filled the rifling on the inside of the gim barrel with copper and caused jamming of the shells. The possibilities for the use of arsenical copper in the manufacture of rotating bands were investigated, and numerous suggestions for improving their quality were ftu-nished the Army. Gun Barrel with Gored Rifling One of the special problems investigated was the cause of failure of a 75 mm gun in which a gash had been torn in the rifling when the gun was fired. It was important to discover whether this failure had been due to some imperfection in the metal out of which the gun had been constructed or because of an imperfect shell. In order to do this the gun was tested coincidentally with a 155 mm howitzer which had proved satisfactory in service. The gun with the flaw performed as satisfactorily in the physical tests and microscopic examinatioii as did the other, which appeared to indicate that the flaw resulted from a faulty shell and not from any defect in the manufacture of the gun itself. 196 WAR WORK OF THE BUREAU OF STANDARDS French and American Trench Mortars An extensive comparative investigation was conducted of the relative properties of 240 mm trench mortars of French and Ameri- can manufacture. The study included physical, chemical, and metallographic examinations. The French steel was found to have a finer and more uniform texture with over 10 per cent greater strength than the American material. This superiority of the French material appeared to be the result of better heat treatment, as the foreign steel had a lower carbon and manganese content, which would normally indicate lower strength. Railway Mount Recoil Piston Rods A number of 6 and 8 inch diameter recoil piston rods for railway- mount howitzers were prooftested by the Bureau to loads twice those brought upon them in firing. These rods necessitated the use of approximately the maximum load which can be appUed with the large Emery hydraulic machine, namely, i 1 50 000 pounds in tension. The elongations in a 5 or 6 foot gage length were obtained as successive loads were applied and stress strain curves were plotted for the rods. This is one of tJie few laboratories in the coimtry which is equipped for appl5dng large tensile loads to specimens of great length. Wheels, Investigation of Artillery, Truck, and Airplane This subject is described under the above heading in another part of the report. Magnetic Analysis of Rifle-Barrel Steel This is treated in the section on " Magnetic Investigations." PAPER Wall and Plaster Board The use of wall board and plaster board as a building material for the construction of military cantonments. Government ware- houses, office buildings, and similar structures was developed to a great extent during the war. Wall board is a general term cover- ing those types of building material which are used as a sub- stitute for wood lath and plaster in partitions and for sheathing on the inside walls of buildings. The construction of the vast number of Government buildings demanded in the prosecution of the war required the use of a material that could be quickly erected, having a low cost as well as adequate supply. This necessity was met by the use of over 100 000 000 square feet of wall board. PAPER 197 Wall boards are made of paper, built up of one or more layers or plies that are cemented together with a binding agent such as silicate of soda (water glass) or other adhesive. Plaster board is made of three-ply material. The inner ply is made of hydrated plaster of paris or gypsum, while the two outer plies are made of a special paper. The finished boards are of thicknesses of three- sixteenths, one-fourth, and three-eighths of an inch. The term "plaster board " is applied to that type of wall cover- ing that is made of two layers of a special paper with fire-retarding properties superior to ordinary wood lath and plaster and far superior to the all-paper wall board. The strength of this material together with its fire-retarding properties and other qualities, makes it an excellent building material and one that will imdoubt- edly come into more extended use as its properties and usefulness become more generally known. The term "wall board" applies to only that type of wall cover- ing that is made entirely of old-paper stock, ground wood pulp, or other paper-making fibers. Through the War Department the Bm-eau was requested to assist in the preparation of specifications and methods of testing to determine those qualities desired. Special methods for testing this wall board had to be devised in order to duplicate as nearly as possible service conditions. Samples of all the commercial grades were secured and tested and their relative suitability de- termined. As the War Department had already used great quantities of this material and was contemplating using much more of it, the need of complete data of the service of this type of building material could not be overestimated. The actual service rendered by this wall board was investigated at various cantonments and in buildings hned with wall board and plaster board which had been erected over a year. The result of this gen- eral investigational work has been to greatly improve the quality of the wall and plaster boards used by the Government. These good results have been made possible largely by the hearty cooper- ation of several of the large manufacturers. In order that further information might be obtained as to the behavior of these boards under service conditions, a question- naire was sent to most of the military cantonments. For tempo- rary structures fiber wall board was preferred merely because it was considered quicker and cheaper to erect and because of its greater salvage possibilities. Plaster board was preferred for permanent construction (that is, over five years) because it was 198 WAR WORK OF THE BUREAU OP STANDARDS considered less subject to changes in temperature and because it makes a warmer building. For hospital use ;plaster board was preferred because of its greater resistivity to fire and moisture. In general, it may be said that wherever quality and performance were considered, the plaster board was to be preferred. Paper as a Substitute for Linen in Airplane Construction An original investigation was undertaken in the summer of 1 91 7 to determine what Materials were necessary to produce an exceedingly strong paper which could be substituted for woven fabrics in airplane coverings. In preliminary tests a mixture of jute arid manila rope was found as most promising and had the advantage of being a material which could be purchased in reas- onably large supplies throughout the United States. The investi- gation was divided into two distinct classes: first, the production of a sheet of strong paper, and, second, the investigation of methods for making it waterproof and fire resistant. A series of runs made on the Bureau's experimental paper machine, shown in Fig. 2 1 , indicated that a paper from the rope stock, comparing favorably in strength with textiles, could be produced. The preliminary experiments demonstrated the limit- ing factors in cooking, beating, and machine treatment, and made possible definite standard methods which were adaptable to prac- tice anywhere, making reproduction of the sheet possible. The result of the preliminary investigation was the production by the paper section of a paper of quality equal to the samples submit- ted and of such character that it could be manufactured in widths up to 144 inches continuously on a commercial-paper machine. The survey also showed that there was a reasonable supply of the necessary material available in this country. The method of treat- ment was long and the cost of the paper would be high', but pre- sumably it could be procured at a price lower than woven cotton or linen fabric. The economic conditions were not investigated further than to determine that a supply was easily available. After having made the paper, attempts were made to develop a method of laminating two or more sheets together to produce a single sheet, having a weight of 6 ounces to a square yard, since, as manufactured, the sheets weigh about i % ounces per square yard. It was the idea of the paper section that a glue or adhesive could be developed which would not only securely fasten the sheets together, but would also render them waterproof and to a certain extent firepi'oof . PAPER 199 The idea of discovering a material of this type was worked on, and experiments were made with the following agents : Switzer's reagent, glue, casein and blood mixed, sodium silicate, formal- dehyde glue, bichromate glue, common glue, oil emulsion, zinc chloride, sani dry, and halowax. No satisfactory results were obtained in any instance, with the possible exception of the Switzer's reagent. This agent, however, was abandoned because of difficulty in manipulation and lack of uniform results. However, the Bureau is not at all discouraged over the proposi- tion and believes further investigation might develop a product which would be satisfactory for laminating purposes. In July, 1917, the Bureau was formally requested to produce 1000 square yards of rope paper to be forwarded to the Dayton- Wright Airplane Co. This paper was prepared on exigency order and forwarded at the earliest possible moment to its proper destination. Paper Filters for Gas Masks In February, 1918, there was started an investigation to develop a paper suitable for filtering "sneeze gas," to be used in comiection with gas masks. This work was requested by the war gas-investigation division of the Bureau of Mines and was later continued with the cooperation of the Chemical Warfare Service of the Army. The first problem which was assigned was to duplicate in this country a thin, craped, tissue paper which was at that time being manufactured in England and used by the Allied armies. Development work was at once begun at a paper mill in Pennsylvania, and shortly thereafter a paper was successfully made which duplicated that being used abroad. Further work was done at this mill, and later development work was carried on at a mill in Wisconsin, producing a paper of very much the same kind. Both these attempts were highly successful, but in each case the plans of the Chemical Warfare Service were changed so that this paper was never produced in any great quantity for use by the American Expeditionary Forces. However, data were collected so that this material can be made at either of these mills on very short notice, and it is to be remarked that the organi- zations controlling these two mills practically permitted their equipment to be used as an experimental laboratory, although this necessarily interfered with routine work. In order to meet certain changes in the shape of the canister desired by the Chemical Warfare Service, a paper which was very 200 WAR WORK OP THE BUREAU OP STANDARDS satisfactory as a protection against "sneeze gas" was developed on the paper machine of this Bureau, and sufficient data were collected so that this paper can be developed here when needed. However, it would probably take some time to duplicate this paper on a commercial scale, although there is no doubt that it could be done in the course of a comparatively short time under the stress of war requirements. In order to assist in the efficiency of this investigation, apparatus was set up at the Bureau with the help of the Chemical Warfare Service by means of which the paper developed was constantly- tested so as to incorporate in its production any improvements which appeared desirable. In this connection a gas house was built for the purpose ol con- ducting actual service tests of gas-filter paper as well as for determining the concentration of the gas by means of the nucliea- tion apparatus. While this work was in progress a large number of commercial papers were studied to determine whether any grades of paper were being manufacttwed which would be suitable for the purpose. It was felt, however, that a very special paper was necessary and that there was no paper on the market suffici«itly porous to permit a man to breathe through it, and at the same time sufficiently compact to retain the small particles of " sneeze gas " which were injurious to the limgs. As a result of these investigation it would appear that two distinct types of paper can be made in the United States on a commercial scale, both of which would be satisfactory as a protec- tion against " sneeze gas " in the gas-mask canister. One is a very thin craped tissue, about 70 thicknesses of which are used, while the other is a relatively thin porous paper, about 10 thicknesses being needed. In both cases the chief obstacle in the use of paper is the mechanical difficulty of attaching it to the gas-mask canister. Paper Containers for Axle Grease and Saddle Soap In the interest of the conservation of tin the paper section of the Bureau was requested to investigate the suitability of paper containers for materials such as axle grease and saddle soap. With the aid of the war-service committee of the paper industry a line of samples representative of all the cartons now on the market and at all suitable for this purpose was secured. These ranged in size from a small box capable of holding about 4 ounces to a barrel of 30 gallons' capacity. PAPER 20 I In general, the packages submitted could be divided into seven different classes: (i) ordinary cardboard boxes with slip-on covers; (2) cardboard boxes made from a heavily sized stock; (3) boxes lined with paraffin or oiled papers; (4) boxes Uned with parchment papers; (5) boxes impregnated with paraffin; (6) boxes carefully treated with a suitable grade of varnish; and (7) boxes composed of three-ply board, the middle one being an asphalt- impreganted ply. Classes 2, 4, 6, and 7 were found to be the most suitable for the axle grease. Any box seems to be possible for saddle soap, but the paraffiai-impregnated boxes were less desirable because in some cases the heat caused by pouring in the melted soap was enough to cause the paraffin to run. Where a plain unsized or slack-sized cardboard box was used, the soap stuck to the fiber of the box, and thus there was some waste in removing the soap. The very Uquid nature of the soap tended to soak off any parch- ment Uning on the boxes, so that a box of this kind does not have the advantage for saddle soap that it has for axle grease. Asphalt as a coating material was found to be distinctly unsat- isfactory, both because it had little effect in lessening the moisture resistance and because it had a decided tendency to peel off and contaminate the contents of the box. An asphalt-impregnated paper used as a middle ply was reasonably satisfactory, and was free from the disadvantages mentioned above. The varnished box (No. 6) was decidedly the preferred type. The varnish coating used was of such a type as to be very flexible. Some boxes coated with sodium silicate were received, but these were too brittle. The tests were made on the assumption that the boxes of grease would be carried in a warm place, as, for example, under the hood of an automobile truck, or exposed to storage conditions in warm places, as in a steamer's hold. The boxes for soap were tested on the assumption that they would be filled by pouring in the soap in the melted form and allowing it to harden. It is thought that class No. 2 boxes made from heavily sized stock would be sufficiently good for both purposes except under extreme conditions, and they are considerably cheaper than some of the more elaborate forms. Round containers with slip-on covers are considered most suit- able except where economy of shipping space is required. The covers should be of such a nature as to slip on and off very easily 202 WAR WORK OP THE BUREAU OF STANDARDS and without injury to themselves or the box. This ia£t renders the complicated folding and locking covers open to criticism. ■ - . A very interesting field has been opened by the paper barrels which were received in this connection. In general, these were of two types: One is a barrel intended for the shipment of pitch or asphalt and as such has a comparatively low strength: when empty- It is made entirely of pulp and is in one piece. : The other is made of paper wound upon a mandrel under tension and is designed to cover a wide range of use. Paper-board heads are put in after the sides are completed ; This barrel has considerable strength and will bear the weight of a man standing upon its side, even when empty. PHOTOGRAPHY General Photography It is well known that the ordinary photographic plate when exposed in a camera does not reproduce exactly what the eye sees. This is because the ordinary plate is affected by only blue and violet light, for which the eye is relatively Insensitive, and is scarcely at all affected by the green and yellow, for which the eye has its maximum sensitivity. These ordinary photographic plates are practically insensitive to red light, and are, therefore, handled with safety in light from a ruby lamp. It has long bpen known, however, that photographic plates may be made sensitive to green, yellow, and red light by the admixtture of suitable dyes to the emulsions coated on the plates, but the incorporation of such dyes in the emulsion generally reduces the sensitivity of the plate to blue and violet Hght, so that until very recently the com- mercial orthochromatic and panchromatic plates have been slower than the ordinary plates. Their use was, therefore, Umited by the relatively long exposure required, until recent develop- ments along the general Hues to be described in the following paragraphs resulted in remarkable increases in the sensitiveness to yellow and red Ught. A plate sensitive to these longer waves possesses some marked advantages over the blue and violet sensitive plate. A pan- chromatic plate reproduces more faithfully what the eye sees, and when exposed behind a ray filter or color screen it portrays objects which, on account of haze, smoke, etc., would be hidden from an ordinary plate, and also gives different contrasts depending on the nature of the light transmitted to the plate through the filter. The phenomenon of haze penetration by the longer waves is known to most outdoor photographers. Not only haze due to water vapor, but smoke haze, which looks bluish to the eye, is PHOTOGRAPHY 203 largely eliminated by using light of longer wave length than that which is most effective in case an ordinary unscreened plate is used. Objects completely obscured by haze or smoke on an ordinary unscreened plate may be shown in good detail by using a panchromatic plate and a red filter. Such red sensitive plates, of course, can not be handled in the presence of the well-known dark-room ruby lamp, but must be developed in complete dark- ness or in a very week green light. To obtain the long-wave advantages of haze penetration and contrast mentioned above, the photographic plates must have the greatest possible sensi- tiveness to yellow, orange, and red light if the exposure times are necessarily short on account of rapidly moving objects or if the camera itself is in rapid motion, as is the case in aerial photog- raphy. When corrected camera lenses giving very bright images are not to be had, there is another argument for using photographic plates of the highest possible sensitiveness. Photography as Applied to Spectroscopy It is common knowledge that white light consists of a mixtiu-e of colors of different wave lengths of light, and most colors in nature are also compounds of several colors. These facts are observed when such soiurces of light are dispersed or spread out into spectra by suitable apparatus such as prisms or defraction gratings, and the analysis of such composite Hght is the domain of spectroscopy. The regular observations in spectroscopy in all spectral colors are now nearly always made by the use of photog- raphy; hence every modem spectroscopist is a photographer, and to photograph the entire range of spectral colors from the shortest ultra-violet to the long-wave infra-red requires a much broader knowledge of the numerous complex underlying elements of photography than that possessed by the average professional photographer. The names of Huggins, Draper, Vogel, Eder, Abney, and Schuman call to mind some remarkable applications of photogra,phy which were inspired by spectrocsopic researches. Photographic Spectroscopy at the Bureau of Standards The spectroscopy section of the Bureau of Standards began in 1 91 4 an extensive program of measuring standard wave lengths throughout the entire range of the spectrum which can be ob- served photographically, and some of this work was published in the following scientific papers of the Bureau: S 251. Interference measurements of wave lengths in the iron spectruni (28S1A- 3,70iA) with notes on comparisons of lengths of light waves by interference methods, and some wave lengths in the spectrum of neon gas. 204 WAR WORK OP THE BUREAU OF STANDARDS S 274. Interference measurements of wave lengths in the iron spectrum (3233 A- 6750A). S 302. Wave lengths of stronger lines in helium spectrum. S 329. Measurements of wave lengths in the spectrum of neon. Then it was observed that although observations of spectra were extensive in the so-called chemical regions — ^that is, in the short-wave lengths from ultra-violet to blue — and were fairly- numerous in the green and yellow, there existed a great scarcity of data for the red and infra-red, these latter being the regions for which highly sensitive photographic plates could not be pur- chased. Accordingly, the Bureau of Standards decided to pay special attention to spectroscopic observations in the neglected regions of long waves. This involved the preparation and use of photographic plates specially treated to make them sensitive to red and infra-red light. Such sensitized plates have been success- fully used at the Bureau since 191 4 in a systematic study of the spectra of about 50 of the chemical elements. Some of the results of these studies have appeared as scientific papers, among which may be mentioned No. 312, "Wave-Length Measurements in Spectra from 56.00A to 9600A;" No. 324, " Wave-Ivcngth Meas- urements in the Red and Infra-Red Spectra of Iron, Cobalt, and Nickel Arcs;" and No. 345 "Measurements of Wave-Lengths in the Spectra of Krypton and Xenon." Applications of such sen- sitized photographic plates were also made in connection with other scientific problems, as, for example. Scientific Paper No. 327, "Measurements on the Index of Refraction of Air from 2218A to 9000A," and especially to astrophysical problems. These later include Scientific Paper No. 318, "Application of Dicyanin to the Photography of Stellar Spectra," and "Photog- raphy of the Solar Spectrum from 6800A to 9600A," Astrophys- ical Journal (47, p. i, 1918); "Solar and Terrestrial Absorption in the Sun's Spectrum from 6500A to 9000A," publications of the Allegheny Observatory, and an attempt at Baker City, Oreg., to photograph the red and infra-red flash spectnun of the stm at the time of its total eclipse on June 8, 1918. Military Problems in Photography During the war aerial observation early came to play a promi- nent part, and it became a matter of great importance to have a suitable panchromatic plate for aerial photography. Labora- tories and research organizations in the allied countries were busily engaged in trying to produce plates of extreme sensitive- ness or speed. The best commercial orthochromatic (sensitive PHOTOGRAPHY 205 to blue, green, and yellow) and panchromatic (sensitive to all colors) plates were too slow for most purposes. For many years, however, physicists had used yellow and red-sensitive plates which greatly exceed in speed any of the commercial plates. These plates are prepared by bathing an ordinary photographic plate (sensitive to violet and blue) in certain solutions of aniline dyeg, and at the expense of keeping qualities such plates may be made much more sensitive to red light than any plate known to commerce. After several years of experience with such plates in spectroscopic investigations by the Bureau, it seemed worth while to make an application of them to landscape photography when the United States declared war upon Germany. Accord- ingly, experiments were made in 191 7 to determine the practi- cability of using the bathed plates for landscape photography with the hope that either this method might be a useful addition to the photographic methods of our military forces or that the experiments would lead to improved conunercial plates which would incorporate sensitizing dyes in the photographic emulsions. With apparatus that was inadequate to test the real merit of the bathed plate, a number of photographs of various objects were made which were encouraging enough to warrant a con- tinuation of the investigation, using better apparatus (more suitable cameras and ray filters) and applying to the staining bath new dyes of British manufactxure. The attention of the science and research division of the Signal Corps was directed to these early results, and the Bureau was given the opportunity of comparing the stained plates with various plates used by the military forces. Among these were several types of British-made panchromatic plates which were then in use on the western front for aerial photography. A series of experiments soon showed that pinacyanol stained plates were at least four times as fast as the best commercial panchro- matic plates then in use. In particular, plates stained with some new dyes of British manufacture were found to be much superior to any commercial panchromatic plates, not only in speed but also in the range of color sensitiveness. These encouraging results led to the offer of the Bureau to supplement the photographic w6rk of our military Air Service by adding to it a new method which would probably be of great military importance in special photographic work, such as the penetration of haze and smoke, detection of camouflage, etc., whore the ordinary commercial plates could not be satisfactorily 206 WAR WORK OF TJHE BUREAU OF STANDARDS used. The Bureau received the assurance of cooperation from the War Departanent and was invited by the Signal Corps to test the bathed plates in aerial photography as soon as facilities per- mitted. Extensive experiments were made at Langley Field during the spring and summer of 191 8 under the generous auspices of the science and research division of the Signal Corps, and after the armistice was signed the flying facilities of Boiling Field were used to a limited extent to complete the experiments. Early in 191 8 a representative of the spectroscopic section of the Bureau, who was in France, was requested by the French military authorities to assist the French photographic section in making fast color-sensitive plates. A temporary laboratory was installed similar to the ones at the Bureau of Standards with the exception that the plate-drying cabinet was arranged so as to use air heated by a gasoline stove. This was necessary on account of the low temperature and high humidity prevailing in northern France in winter. The following experiments with bathed plates were conducted by this laboratory. Ordinary fast plates were stained with pinaverdol and pinacyanol with ammonia, ortho- chromatic plates were bathed in pinacyanol with ammoma, pan- chromatic plates were bathed in ammonia. Each type of plate was used in three ways: Unscreened, with a deep yellow screen, and with a red screen. The bathed plates were foimd to be somewhat faster unscreened than any plate actually used by the French. Using screens, the bathed plates were shown to be much faster, particularly for red photography. This matter was also considered of very great importance by the American Aviation Section, and Col. Dunwoody requested that one of the Bureau's physicists cooperate with them on experiments of this kind. This cooperation was offered, but was not strongly supported by the military authorities in America, largely on account of the fixed idea that bathed plates had only reached the experimental stage. Considering the special equipment and technique required by such plates, their use at the battle front was held in abeyance with the hope that commercial panchromatic plates of sufficient speed would be forthcoming. The success with which dye-sensitized photographic plates had been used in the Bureau's laboratories, however, convinced those engaged in this work that such plates had possibilities of extreme importance in special military applications and impelled the Bureau to demonstrate their value by careful measurements in the laboratory and tests in the field. In addition to the problems PHOTOGRAPHY 207 of ordinary photography, the use of bathed plates with light filters involved (r) a study 'of photosensitizing dyes and of the spectral sensitivity which they imparted to ordinary photographic plates, (2) experiments to increase still further the speed of color-sensitive plates, (3) investigations of the spectral (iistribution of energy reflected from landscape, etc., (4) design and construction of new photographic lenses for use with red light, and other similar problems. These problems have been investigated by the Bureau with such success that the importance of dye-sensitized plates in military photography is now admitted by all, and the practi- cability of their use has been thoroughly demonstrated. The great advantage of these special plates in the taking of photo- graphs from an airplane under adverse conditions is illustrated by Fig. 22. Photographic Method of Detecting Camouflage The Bureau, in cooperation with the United States Signal Corps and the United States Air Service, utilized a photographic method of detecting camouflage. The method consists of making two photographic negatives of the group of objects, one before the change and the other afterwards. A positive is made from one of the negatives, and this is superposed upon the other negative. If no change had occurred between exposures, the combination would form a field of practically uniform photographic density; but cha;hges are plainly indicated by discontinuities of density. It is believed that the method has considerable commercial value for engineering and detective work, etc. Protection of Moving-Picture Film from Heat of Lamp A short description of the work done by the Bureau on this problem will be found in the article on " Radiometry." Metallic Mirrors for Cameras This subject is likewise treated in the article on " Radiometry." Automatic Regulation of Diaphragm by Intensity of Light A paragraph dealing with some work done by the Bureau on the practicability of a device for this purpose is described in the article on " Radiometry." Production of Dense Barium Crown Glass for Lenses The experiments which have been conducted in the Bureau's glass-maldng plani; on the : production of glass for photographic lenses will be foimd in the article on " Qptical Glass and Optical Instruments." 208 WAR WORK OP THJB BURBAU OP STANDARDS Tests of Photographic Lenses and Camera Shutters The experimental work performed in connection with the testing of photographic lenses and shutters is treated in the article on "Optical Glass." PROTECTIVE COATINGS Electroplating Investigations Even before this country entered the war, numerous inquiries for information upon plating of mihtary supplies were received by the Bureau, and a few preliminary investigations were made upon special appKcations of plating; for example, the production of copper propeller tips by electroplating. Soon after the begin- ning of the war it became evident that electroplating was being extensively required upon military suppUes of most varied; de- scription, and that th^re were almost no specifications or methp^s of inspection relating to such work. Owing to the fact tha^ in almost every case the plating was considered as an incidental and minor part of the manufacturing process, very Uttle consideration was given to this subject by military officials in the early part of the war. In consequence, confusion and misunderstanding often arose in connection with the plating operations. There was lield at the Bureau of Standards on March 27, 191 8, a conference of military officials and manufacturers interested in electi-oplating. After a detailed discussion of the methods and requirements of the plating operations certain definite recommendations were made which were incorporated in a report of the conference, copies of which have been widely distributed. Among the recom- mendations of this conference was that one or more experienced platers should be engaged as plating advisors on military supplies. In accordance with this suggestion two experienced electroplaters were engaged by the Biureau and devoted most of their time to visiting various plants engaged in the manufactm-e of supplies, to assist them in producing satisfactory plating. To a great extent the fxmction of plating upon mihtary sup- pUes was to protect metals, especially steel, from corrosion. Therefore it was necessary to devise and adopt methods of testing by which the relative value of plating as compared with other protective coatings could be determined. This section therefore cooperated actively with other sections of the Bureau in the preparation of specimens to be used in comparative corrosion tests. The principal test used for this purpose was the salt-spray test, by means of which it was readily shown that zinc coatings Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 22. — Views from an airplane taken with an ordinary photographic plate, and one specially prepared at the Bureau's laboratory The penetration of haze by these special plates is clearly illustrated in the two pictures, taken at the same time of the same objects. Such plates were orifiinally developed for use in spectroscopic work, thus illustrating the practical application of a scientific achievement Miscellaneous Publications, Bureau of Standards, No. 45 *f f*-^^. Fig. 23. — Radio direction finder or coil aerial used in measuring the distortion of radio waves in the vicinity of trees, towers, wires, buildings, etc. Fig. 24. — Model set-up in the radio laboratory to illustrate the use of the radio direc- tion finder Radio signals are sent out from the two lighthouses and the direction from which they come is determined Idv the apparatus on the ship. The latter can therefore determine its position even in fog and under other conditions which render the lighthouse lamp invisible PROTECTIVE COATINGS 209 exert the best protection against corrosion of steel, and that satisfactory zinc coatings may be produced by either hot dipping, shei^dizing, or zinc plating (so-called electro galvanizing) . Accordingly, in various specifications the Bureau urged the use of zinc coatings, not specif}Kng the method of producing these coatings, but requiring a certain resistance in the salt-spray test. In connection with the cooperative work with the Ordnance Department of the Army and with the committees on standardiza- tion of supplies for the War Department, it became evident that there was a lack of coordination between the various protective coatings required for different articles and for different kinds of service. A suggested classificauon of finish for metal parts was therefore proposed for use by the Ordnance Department and other branches of the War Department. The basis of this classification was the use or degree of exposure to which the articles would subsequently be subjected. While the proposed plan was never formally adopted by the military officials, it served in numerous cases as a guide in the specification of metal finish. This classifi- cation, together with a review of military applications of electro- plating, was presented at a meeting of the Electro-Chemical Society and was also published in Metal Industry for November, 1918. The three kinds of plating which were of most importance from the military standpoint were zinc, lead, and black nickel, upon all of which investigations were conducted at the Bureau, and information was furnished to manufacturers and to military officials. When the value of zinc coatings for protection of steel against corrosion was fully realized, its use was extended to a wide variety of articles. For example, naval airplane parts, material used in shipbuilding, hardware for ammunition boxes, and parts of the fuze mechanism for high-explosive shells. It was found that satisfactory deposits could be produced, from both cyanide and sulphate solutions, but that the cyanide solutions were usually more satisfactory for irregular shapes or parts with deep recesses. Except for a few plants engaged in the manufacture of storage- battery fittings, lead plating was until recently almost a scientific curiosity. It was used very satisfactorily, however, on an exten- sive scale for plating boosters and adapters for gas shells and for lining certain of these shells. Another important application of lead plating was on the inside of underweight shells, whereby thousands of otherwise rejected shells were salvaged. A pre- 26035°— 21 14 2IO WAR WORK OP THE BUREAU OP STANDARDS liminary circular giving detailed information upon the applica- tions of lead plating was prepared by the Bureau and extensively circulated. In connection with this work the application of lead plating for lining chemical apparatus was investigated, and in cooperation with the officials of the Edgewood Arsenal a large tank was plated with lead to a thickness of 0.07 inch. Unfortu- nately, this experiment was finished too late to permit the general application of the process on the equipment at Edgewood before the signing of the armistice. A large amount of hardware and equipment used by the Gov- ernment is required to have a black or gray-black finish, the so-called "Government bronze." This is usually produced by the process known as black nickel plating. It may be applied to brass directly or after copper plating and to steel which has been previously plated with copper or with zinc. Great difficulty was experienced by the manufacturers in furnishing a satisfactory finish, but after a great deal of work the Bureau was able to throw considerable light upon the behavior of such solutions and to devise simple methods of operation and testing. There were many demands for information on copper, nickel, and tin plating which frequently required visits to the plants in order to determine and remove the cause of the difficulties experienced. In addition to the normal applications of plating many imusual problems arose in connection with which the Bureau was able to furnish some assistance. Among such problems may be men- tioned the following: In cooperation with the Bm-eau of Mines the Bureau assisted in the development of a process by which it was found possible to produce heavy nickel articles, such as seamless tubes, etc., for which there was a considerable demand in the nitrate manufacture and the separation of helium. It was arranged that the Govern- ment was to have the use of this process during the war and for six months thereafter. So far as is known this method has never been operated upon a manufacturing scale. It was found that by a simple electrotyping process it was possible to reproduce at a very slight cost the master range finder scales used for producing the lines upon range finders. The Bureau cooperated with the New York Navy Yard and assisted them in securing the equipment and personnel for pro- ducing various plates by electrodeposition. PROTECTIVE COATINGS 211 The Bureau was requested to investigate the possibility of relining worn-out guns with a sufficiently heavy coating of nickel, iron, or cobalt to permit' the rerifling of the gims. The limited force available prevented any exhaustive research upon this problem, the solution of which on a large scale does not seem to be very promising. More recently a similar problem has been submitted, namely, of relining defective recuperator cylinders used on 75-mm. or larger guns. In the preparation of these cylinders it is necessary to have an absolutely smooth surface, which is obtained by "lapping." In the process of manufacture many defective cylinders have been rejected, and it is desired, if possible, to salvage these cylinders by plating a sufficient coating of nickel or copper to permit relapping to a true surface. The possibilities of this process are now being investigated. The Bureau was called upon on numerous occasions to assist in the design of plating equipment to be used for military sup- plies both in privately owned and Government plants. Among the latter may be mentioned the New York Navy Yard, Marine Depot in Philadelphia, and the Naval Aircraft instruction schools. In addition visits were made to the principal arsenals and navy yards and recommendations were made regarding the equipment and processes for such work. Bituminous Materials The bituminous products are remarkably resistant to acid and alkaline attack, their physical properties vary from Hquid asphaltic oils and tars to hard and brittle asphalts and pitches, and they are susceptible to modification through this range by blending and fluxing. These physical characteristics have given them a wide and general application in building construction, and they were called into important uses by the military authorities not only for such recognized pmposes as prepared and built-up roofing, preserva- tion of timber and waterproofing, but specifically as a coating for steel and concrete ships, a coating for hand and rifle grenades, for marine glues used in the construction of pontoons for naval air- craft, etc. In the preparation of spedflcations for prepared roofing it was reaHzed that war conditions had seriously interfered with the normal production and quality of the raw material used in its manufacture, and it was necessary therefore to make the specifica- tion less exacting than usual so as to obtain the essential produc- 212 WAR WORK OE THE BUREAU OF STANDARDS tion. The pi'esent specification was developed in cooperation with the War Industries Board and manufacturers producing prepared roofing and includes instructions to inspectors as well as methods of inspection and test. lyarge quantities of material were ob*- tained imder this specification which were used on cantonments, hospitals, and warehouses built in various sections of the country. The inspection and testing of this material required a great deal of laboratory work, some thousand samples having been examined, representing a milUon squares of roofing. As might be expected many leaks developed in these roofs, but from such information as could be obtained they appeared to be due to faults or carelessness L in laying rather than defective material. Several cases were investigated by members of the Bureau in cooperation with ofiicers of the construction division of the Army and recommendations for repairs were made which did not necessitate rela3nng. The recommended method of repairs ^ in- cludes a specification for a roofing cement to patch holes or larger areas in which the exact position of the leak can not be located. This practice is now being followed and results in satisfactory repairs at minimum expenditure of time and money and with maximum convenience. jj A specification for built-up roofing is still in process of develop- ment. This type was not so urgenty required ..during the war, and with the signing of the armistice the immediate need passed. There is a normal demand for a specification for this type of roofing by the Government, and it is expected that those inter- ested in the subject will be called together to consider the develop- ment of a uniform specification as soon as other more urgent demands permit. For covering the interior of steel ships, including the coal bunkers, it was necessary to produce a coating which would protect the metal effectually against corrosion in places; which were not readily accessible for inspection and which were always wet with salt or bilge water. Likewise the coating must adhere without cracking or flowing when the vessel is laid up during cold winter months and must not be injured by the severe conditions encountered in the engine and boiler rooms. Such investigations required a study of the service conditions, the development of tests which would duplicate these conditions, the production of material in the laboratory to meet these tests, and finally the preparation of specifications to insure the quality of the material accepted. PROTECTIVE COATINGS 2 1 3 Considerable work has also been done in connection with con- crete ships. Such vessels had never before been used to any extent, and required special treatment. A ship is subjected to great and various strains and can not be made perfectly rigid, so that, although concrete is usually considered a rigid material, it was called upon in this case to act as an elastic substance, and with proper reinforcement was found to stand much distortion without failure. Small hair cracks may develop in the concrete, and one of the problems presented was to cover the outside of the structure with an elastic waterproof membrane which would bridge such small cracks. The coating must not be affected by water, must not flow when subjected to the pressure of the water when the vessel runs at maximum speed, and must present a suitable surface for the application of antifouling, boot-topping, and camouflage paints over it. All of these conditions were considered and a number of products tested in an effort to obtain one which seemed to meet conditions of service best. Another problem in connection with concrete ships was the production of a satisfactory inside coating for oil tanks. Many of these ships were to be used as tankers for the transportation of a variety of crude and refined oils. The problem was compli- cated by the presence of water in the oil and also by the fact that these tanks might frequently be filled with water as ballast on return trips. This, of course, made necessary a coating resistant to both water and oil. After an extended investigation it was recommended that a long oil varnish, commonly called spar varnish, be adopted for the inside coating, and a bituminous coating for the outside. It was also found that treating the concrete surface with a solution of magnesium fluosilicate was an advantage, as it had a hardening effect and gave a better bond. A satisfactory marine glue was required both by the Army and Navy to cement canvas to light pontoon construction, as the foreign product was no longer available. The material must be adherent and elastic after frequent wetting and drying, must remain tacky and pliable and be waterproof and durable. Many commercial products were tested and analyzed so as to know what materials could be used in producing a satisfactory product. As a result, a specification was developed in which the require- ments were based on performance so as to insmre a satisfactory material with the minimum restriction on composition. 214 WAR WORK OF THE BUREAU 0;p STANDARDS Hand and rifle grenades required a coating which would dry hard on a greasy surface, as it was not practicable to clean them after machining. A hot coal-tar pitch of relatively low penetration and high melting point was applied to both the inside and outside of the grenades and excess drained off. The coating for the inside of high-explosive shells required different characteristics. It had to be free from any lead or manganese compounds, should not dry hard, but remain tacky, so as to bond with the molten T.N.T. with which the shells were charged, and should not have any effect on the efficiency of the explosive. A Gilsonite fluxed with a soft Bermudez asphalt and thinned with a slight volatile petroletun product appeared to give a satisfactory coating. A war-time specification was also prepared for the construction division of the Army for creosote oil for wood preservation. This specification was not put in general use, as contracts had been placed prior to issuing of the specification, and much inferior material has been used on that accoxmt. Varnishes Early in 1 9 1 7 the Bureau rendered a report to the Aviation Section of the Signal Corps showing the relative waterproofing qualities of different treatments on wood. This information, which was re- quested by tiie chief signal officer, showed that a high grade of spar varnish was the most satisfactory transparent coating for waterproof- ing the wooden parts of airplanes and forprotectingthedopedfabric of the wings. The varnish which seemed to answer the require- ments most satisfactorily consisted of a china-wood, oil-resin varnish, and at this officer's request a specification, based entirely on physical requirements of the film, was drawn up, which appears to have been the first specification for spar varnish the basis for which was performance rather than composition. In October, 191 7, there was a meeting at the Biu-eau of representatives of the various varnish companies and of the Bureau of Standards. This committee constituted the varnish subcommittee of the standard committee of the Society of Automotive Engineers, and the com- mittee adopted a specification very similar to the one furnished by the Bureau's laboratory to the Signal Corps. About this time the various branches of the Aviation Section began calling for lists of varnishes that had been approved under this specification. The Bureau had made examinations of many samples of varnish under this Specification and had complete PROTECTIVE COATINGS 215 data on exposure tests, so that a small list of approved varnishes was immediately available. A very large number of samples to be examined continued to be submitted, and after exposure tests were completed, those which proved satisfactory were added to the approved list. As a result there was always available for the chief signal officer a continually growing list of approved varnishes that could be ordered by any of the airplane manufacturers on brand. In December, 191 7, a chemist from the Signal Corps laboratory, part of which was then located at the Bureau, was detailed to help in carrying out the very great amount of experimental work and testing which was required by the more than 200 samples of var- nish which were submitted for test. In May, 191 8, when the Signal Corps laboratory was moved to Pittsburgh, all the work which had been carried on at the Bureau was transferred to the new laboratory. At a conference with representatives of the allied nations, the Specification for spar varnish for airplanes, with practically no changes from that originally prepared, was approved and adopted by the International Aircraft Board. The same specification has since then been adopted by practically all branches of the Govern- ment, including the Bureau of Construction and Repair, Navy Depaortment; the United States Shipping Board; the United States Railroad Administration; the General Supply Committee; and the War Department. This specification has proved to be one of the most generally applicable to all classes of high-grade varnish for all purposes, although the specification finally adopted by the Signal Corps of the Army contained several slight modifications, which were made after the removal of its laboratory to Pittsburgh. In order to prevent the explosive ■ charge in high-explosive shells from acting on the metal and to prevent rotation of the charge independently of the shell, it is necessary that the interior cavity of the shell be coated with a varnish. Such a varnish should be free from lead and other heavy metals, so that no reaction will set in between the explosive charge and the coating. The specifi- cation to cover a satisfactory varnish of this type was furnished in April, 1 91 8, to the explosives section. Engineering Division, Ordnance Department of the Army, for a shellac-resin mixture cut in denatured alcohol. Such a varnish had previously proved sat- isfactory. Another specification covering material for this pur- pose, but which did not limit the manufactturer to any specific 2l6 WAR WORK OP THE BUREAU OP STANDARDS coftiposition and which would have the proper softening point to form a good bond between the shell and the explosive in hot loading, was also furnished to the Ordnance Department. . For coating the metal ends of powder containers the Ordnance Department had specified an acid-proof lacquer. This caused considerable confusion. Since all oily resinous materials contain a certain amount of organic acidity, and the Bxu-eau was called upon to settle this question of dispute between the manufacturers furnishing the lacquer and the War Department inspectors, who were guided by the existing specification. As a result the Bureau recommended spar varnish conforming to the airplane spar var- nish specifications as the most suitable coating for the metal ends as well as for the fiber container, where it is crimped into the endsi The coating of the interior of rifle and hand grenades presented a somewhat different problem from the coating of the high-explo- sive shells. In the process of manufactiure the grenades are sent to the loading plant in a very greasy, dirty condition. In order to expedite production it was not feasible to clean the grenadesi and a coating material was therefore desired that woidd adhere to the greasy interior surface. For this purpose certain loading stations had been using a varnish which consisted of a solution of resin in a certain type of gasoline which resulted as a by-product in a manufacturing process. This gasoline was available only in limited quantities, and in order to overcome this difficulty samples of the gasoline were submitted to the Btu-eau. It was found that it was nothing but ordinary automobile gasoline containing a con- siderable quantity of sulphur, free and in solution. In experi- ments with the varnish made from this gasoUne and resin it was found that the varnish had no particular virtues in adhering to greasy surfaces and that good continuous films on the interior of grenades were not always obtained. However, when a solution of resin in solvent naphtha was used, a much better film was obtained. A recommendation to this efiEect was made to the trench-warfare section of the Ordnance Department. It later developed that the manufacturing process allowed heating of the grenades, and this enabled them to use a bituminous varnish which appeared to be more durable. Enamels Soon after extensive varnish testing was under way, it was realized that the use of an enamel would much better protect the doped fabric of airplane wings, and such an •enamel should be PROTECTIVE COATINGS 217 light in weight, opaque, and very durable; so experiments on various types of enamels were begun. As a result of the prelimi- nary information obtained the Bureau of Construction and Repair, Navy Department, drew up a specification for airplane-wing enamel which required the necessary physical qualifications of a satisfactory enamel and limited the manufacturer to a specific composition of pigment. It developed that it was impossible to use certain types of pigments with airplane spar varnish, which had been proved to be the most satisfactory vehicle. Further experiments were made to determine which pigments worked most satisfactorily in a vehicle of spar varnish, and in cooperation with the Biureau of Construction and Repair, which made similar experiments, the airplane-wing enamel specification was finally changed so as to retain the physical requirements of the original specification, but eliminated the composition requirements of the pigment. SpecifijCations for high-grade air-drying and baking enamels were also drawn up for various branches of the War Department. These had for their basis the spar-varnish specification and in general called for suitable pigments in a vehicle for spar varnish conforming to the airplane spar-varnish specification. The air- drying enamels, especially the low visibility gray, were used exclusively on the hydroplanes manufactured by the Navy Department, and black enamels, both air-drying and baking, were used for finishing the metal fittings on hydroplanes as well as for giving a black finish on military hardware parts. A blue air-drying enamel consisting of a pigment in airplane spar varnish is used for coating the wires and cables of airplanes. Several samples of baking Japan that were to be used for coating tte cases of submarine mines were examined for the Bureau of Ordnance of the Navy Department. These Japans were backed on at very high temperatures (between 250 and 300° C) and had to resist the severe action of being submerged in salt water. Slushing Oils Shortly after the armistice was signed, the Bureau received many requests on rustproofing compounds which coidd be used in storing large quantities of equipment of ah types. Examina- tion of some of the samples of rust preventives which had been submitted to the Bureau showed that they varied greatly in com- position and in quality. After many practical tests specifications were drawn up to cover transparent slushing oils of brushing con- 2l8 WAR WORK OF THE BUREAU OP STA>fDARDS sistency and dipping consistency, and opaque slushing oils of brushing consistraicy. These specifications did not limit manufac- turers in any way as to the composition, but were based entirely on practical tests. A small mimeograph pamphlet on " Notes on the Protection of Metal in Storage" was prepared. This dealt in a general way Avith the question and included the specifications. Technologic Paper No. 176 has been issued recently and deals more fully with slushing oils. Fireproofing of Fiber Board Some of the Army hospitals were constructed from a water- proofed wood-fiber board. This construction proved to be a very great fire hazard and resulted in the complete destruction by fire of one or more of the hospitals. The problem of fireproofing the material was given to the Bureau, and experiments showed that when treated with a solution of sodium silicate and water the combustibility was very much reduced. It was recommended that as far as possible the buildings constructed of this board be discarded and ftirther use of this building material be dis- continued, but that where the buildings could not be vacated, all the interior walls be treated with a spray coat of a fireproof- ing solution of water glass. Fire-Retarding Paints In the summer of 191 7, at the request of the Bureau of Yards and Docks, Navy Department, work was commenced on a number of samples recommended as fire-retarding paints. On November I, 1917, a report was made to the above Bureau, and since that time a number of other fire-retarding paints have been tested for various branches of the War and Navy Department?. The sub- sequent work has only tended to confirm the original conclusions, which are as follows : Since this investigation was intended to give information regarding fire-retarding treatments for interior surfaces, the tests were not designed to give information as to the durability of the coatings on exposm-e out of doors; but since any interior surface may be subject to an occasional wetting, it is believed that the relative behavior before and after soaking in water is of importance. It should be noted tbat the per cents of efficiency in fire retarding, from the arbitrary methods necessary in making the calculations, are relative only, those having the higher ratings showing the greater resistance; but the difference between the best and the poorest is actually small. The value of any particular treatment can best be judged by reference to its relation to well-known common methods of coating wood. Thus, whitewash ranks among the most resistant materials, Indian red in linseed oil among the least resistant, and white lead and zinc oxide in linseed oil are intermediate. The general conclusions frpm this investigation are that while practically all paint coatings have some fire-retarding action, none of the materials examined afford very great protection. All the samples tested were materially damaged by application PROTECTIVE COATINGS 219 of the flame for a few seconds, It is noted that both sodium silicate and whitewash rank relatively high. Both of these have the advantage of cheapness, and there is no reason why both can not be used on the same surface. It is believed that for interior wood surfaces the application of sodium silicate solution followed by a coat of whitewash may be more efficient than either one used alone; but, as mentioned before, no treatment of wood after erection can be expected to act as an efficient fire preventive, and the use of such materials should not be taken as an excuse for omitting any precautions tending to prevent dai^er of fire starting or providing ample facilities for stopping any fire as soon as it starts. It is believed that wood construction, no matter how the wood is treated either before or after erection, will involve serious fire risk, and whenever practicable noncombus- tible material such as metal, cement, etc., should be used. We, therefore, now recommend that claims as to fire-retarding properties of paints be entirely disregarded, since practically all paints have some slight value in this respect, and the difference between the best and the poorest is practically negligible. Paint Specifications A very large amount of work was done in preparing specifica- tions for paints for various purposes. For example, at the request of Emergency Fleet Corporation, specifications were prepared for white paint, outside slate-color paint, boot-topping paint, and copper paint for use on wooden ships. These were sent with a letter to the United States Shipping Board, August 3, 1917. The Bureau cooperated with the Paint Manufacturers' Associa- tion in examining numerous specifications for paint for steel ships recommended to the Shipping Board. Specifications for a paint for steel helmets were prepared for the Chief of Ordnance of the Army. After testing a number of samples of very rapid-drying paints for use on projectiles, a specification was prepared and sent with a letter to the Edgewood Arsenal at Baltimore. All projectile paints of which the Bureau has any knowledge that were used diuring the period of heavy production were of the type covered by the above-mentioned specification, but the Bureau has been unable to learn whether the proposed specification was ever issued. The War Department has, however, specifications for projectile paints of a different nature from the above, as they are much slower drying and it has not been learned whether pro- jectile paints according to the issued specifications have been used or not. Numerous paint specifications were prepared for the various branches of the War Department, the War Industries Board, etc. Paint Testing A very large number of samples of paint have been tested for the War and Navy Departments, and this work still continues. 220 WAR WORK OF THE BUREAU OP STANDARDS The Railroad Administration specifications for all paint mate- rials prescribe that methods of test approved by the Bureau shall prevail in all cases of dispute, and considerable work has been done in advising the railroad chemists regarding methods. Since the signing of the armistice the Bureau has been called upon by the Ordnance Department of the Army for advice r;elar tive to the painting of guns, shells, and other equipment for stor- age. Experiments were carried o'ut to determine the thickness of paint films on shells a,fter brushing and spraying; it was found that three coats of a suitable paint could be applied to shells without increasing the diameter sufficiently to exceed the toler- ance permitted, thereby obviaiting the necessity, of cleaning the shells should their use be required, PUBLICATIONS AND INFORMATION New Scientific Publications The Bureau of Standards prepared and published during the war many special reports, most of them confidential, on military technologic subjects. For example, a series of confidential reports on the strength and other properties of aluminum alloys; a series of special aeronautic instruments circulars, describing a wide vari- ety of aeronautic instruments, their design, construction, a:nd methods of test and calibration; and a series of power-plant re- ports, relating to the design and test of aeronautic engines and auxiliaries, such as radiators, carburetors, etc., and the relative merits of various types of airplane fuels. A series of confidential news bulletins recounting recent developments in the Bureaii's military work, besides many thousands of scientific and techno- logic papers and circulars were distributed regularly to technical ofl&cers and employees throughout the military services. Metric Literature General Order of the United States Army, dated January 2, 1918, reported that the General Staff of the American Expe- ditionary Forces in France had adopted the metric system. The order read in part: The metric system has been adopted for use in France for all firing data for artil- lery and machine gnns, in the preparation of operation orders, and in map construc- tion. Artillery and machine-gun material intended for service abroad is being graduated accordingly. Instruction in the metric system will be given to all concerned. The urgent demands for metric pamphlets, charts, paper com- parison scales, and tables of equivalents began at once and kept up briskly throughout the war and, in fact, have not yet ceas«i. PUBLICATIONS AND INFORMATION 221 The metric literature issued by the Bureau and thus distributed by the thousands of copies among the technical services of both the Army and Navy in France and America included: 1 . A graphic wall chartof the international metric system (44^^ inches by 28X inches) ; giving tinted graphic and numerical comparisons of customary ^d metric measures, and metric tables. 2 . A descriptive pamphlet of the international inetric system showing the uses of the various tmits, giving a brief summary, the important equivalents, and a statement of the legal status of the metric system. 3. A 30-cm (12-inch) comparison scale printed on paper, permitting direct visual translation from centimeters or millimeters to inches or binary fractions of the inch. 4. A document entitled "Units of Weight and Measure — Definitions and Tables of Equivalents, ' ' a rather complete statement of tables of equivalents covering all the more commonly used units, together with oflScial definitions and equivalents of basic units of length, area, capacity, and mass. 5. The Metric manual fof soldiers, which is described in the next section. Soldiers' Manual of the Metric System When the General Staff of the American Expeditionary Forces adopted the metric system of weights and measures for use of otur forces during the war, the Bureau, as mentioned above, had many demands for metric material for instruction work here and abroad. To meet this demand in a manner especially adapted for the use and information of the average soldier, the Bureau prepared a metric manual of 16 pages (10 by 15 cm). This was known as Miscellaneous Publication No. 21 of the Bureau of Standards, entitled "Metric Manual for Soldiers— The Soldier's Manual of the System — ^An International Decimal System of Weights and Measures, adopted as the legal standard by France and 33 other nations and in world-wide use. ' ' The aim was to give the American soldier a practical grasp of the system by graphic examples of the units, citing dimensions of objects familiar to the average soldier. The publication included a sketch of the origin of the metric system with brief tables of equivalents and a glossary. About 100,000 of these manuals were distributed, and the plates were made available to the bureaus of the Army or Navy desiring to use them for printing special editions. The Superintendent of Documents also placed the manual on sale. Abundant evidence has shown the usefulness of this little manual in teaching the metric system of weights and measures, which was used with such success by the allied forces in the war. Technical Information Activities On account of rapidly changing conditions during the war with its extensive use of all scientific and technologic resom-ces, the methods of combat and communication were being constantly improved or replaced through the use of new inventions or as a 222 WAR WORK Olf THE BUKBAU OP STANDARDS result of scientific discoveries. The rapid and continuous circu- lation of accurate technical information to the workers in scientific and technologic fields having military application therefore became of paramount importance. Pertinent technical information was available from authorities and experts in all branches of physics, chemistry, and engineering, from the various offices of the War and Navy Departments and from research agencies of civilian character engaged on military problems here and abroad, such as the National Research Council, National Advisory Committee for Aeronautics, etc.; as a means for disseminating this material an "information section" was inaugurated in the central office of the Bureau of Standards in which confidential information of the type just described was re- ceived. All such information was carefully digested with reference to the needs and activities of every section of the Bureau, and notices of receipt and abstracts of such information were distributed promptly to the experts concerned with the particular subject of each document. In this way a means was provided by which timely and effective contact could be accomplished among workers in related subjects of military research, and results were obtained in the way of speeding up exigent work which otherwise would have been quite impossible. The information work was further extended by the publication at frequent intervals of a series of confidential bulletins giving timely news items regarding military work carried on at the Bureau of Standards. This series was made up of 26 numbers, approximating 65 000 words. These were sent to a mailing Ust comprising the principal technical officers and employees in the Army and Navy organizations in this coimtry and in France. These bulletins give short abstracts of the method of solution and often also a r^sum^ of the results obtained on new researches and investigations, and so afforded a ready means of supplying tech- nical experts throughout all services vnth up-to-date information regarding recent advances at the Bureau in their respective specialties, aiding also to set up contacts by which experts in any field included might be enabled to obtain cooperation or assistance. RADIO COMMUNICATION Radio in Warfare The use of electric waves for telegraphy without wires had become recognized in 1917 as indispensable to warfare. It made possible commtmication with remote places and rapidly moving RADIO COMMUNICATION 223 troops. The United States was far behind other countries in the appliciation of radio communication to the above purposes, but its immediate importance was clear. The absolute necessity of radio in modem warfare is apparent when one considers the tremendous complexity of the fighting methods. The firing of artillery no longer consists merely in the pulhng of a trigger by a man who looks at his target; the use of heavy gtms depends not only on the properties of explosives and the application of ballistics, but also on aviation, radio, meteorology, map making, and many other sciences and arts. The gunner does not see the object at which he is shooting nor does he even see the airplane which is watching the mark and telegraphing the results of his shots to him. All methods of conveying signals are used, from the most primitive to the most advanced, from a human cornier to electric waves. Radio has been used not only to give orders, direct battle, and listen to the communications of the enemy, but also for very different purposes, such as the issuing of propaganda, the conduct of armistice negotiations, and the saving of ships at sea. In the several allied coimtries large labora- tories were estabhshed for the investigation of radio waves and the development of apparatus which would utilize them in the most efficient manner. Existing scientific laboratories were uti- lized for research on the more fundamental principles and instru- ments, and the military services established special laboratories of their own to design the military equipment. When this country entered the war, the Bureau of Stadnards was ready with methods, apparatus, and trained personnel for the solution of many of the fmidamental problems which confronted military men. Among the problems which had to be solved, and solved quickly, were: (i) The establishment of high-power trans- oceanic radio systems for use in case all the cables should be cut; (2) the development of low-power radio eqtiipments which should send out just enough but not too much power for commimication in the congested area of any given sector at the front; (3) a means for the location of enemy radio stations and airplanes, submarines, and ships; (4) apparatus for communication with and from sub- marines, particularly when totally submerged; (5) simple and reliable apparatus for radio telephoning; (6) the production of radio apparatus which could be easily carried and yet comprise everything necessary to make the most effective use of radio waves; (7) the training of great numbers of men in a complex and rapidly changing subject. 224 WAR WORK OP THE BUREAU OK STANDARDS Fortunately, certain radio devices existed which gave pronuse of solving ;a number of the more important problems. Among the most noteworthy of these devices are the electron tube and the direction finder. Both were being developed at the Bureau of Standat;ds. The first of these, the electron tube, is a device which makes possible radio telephony. The direction finder is a simple apparatus, which not only receives the radio waves, but which can also be turned in such a way as to determine the direc- tion from which the waves come. At the beginning of the war, although these devices were imperfect and their principles but little understood, both scientists and military engineers saw in them the promise of great utility and means for the solution of problems which would give oiu: Army great superiority over the enemy. Much has been done in the application of the best scientific knowledge to the development of these instruments. Relation of Bureau of Standards Radio Laboratory to Military Services The Bureau of Standards has been a center for radio research work. Besides its own radio laboratory, there were located at the Bureau during the war the following Government establish- ments: Naval radio research laboratory; naval aircraft radio experimental laboratory; research laboratory of Signal Corps; intelligence division of Army, radio laboratory. The Bureau furnished laboratory facilities for these independent organizations and assisted by testing instruments, etc. Accormnodations were provided for this work in the electrical building of the Bureau and in another building which was erected on the Bureau's grounds exclusively for radio work. Equipment was provided and research undertaken for the many lines of specialized measurements which were required. The presence of the Signal Corps and naval laboratories assisted the Biu-eau in planning its own radio work along lines which would assist the military services. Besides contributing directly to the military radio problems, much of the work on the Bureau's radio research program was necessarily of a fundamental nature on principles, measurements, and the efficiency of design of apparatus. Without such work the Bureau would have been unable to advise upon many of the problems submitted to it or to effectively carry out its othei- undertakings. The Bureau was able to fimiish information, advice, and assistance upon radio matters to the National Research Cotmcil, RADIO COMMUNICATION 225 as well as to the various branches of the Army and Navy Air Services. The most direct relations were maintained with the Signal Corps, a considerable part of the Bureau's work having been done specifically for this organization. This work was supported in part by fvmds transferred to the Bureau by that Corps. During the war period the staff of the Bureau's radio laboratory numbered about 40 men, and besides these, 5 men were detailed from the Signal Corps to assist in special investiga- tions. Six additional persons were detailed for short periods by the Signal Corps and the Navy to receive instruction in laboratory work. That the Bureau's cooperation was appreciated by the military services is indicated by the following statement contained in the report for 19 19 of the Chief Signal Officer to the Secretary of War: The outcome of this research work has been of vital importance to the Signal Corps, and it is fejt that every additional facility should be afforded to the Bureau of Standards so that it may continue to collaborate with the Signal Corps on these special problems. Cooperative Activities A notable event following America's entry into the war was the visit of the French Scientific Mission early in 191 7. The Mission included Prof. Henri Abraham, of the University of Paris, who has been a leader in applying electrical science to radio apparatus for war purposes. The French mission brought a large quantity of scientific military apparatus and deposited it at the Biureau of Standards. This was demonstrated to Army and Navy officers and scientific men employed upon military probleins. In 1 91 8 an interdepartmental radio conference was organized which held meetings every two weeks at the Bureau of Standards. This conference was composed of the technical radio men of the Navy, Signal Corps, and the Bureau of Standards, and at each meeting one of the important problems of radio was discussed. These meetings were very valuable, both for the interchange of information and because through this means the radio men of the Government were kept in touch with one another's work. In a number of the problems upon which the several radio organizations were working along related lines, cooperation was furthered by informal conference. Many investigations were carried on directly for the military services by the Bureau, as detailed later in this report. Besides those mentioned, many 26035°— 21 15 226 WAR WORK OF THE BUREAU OP STANDARDS Special, researches were undertaken at the request of the Army and Navy. Information and Reports to the Army and Navy A considerable part of the time of the radio laboratory was devoted to furnishing information by letter and informal con- sultation, and to aid in this work a systematic file of technical data on radio subjects was established. The subjects covered in this consultation work included principles of radio waves and apparatus, scientific and commercial data, radio organizations, instruction material, standard symbols and terminology, bibliog- raphies, applications of radio to various piurposes, and the whole range of technical subjects covered in the laboratory work. On a number of occasions demonstrations of radio apparatus and instruction in radio methods and measiu-ements were given to Signal Corps officers. General reports on the Bureau's radio work for the Signal Corps were prepared and furnished to that organization imder the dates of July 31, 1918; October 11, 1918; December 17, 1918; and July 30, 1919. A general report on the radio work for the Navy, of December 20, 191 8, was supplied to the Navy Department. Summaries of the radio work in progress, dated August 7 and November 6, 19 18, were sent to the Signal Corps, the Navy Depart- ment, Air Service, and to the other military organizations through the National Research Coimcil. Outlines of the radio work in progress and planned were sent to the Signal Corps. Numerous copies of about one himdred technical reports on the various phases of the radio laboratory work were sent to the military services. These reports varied from brief descriptions of work done to extensive treatises on radio subjects. Radio Instruction Among the problems which the Signal Corps- had to solve immediately upon entering the war was the training of great numbers of men for technical service in the field. This was a task of peculiar difficulty so far as radio is concerned, since it is a complex and rapidly changing subject. The available literature was not up to date, either for students or instructors. Instead of the few scores of trained radio operators and a handful of experienced radio engineers that were available before the war, there was need for thousands of men skilled in this subject. Intensive training was carried on at the cantonments and in colleges and schools designated by the, War Department throughout RADIO COMMUNICATION 227 the cotmtry. An acute need for suitable radio textbooks was felt by the men secured to act as instructors in these schools. The general plan of radio instruction was outlined by Signal Corps officers at a conference of college and university representatives, held at the Bureau of Standards on December 29, 1917, and, although the Signal Corps courses of radio instruction were decided upon at that time, but little instruction material was available. From the above date the Bureau of Standards' radio staff assisted in the preparation of instruction material. This was accomplished by consulting the data compiled by the Signal Corps officers and by preparing original material. The principal contribution of the Bureau was the preparation of a radio reference book and a textbook for elementary training, both of which are described below. A number of brief instruction pamphlets were prepared by the Bureau of Standards and submitted for use in connection with various parts of the Signal Co,rps course. In addition, pamphlets compiled by the Signal Corps were submitted to the Biureau for criticism and were revised to a greater or less extent by the Bureau's radio staff. "Radio Instruments and Measurements" At the December 29, 19 17, conference, the Bureau of Standards presented the nearly completed nianuscript of a treatise on radio principles, measurements, and theory. This was offered as a reference book which would be useful to the instructors in giving the radio courses and to students training for radio officers. The Signal Corps approved this manuscript and requested its speedy completion and publication. By concentrated effort the manu- script was completed and sent to the printing office in January, It was finally issued on March 23, 1918. The Signal Corps was supplied with 2000 copies. This book, "Radio Instruments and Measurements," Circular No. 74 of the Bureau of Standards, covers the ftmdamental theory of radio and the more important instruments, measiu-ements, and formulas actually used in this work, subjects not covered by any other publication. Thus there were made available unpublished results of radio work which had been carried out in the laboratories of the Biureau. One of the chief characteristics of the book is its treatment of the underlying principles of radio by the simplest methods of ordinary alternating-current theory. Previous books have presented radio as a complicated subject based upon abstruse 228 WAR WORK OP THE BUREAU OP STANDARDS theories, thus making it mysterious and dif&cult to grasp. The book, commonly referred to as "Circular 74," is a publication of 341 pages, with an index and 224 illustrations. Thousands of copies of "Circular 74" were furnished to men in the universities, the Navy, and the Army, besides those furnished to and through the Signal Corps. It has been used in the training of thousands of students. Many letters have been received by the Btireau from various sources, military, commercial, and academic, commending Circular 74, and stating that it has been found very useful in many lines of work. One criticism of Circular 74 was that it was issued in paper covers, uniform with all other Bureau of Standards publications. The demand for a cloth-bound edition was recognized by a com- mercial pubhshing company, the Wireless Press. This company reprinted Circular 74 with a smaller page size and cloth covers and has sold a great many copies. "The Principles Underlying Radio Communication" Circular 74 having been found satisfactory as a radio reference book for the use of instructors and officers, the training section of the Signal Corps requested the Bureau of Standards in April, 191 8, to prepare an elementary textbook for general use by en- listed men and others. It was desired that this book cover the fundamental principles of electricity and dynamo-electric ma- chinery, as well as radio circuits and apparatus. Such a book was needed as the basis for various pamphlets to be issued on Signal Corps radio apparatus. None of the previously available books had been found satisfactory. An understanding as to the outline of the book was reached with the Signal Corps, and the active work of preparation was begxm in June, 1918. The book was required within three months. It was prepared by a syndicate scheme of authorship which was rather an innovation in the preparation of scientific textbooks. The regular radio staff was temporarily augmented by six members of college faculties who were well equipped for this task. Due to the haste necessary in preparation and because of the collabora- tion of a number of writers, coordination of the material was par- ticularly difficult. The work was completed and turned over to the Signal Corps in September. The Signal Corps was imable to proceed with publication until December, and the book was finally issued in March, 191 9. While the book was in preparation, a first edition RADIO COMMUNICATION 229 of 50 000 was contemplated. After the signing of the armistice this was reduced, and the first edition numbered 6000 copies. The book is nonmathematical and profusely illustrated. It contains 356 pages, 4>^ by 7^ inches, m a flexible binding. It is suitable for the instruction of men who have had the equivalent of high-school education. It has been adopted as a radio text- book in the Army and the Navy, a number of colleges, Y. M. C. A., and other schools. The Bureau of Standards has received many favorable comments on this book from schools, manufacturers, inventors, amateurs, and students. Radio Direction Finder A few turns of wire wound around a wooden frame arranged so that it can be rotated constitute an aerial which not only receives radio waves but also determines the direction from which they come. Detecting apparatus connected to this simple device receives the wave in much the same way as the more familiar antenna which is seen in connection with most radio stations. The ciurent circulates around the coil instead of oscillating up and down as in an antenna. This direction finder is a much smaller structiure than an antenna, being, in fact, usually only about 4 feet by 4 feet in size. It is not as powerful a receiver of the radio waves, but its great advantage is that it determines the direction. The wave produces electrical action in the coil only when it is placed in the line of the advancing wave. If it is rotated so that it lies across the line of the wave, no effect is produced. As one turns the coil, the received signal changes from a certain maximtun loudness to a weaker and weaker and finally zero sound. From the position of the coil when the sound is thus reduced to zero, the line of direction of the wave may easily be determined. This simple apparatus was in constant use during the war to locate the position of enemy radio outfits in the trenches, on ships, in the air, and even under the ocean. Among the numerous problems presented for solution in con- nection with the direction finder was that of the accuracy of its indications. Would the direction thus determined make it pos- sible to aim a gun so as to destroy the transmitting station ? Was it sufficient to permit replacing the magnetic compass on a ship by a direction finder? (Incidentally, this use of the radio direc- tion finder in navigation has led to the name generally adopted among naval and flying men, the "radio compass.") Another important practical problem was the determination of the abso- 230 WAR WORK OF THE BUREAU OF STANDARDS lute direction of the transmitted wave. That is, could the direc- tion finder determine not only the line of tf aiismission of the wave but also from which of the two possible directions ailoilg that line the wave came? Other questions were: What was the best detecting apparatus to use with the direction finder? How did it compare with an antenna of a given size in the receiving of weak signals? Could this kind of aerial be used not only f or jrecdying' but for transmitting radio waves ? From the work done at the Bureau of Standards upon these and similar practical problems, it was found that direction could be determined to about i °. This showed that the device could b^ used to prevent the interference of radio waves. That is, it be- came possible to receive the particular wave desired and exclude others. By turning the direction finder the waves from any particular direction may be excluded, and a wave from any other direction may be received without interference. The indications of the direction finder are affected by its sur- roundings. A careful study of this subject was made, and it was found that metallic objects, such as a tower, bridge, or trolley wire cause yery large errors of direction, since ciurents produced in these structiwes by the waves affect the direction finder. (Fig. 23.) Nonmetallic objects, such as a tree or bank of earth, have smaller effects. The effects of the surroundings can be corrected for if the direction finder remains fixed relative to the surroundings. It was found that when very long radio waves are used their direction varies with time, especially at night. With the rela- tively short waves used on ships and in military operations, how- ever, the variations are negligible. One of the applications of the direction-finder to navigation developed by the Biu-eau is its use in fog signaling. (Fig. 24.) The lighthouse lamp, made useless by a heavy fog, can be supple- mented by a radio system, consisting of automatic apparatus for transmitting radio waves as regular signals at definite intervals. A ship equipped with a very simple radio direction finder can easily steer toward the lightliouse or determine its location. Coil Aerial as Transmitting and Receiving Device The simple qoil aerial used as a direction finder was carefully investigated in comparison with the ordinary large antenna as a radiator and receiver of radio waves. It is in general not as power- ful as the large antenna, but when a sensitive amplifier is used with the receiving apparatus the coil aerial is sufficiently powerful RADIO COMMUNICATION 23 1 for many classes of radio communication. The coil is used in series with a condenser, and to this simplest type of radio circuit is connected the generating or detecting apparatus, no connection to the ground being necessary. On account of the small size of the coil aerial, the complete radio outfit may be contained and used in an ordinary room. The coil aerial has advantages over the ordinary form of antenna for certain applications, as in the Submarine work described in a subsequent paragraph. A study of the principles of action of the coil aerial and the ordinary antenna led to the development of transmission formulas by which it is possible to calculate the current produced in the receiving apparatus for any distance from the transmitting station. The formulas have been verified by experiment. Laboratory studies were made of the design and electrical properties of coil aerials, thus leading to a knowledge of the best way to construct such an aerial for any particular purpose. In field tests with the coil aerial it was fotmd to have directional properties as a transmitting as well as a receiving device. That is, it transmits more intense waves parallel to its plane than at 90° to it. This valuable characteristic, together with the facts that more current can be gotten into it and that it can emit a narrower band of wave lengths than the ordinary elevated antenna, makes it valuable for certain applications, but its relative lack of power precludes its general use as a transmitting aerial. When used as a receiving device, it will pick up waves coming from a source a surprising distance away. By the use of exceedingly sensitive a,mplifiers, with an aerial consisting merely of a few turns of wire located inside an ordinary room, it is easy to receive in Washing- ton signals txansmitted from Germany or from the Pacific coast of the. United Sta.tes. Submarine Radio Signaling It was generally believed that radio communication could not be carried on below the surface of the ocean, since the water was thought to be so good a conductor as to be impenetrable by the radio waves. In the navies of this and other countries experi- ments had failed to develop a successful radio system for use under water. Nevertheless, the Bureau investigated the coil aerial as a possible solution of this problem. In November, 1917, the Bureau of Standards placed a direction finder coil under water and received signals virtually as good as with the coil in the air. Additional experiments with the use of 232 WAR WORK OF THE BUREAU OF STANDARDS the apparatus on a submerged submarine soon followed. Permis- sion was secured from the Navy Department to make experiments at the submarine base, and after experimental work on sub- marines had been conducted for several months very simple but effective radio apparatus was developed. The essential part of the apparatus is the antenna, of the coil or loop t3^e, consisting of a single turn of wire, heavily insulated, its circuit being completed by grounding to the metal hull of the submarine. On June 20, 191 8, with every part of the submarine and its equipment submerged, signals transmitted from Germany were recdved in the submarine. Signals from Paris, Rome, and California were received with equal ease. Furthermore, the appa- ratus was found to be a satisfactory direction finder just as when used in air. It was later found possible not only to receive but to transmit radio messages when the submarine was totally submerged. Transmission can not be carried on over very great distances, but it is possible for a ship and a submerged submarine to exchange recognition signals. This is the only thoroughly successful system that has been developed for communicating with submerged submarines, and the Navy has equipped a number of submarines with it, as shown in Fig. 25. Signaling System for Airplane Landing The radio direction finder is very useful to the aviator. The magnetic compass and other instruments to assist him in steering his course are affected by the rapid changes of motion of the air- plane and do not give him reliable information about his location with respect to a known point on the ground. By using the radio compass, he can steer toward his landing field, but a need was felt for a device which would supplement the radio compass by telling the pilot when he was exactly over the landing field even in fog or darkness. The need for this had been felt particularly in the Air Mail Service of the Post Office Department. A localized signaling system which fulfills the above requirement has been worked out as a result of the Bureau's experiments. Either of two methods may be used. In one, a large coil of wire consisting of one or a very few turns with a length of about 1000 feet on a side is laid on the landing field. This coil carries several amperes of alternating current of about 500 cycles per second. Magnetic induction from this coil extends over the space above the landing field, and produces a signal in telephone receivers connected Miscellaneous Publications, Bureau of Standards, No. 46 Ml ;^Sij r ^ ift:':iio^ 3^ Fig. 25. — United States suhmarine equipped with coil antenna By this means the submarine can send and receive radio messages even when totally submerged. The system of submarine radio transmission developed at the Bureau has been applied to several vessels by the Navy Department Miscellaneous Publications, Bureau of Standards, No. 46 A Fig. 26. — Nwmerotcs forms of electron tubes This Httle device has revolutionized radio communication, has made possible radio telephony, and has aided in securing satisfactory multiplex wire telephony where as many as five cunnections are possible over one pair of wires RADIO COMMUNICATION 233 to a coil of wire on the airplane. The aviator hears the signal only when he is over the landing field. In the second method the result is produced by the use of specially constructed coils of much smaller size on the landing field, employing radio-frequency current. Satisfactory signals are heard by the aviator at distances up to a mile above the field. The advantage of this system is not alone the actual utility to the aviator, but also the added sense of security against his enemy, fog. Electron Tubes The use of electron tubes was practically unknown to the mili- tary forces of the United States prior to 1917. All of the radio- apparatus then used was of the damped-wave type. The develop- ment of military radio equipment employing electron tubes was given the needed impetus by the visit of the French Scientific Mission in the spring of 191 7, since they brought with them, and exhibited at the JBureau of Standards, a great variety of equip- ments for radio communication which had been developed around the electron tube as the essential element. The three-electrode electron tube is a simple device, known by a great Ihany names, such as vacuum tube, audion, valve, lamp, ther- mionic bulb, triode, valve relay, etc. It looks much like an ordinary iiicandescent lamp bulb (Fig. 26) , but has in addition to the hot filament two other metal structures, a simple wire grid and a metal plate or cylinder. The operation of this device depends upon the fact that every hot object is continuously giving off small particles of electricity, called electrons. Some of the electrons emitted from the hot filament go to the metal plate, and this flow of elec- trons constitutes an electric current. This electric current can be influenced or controlled by the flow of electrons to the metal grid. This control of one electric current by another very small electric current makes it possible to use the tube as an electric relay or amplifier and for many other piuposes. A simple descrip- tion of the fimdamental theoiy of operation of the electron tube is given in Circular No. 74 of the Bureau of Standards, " Radio Instruments and Measurements," pages 200 to 215. This device was invented less than 10 years ago, and most of its development has been accomplished during the war. It is a remarkable instru- ment if only the variety of its uses is considered. Thus it serves as the detector of radio waves, as a very powerful amplifier of radio or any other electrical ctirrents, as a generator or producer 234 WAR WORK OF THE BUREAU OF STANDARDS of radio waves, and as the means for converting speecll into a modulated radio wave which can be again changed into speech by a receiving radibapparatus. The reader will readily see that these important applications justify the most extensive and pro- found research, development, and application. Thus the principal work of the great New Jersey radio laboratories of the Signal Corps was the development of apparatus using these eleclron tubes. Certain research work and the standardizing of tubes and methods of testing were carried on by the Bureau of Standards. The principles of the operating and functioning of the tubes are, by contrast with the structure of the tubes themselves, complicated and difficult to determine. When one considers the extensive applications, of these devices which have already been made, it seems strange that so little is known regarding the principles of their operation. The importance of electron tubes in military work is apparent from the fact that for the American Armies alone 25 000 tubes were made each week. These devices have revolutionized all branches of radio and have produced many advantages for which investigators have sought in vain for years. Electron tubes have many applications outside of radio, as, for instance, their applica- tion to wire telephony, where they make possible five simultaneous convereations over one pair of wires. On account of their great sensitiveness as receiving devices, radio apparatus can be made very compact. These devices must be credited with a considerable share in the achievement mentioned above in describing direction finders, namely, the receiving of messages from a distance of thou- sands of miles with a small apparatus contained in an ordinary room. In fact, apparatus cS,pable of concealment about a person's clothing may now be constructed by which one can receive the radio messages which are passing through space. The electrqii tube has made cornmunication between airplanes successful. Obviously, airplane apparatus must be extremely light in weight, and this is possible ijith the very sensitive electron, tubes. Airplane pilots can now talk to one another, using appa- ratus that adds only a few povmds to the weight of the machine. Not only can the communication be carried on from one airplane to another and from an airplane to a ground radio set, but the apparatus can be connected to the ordinary telephone, lines. The radio telephone upon which the above apparatus is based was brought by the French Scientific Mission to this country in RADIO COMMUNICATION 235 191 7. Improvements in many points have been made since that time and have rendered the apparatus more reliable and effective. On account of the Bureau of Standards' work on the charac- teristics of tubes, amplifiers, and other electron-tube apparatus, and on methods of measuring their constants and characteristics, the Signal Corps requested the Bureau to take over the general work of electron-tube standardization. The work included the meastirement of numerous characteristics of tubes in order to check results obtained by field inspectors, a study of experimental tubes under development by the various companies, and of the methods of testing tubes in the field and in the laboratory, a determination of the causes of behavior of power tubes in various circuits, and a study of various amplifier circuits. The Signal Corps was assisted in its inspection work by instructing a number of electron-tube inspectors who were detailed to the Bureau of Standards for that purpose and were given instructions for a period of six weeks in the methods of testing tubes. Special Military Researches on Electron Tubes A number of rather extensive researches were carried on fol- lowing a definite program laid out by the Bureau after receiving a specific statement of the problem from the Signal Corps. One of these was on the efficiency of electron-tube generators of high- frequency current as affected by the production of harmonics. Investigation had shown that practically all the generating cir- cuits used in Signal Corps apparatus were heavily loaded with harmonics, so that the output hot-wire ammeter reading gave no indication of the useful power output which is all contained in the fundamental frequency. This was investigated both by measure- ments of current in tuned circuits and by direct observation of the wave forms in the cathode-ray oscillograph. A complete solution of the problem was obtained. The principles were developed by which the amount and effects of the harmonics may be determined, and by which apparatus can be so designed as to reduce them. A critical study was made of the Signal Corps testing sets for receiving tubes. These sets had served well as factory testing sets, but it was recommended that the Biireau make a critical study of them, looking toward improvements in design, practice, or method, and having as particular objects increased uniformity of the inspected product and decreased difficulty of manufacture. It was found that a number of improvements could readily be 2^6 WAR WORK OF THE BUREAU OF STANDARDS made in the amplification and in the detector test sets. It was shown that the generator test set did not make a real test of the generating properties of a tube, as it would be difficult to find a tube which would not oscillate in this set. Furthatnore, it was easy for the operator to connect up the set so as to change its indications radically. The following suggestions were made: (i) A circuit which really tests the generating properties and which is similair to actual receiving circuits should be used, and (2) a dif- ferent method of indicating the presence of generated current should be employed. In using receiving tubes as high-frequency amplifiers the Signal Corps had noticed that the input impedance of the tube, particu- larly that part contributed by the capacity of the grid to the other electrodes, seemed to be a function of the electrical constants of the plate circuit. A furthier knowledge of this was very desirable, and the Signal Corps recommended that the Bureau make an experimental study of the subject, using in particular the VT-i tube. A complete solution of the problem was attained by a theoretical; investigation, checked experimentally. It was found that the input impedance of a tube can be represented as a series resistance and capacity, the values of which depend in a definite and determinable manner upon the electrical characteristics of the plate circuit and the capacities between the tube electrodes. At the request of the Bureau of Steam Engineering, Navy Department, a study was made of the use of high-power three- electrode tubes for producing power at radio frequencies, par- ticularly for radio telephony. For supplying the direct-cuirent power to the tubes without use of commercial high^voltage direct- ciurent generator^, a rectification system was developed, using three-electrode tubes in place of the two-electrode tubes usually used as rectifiers. It was found that the third electrode, or grid, Gould be used either for smopthiug out the pulsation in the output current or for increasing the efficiency of rectification, but not for both at the same time. Accordingly, attention was directed especially to the use of the grid control for increasing the efficiency, while the desired smoothing out of pulsation was obtained by other methods. For the latter purpose the output load was shunted with a condenser and inductance in series, tuned to the principal frequency to be eliminated. For converting the direct^ current power thus obtained into radio-frequency power, a study was made of the use of large power tubes as radio-frequency am- plifiers. In this system the tube or set of tubes operates into a RADIO COMMUNICATION 237 high-power resonant ctfcuit or antenna, while the grid alternating voltage is supplied from an independent low-power resonant cir- cuit which is excited by a low-power electron-tube radio tele- phone set. This radio-frequency power amplifier was found to possess many advantages, especially in simplifying the design of the high-power output circuit, and in placing most of the com- plicated apparatus in the auxiliary low-power radio telephone set. Radio Measurements and Tests An interesting portion of the work consisted in development along the lines of technical measurement. The basic facts, the knowledge of which has made possible the improvements above described, could never have been discovered without careful measurements of the phenomena which take place in connection with the use of high-frequency radio currents. These measure- ments are far more complicated than most of the ones with which we are familiar. In this matter of measurements the Bureau naturally is the recognized leader in the country, and radio meas- urements in particular have been very highly developed by its laboratories diu-ing the last few years. The beginning of the work for the military forces was contempo- raneous with improvements in radio measurements made possible by the use of the electron tube. Tubes were put into service in the laboratory as a source of ciurent for meastirement ptuposes, particularly as generators of very steady high-frequency currents, which have made possible measiurements to an acciuracy of i or 2 per cent, whereas 5 or 10 per cent has been the previous limit. This greater accuracy of measurement made it possible to develop more accurate radio instruments, and thus called for the devel- opment of special methods of measuring, special means for avoid- ing small errors in the measiurements, and special measxuring instruments. Standards for the important measurements were established. The theory of radio measurements was studied extensively in the Bureau's laboratory, and the results of this work were made available to the Army, Navy, and manufacturers through special laboratory reports and printed publications. A great deal of the measurement work was concerned with the fundamental components of all radio circuits, capacity, induc- tance, and resistance. A typical instrument set-up for this work is shown in Fig. 27. The elimination of errors in the laboratory by carefully devised shields and other precautions was necessary. Apparatus for the production and measurement of large current 238 WAR WORK OF THE BUREAU OF STANDARDS and high voltage at radio frequencies was developed. (Fig. 28.) For the measurement of small current, thermoelements were used, while for the very small currents, such as those in a receiving antenna, electron tubes were employed. Special mention should be made of the employment of these tubes in a special method of comparing the received with a locally produced signal. , Numerous types of radio apparatus were designed, and this work made necessary an extensive study of the inductance coil, one of the. principal ele- ments of every radio circuit. Electrically, a coil is a complicated apparatus, and there has been no comprehensive information on its behavior in a radio circuit. A number of tests of radio operating apparatus, such as trans- mitting sets and antennas, were made. A particularly interesting one consisted in a complete test of a Signal Corps airplane radio- transmitting set known as type SCR-73. It was modeled on a French set, and comprised a self-excited, 900-cycle generator, wind-driven by a small propeller, and a radio set of the spark type. Provision was made for rapid change to any of 16 wave lengths or to any of 6 signal pitches, for the purpose of reducing interference with other transmitting sets. The tests included characteristic curves and oscillograms of voltage and current from the generator power output and tests of the self-excitation and other feattxres of the generator, study of capacity to produce best spark tone and maximum quenching action, nattu-e of wave radiated, and general criticism of design. Among other interesting tests was that of a novel type of quenched-gap transmitting set. The operation of special features of thp set was clearly revealed by the use of the cathode-ray oscillograph. The determination of the constants and behavior of a submerged cable used as a receiving antenna formed another investigation. Quite a number of tests were conducted and studies made of the design of wave meters used as radio measuring instruments. Detailed comparative studies were made of the design of commercial. Signal Corps, and French wave meters, and the Bureau's experience in measurements and design enabled it to suggest improvements. One of the tests of wave meters in- tended for field or air use was a shaking test, in which was deter- mined the effect on the instrument caused by placing it in appa- ratus which simulated the vibrations of an airplane. The routine tests frdln May i, 1917, to June 30, 1919, included 467 electron tubes, 279 decrimeters, 78 wave meters, 46 conden- sers, 39 inductors, 13 ammeters. Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 27. — Apparatus for measuring the properties of inductance coils used in radio transmitting and receiving sets The instruments are inclosed within a cape of galvanized wire screening in order to avoid inter- ference by waves from near-by transmitting or generating sets Fig. 28. — Apparatus for generating high voltages at radio frequencies used in the study of the radio properties of insulating materials for radio apparatus Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 2g. — Apparatus for testing self-luminous dials By means of this equipraent it is possible to determine whether a sample possesses the requisite brightness RADIO COMMUNICATION 239 Cathode-Ray Oscillograph There is no mechanical instrument which can follow the exceed- ingly rapid alternations of the current used in radio circuits. The only way this can be done is by the use of a stream of electrical particles, as in the cathode-ray oscillograph. In this instrument a beam of cathode rays is deflected by the currents, and the motion of the beam is made visible on a fluorescent screen. Thus the wave form of radio current can be seen, and the phase relations of currents and voltages in radio circuits can be directly determined, . Two general types of cathode-ray tube have been used at the Bureau of Standards, those using a high voltage to cause emission of a stream of electrons from the plane cathode of metal and those having a cathode consisting of a heated filament from which the electron emission is spontaneous. Tubes of the former type are suitable for the measurement of currents of the order of magnitude of I ampere. The hot-cathode tube is suitable for measiiring the electron currents obtained in three-electrode electron tubes, which are of the order of magnitude of 0.05 ampere. Oscillograms were made with these tubes as a part of the tests performed upon radio transmitters submitted to the Bureau of Standards. The quenching action of different spark gaps is determined direclJy by passing the gap cmxents through suitable deflecting coils placed around the tube. A time axis perpendic- ular to the ciurrent deflection thus produced is provided by con- necting the electrostatic deflecting plates across the capacity in the closed oscillatory circuit. The cathode-ray oscillograph was also used in the investigation of the harmonics produced in antenna circuits by electron-tube generators. D5mamic characteristic curves are obtained for a given generator by impressing the grid voltage upon the deflecting plates and allowing the electron currents to flow through the de- flecting coils. The wave forms of the antenna cturents are deter- mined by providing a sinusoidal time axis from a separate generator timed to a lower frequency, of which the frequency of the antenna current is a suitable multiple. The combination of an electron- tube generator, which is capable of furnishing currents at any frequency up to that corresponding to a wave length of a few centimeters, with a cathode-ray tube having a slow-moving electron stream, whose deflections are closely proportional to the instantaneous values of these currents, offered an opportunity for much profitable research. By the use of one of these tubes a study has been made of the charac- teristics of electron tuhes under actual operating conditions. 240 WAR WORK OP THE BUREAU OP STANDARDS Insulating Materials The work which was in progress at the Bureau of Standards on instilating materials during the war included the measurement of all the properties which are of importance in determining the value of the materials for radio apparatus. Both in regard to the number of properties thus investigated and the number of materials, the investigation was a comprehensive one. The work was mainly on the class of materials most widely used in radio equipment, known as the phenol or bakeUte type. Much of the work of measurement was carried out by other sections of the Bureau, the only meas- tirements made in the radio laboratory being those involving the electrical properties at radio frequencies. While a wide variety of electrical and mechanical properties of the materials were meas- ured, the object principally in mind was the determination of the suitability of the materials for radio uses. The work involved in this investigation can not be appreciated without an understanding of the natiure of the materials. They all contain a phenol resin or varnish of the general t)rpe invented by Dr. Bakeland and may be divided into two general classes, the laminated and the moulded ma:terials. It was fortimately possible to make the meastu-ements of elec- trical, mechanical, and thermal properties on the actual sheets of material as regularly supplied for commercial use. Conseqiiently, the data obtained should be directly applicable, as it was not necessary to manufacture samples especially for test purposes. Thus, for instance, for the radio measin-emerits condensers were made up using an actual slab of material such as is employed in the construction of apparatus; that is, regular stock as furnished by the manufacturer. This is a far better way to conduct a test than by making up a special condenser from specially prepared thin sheets of the material. Mechanical properties which were measured included tensile strength, transverse strength, modulus of elasticity, permanent warping, brittleness, hardness, density, and moistine absorption. The thermal tests included thermal expansivity, a little work on the effect of temperature on hardness, and some work on the effect of temperature on the disintegration of the material by carbon- ization, etc. The electrical properties included volume resistivity, surface resistivity, power factor, dielectric constant, and the effects of high voltage at radio frequencies. Investigations have been made of the effects of voltage at radio frequencies upon insulating materials. This work has required the RADIO COMMUNICATION 241 development of entirely new apparatus for the production of con- stant high-frequency voltages, for their application to insulating specimens, and for their measurement. Essentially, the method consists in placing a sample of material in a radio circmt with electrodes upon its surface in parallel with a condenser and meas- uring the voltage required to produce certain effects, such as the appearance of corona, flash over, and also pimcture of the material. The flash over and puncttire voltages are of very different magni- tude at high frequencies as compared with other values at low frequency for the following reasons: Very much lower voltages produce these effects at radio frequencies than at low frequencies, because the dielectric carries a considerable dielectric current. This current heats the specimen by virtue of the absorption phe- nomenon or dielectric loss in the material and soon raises the tem- perature to a point where breakdown occurs. The effect is then not a punctmre or rupture of the low-frequency type at all, so that instead of himdreds of thousands of volts being required to break down or flash over a specimen, 10 000 volts more commonly suffices. Measurements of the various properties which have been enu- merated were made upon several hundred samples of material representing various grades and thicknesses of the standard products. These were mainly laminated materials, but some of the work has been upon the molded phenolic materials. In general, the more expensive grades are superior in their electrical properties. The reverse is true in regard to mechanical strength. For example, the cheaper materials have higher tensile strength than the more expensive. The expense of the materials is usually determined by the percentage of phenol varnish to paper, the more expensive materials having a larger percentage of varnish. The thermal properties do not in general fxumish any means of differentiating between the several materials. Above 50° C the variation of dimensions with temperature is not reversible, and many peculiar expansion and retraction effects have been noticed upon raising and lowering the temperature. All the phenol materials have smaller coefficients of expansion than hard rubber. The insulating materials of the phenol type came into use largely because a substitute for hard rubber was wanted, the value of the latter material for electrical apparatus arising; from the fact that it is easily machined, has very small power loss or phase difference, and very high puncture voltage at radio frequencies. It has, however, certain other objectionable features; thus, it shrinks, 26035"— 21 16 242 WAR WORK OF THE BtTREAU OF STANDARDS warps, is brittle, deteriorates in sunlight, and has high thermal expansivity. An interesting result which was observed in the power-factor measurements was an apparent increase of power factor in a particular sample of over 100 pet cent after a lapse of time of six months. Upon investigation this change was found, to be an actual change, and by a baking pirocess the phase difference was brought down again to . approximately the value of six months before. This and similar experiments on other materials sug- gested that the absorption of moisture from the atmosphere changes the properties of these insulators. Permanent Crj^tal Detectors In March, 191 8, the Signal Corps submitted a suggestion for a crystal detector having' a permanent contact and possessing high sensitivity. The plan was to secure on the surface of the crystal, by alternating-current electrolysis, needlelike deposits, perhaps one-half mm high, and making a permanent soldered contact to a number of these needles. It was thought that the additional contacts would give increased sensitivity, besides possessing the advantage of a permanent contact not disturbed by ordinary mechanical jais. This was desirable for field work, as a crystal detector was needed having a permanent contact which could be used in the field without the necessity of continually resetting the contact wire on a sensitive spot of the crystal. Experimental work was carried out along the suggested lines by the Bureau. It was found that deposits could be secured by electrolysis using a 6o-cycle alternating current of the order of 50 milliamperes, and that in some cases deposits secured on car- borundum were sensitive. It was also found that no advantage could be practically attained by using a number of contacts in parallel, and that, in general, the signal obtained using two con- tacts in parallel was less than that obtained using alone the most sensitive one of the two contacts. Difficulty was met in attetnpt- ing to secure sensitive deposits on galena, because most electro- l3rtes seriously affected the sensitiveness of the galena as a de- tector. Since galena is one of the most sensitive crystal detectors, it was very desirable that a permanent-contact galena detector should be obtained, if possible. Later attempts were made to secure, electrol3^cally, deposits on a number of different crystals, using a radio-frequency current of 1 50 to 200 inilliamperes. Some sensitive deposits were secured in this way, but they were not satisfactory, as they were not RADIO COMMUNICATION 243 adherent. In the work at radio frequency it was shown that the sensitiveness of a deposit is dependent on the failure to obtain a perfect contact with the crystal surface, and that really adherent deposits are practically never sensitive. Work with direct-current electrolysis also served to confirm the conclusion that sensitive- ness of a metalUc deposit depends on failure to make perfect contact with the crystal surface. As a possible means of securing permanent contact, soldering was attempted, but without success, except in the case of molyb- denite. No flux was found which would cause the solder to adhere either to galena or silicon. Both arc and spark welding were tried. Wires as fine as No. 40 were welded to galena, silicon, iron p5Tites, and molybdenite by these methods, but in no instance was the welded contact sensitive. Since in welding the surface was melted and the surface immediately surrounding the weld, which had been melted in the process, was found to be insensitive, this method was not successful. A fairly satisfactory detector was constructed by making a contact between a slightly hollowed surface of Wood's metal and the sharp point or edge of a galena crystal, the whole being inclosed in paraffin or sealing wax. The chief difficulty encovm- tered was due to alternation in the contact caused by expansion of the wax with change of temperature. Some experimental detectors so made remained fairly sensitive for several weeks, but after a few months were usually found to be much less sensitive. Two general conclusions were drawn as a result of these investi- gations: With the less sensitive crystals, such as carborundum and molybdenite, plated-contact detectors of fair sensitiveness can be made, but this is no practical gain, since these crystals have previously been in use with fixed contacts. With the more sensitive crystals, such as galena and silicon, the only method which gives promise of success is that of a mechanically maintained high-resistance contact. Peace-Time Value of Work The radio work carried out during the war presents a con- spicuous example of scientific advancement of permanent value. It has been estimated that in two years of war the progress in radio was equal to that in 10 ordinary years. The work of the Bureau contributed to the progress made upon the electron tube, the direction finder, control of radio waves, radio measurements and design, submarine signaling, airplane communication, and radio instruction. Each of these, with the possible exception of 244 "WAR WORK OP THE BUREAU OP STANDARDS submarine signaling, has positive value in the normal pursuits of peade. As a result of the research work during the war, the electron tube was developfed from an unstable, experimental device, used largely by amateur radio enthusiasts, to a dependable instrument of standardized operating characteristics, produced on a com- mercial scale and used in large quantities. Besides increasing greatly the effectiveness of radio-receiving apparatus, it is used also as a generator of alternating current of any frequency from i to loo coo GOO. Reliable long-distance wire telephony has been made possible by the electron tube; through its aid many tele- phone conversations are carried on over one pair of wires. It is used in the construction of amplifiers which magnify very feeble currents or signals hundreds of times. In consequence, radio antennas can be reduced to very small dimensions and the range of radio communication greatly extended. Music, as well as the voice, is readily transmitted without wires, even across the ocean. A symphony concert can readily be given at one point and be sent forth by radio so that it can be received anywhere in the United States. The direction finder has been developed into a very convenient and portable apparatus. It replaces the antenna for receiving radio waves and determines the direction of the transmitting sta- tion. It is useful in determining the location of radio stations, as in the work of Government inspectors when violators of the radio laws are sought. It is of the greatest value as a safety device on either airplanes or ships in fog or other perilous circumstances. Following the end of the war the Bureau of Lighthouses and the Biu-eau of Standards took up actively the development' of a fog- signaling system involving the use of the direction finder. Thus aided, a ship can steer toward a lighthouse or determine its posi- tion during the heaviest fog. The improvements in design of radio apparatus and the advance in methods of radio measurement have increased the efficiency of instruments and equipment. Progress in the unda-standing and control of radio waves has increased the certainty with which the range of communication of any radio set can be calculated. The textbooks and other material prepared for the training of Signal Corps men in radio have been found useful in universities and other institutions. One of the incontestable benefits wliich has been salva,ged from the war is the application of radio to airplane communication. In RADIOMETRY 245 the future of aerial navigation, as in policing the air, radio will play an important part. The electron tube has made possible conver- sation between airplanes and the ground, the direction finder is the compass of the aviator, and radio methods supply the signals by which a landing may be made in fog or darkness. The ordinary wire telephone can now be connected to a radio system, so that conversation between the ordinary city telephone and an airplane is possible. Radio supplements the older systems of communi- cation, and no spot on the earth or in the air is too remote for it to reach. RADIOMETRY Researches The following paragraphs summarize the most important in- vestigations in Radiometry which had a direct military apphca- tion: Radiant Power Life Tests of Quartz-Mercury Vapor Lamps This subject is treated imder the heading "Aircraft Materials." Optical Properties of Balloon Fabrics This subject is treated under the heading "Aircraft Materials." Photoelectric Sensitivity of Molybdenite and Various Other Sub- stances This subject is treated under the heading " Invisible Signaling." Amplification of Bolometer Current for Signaling Purposes This subject is treated under the heading " Invisible Signaling." Infra-red Transmission Spectra of Various Substances The spectral transmission of various substances, including spe- cial glasses, colored fluorite, salts in solution, etc., was determined. The data were published in Scientific Paper No. 325. Some of the substances examined (including material submitted for official tests) were foimd useful in a device used for secret sig- naling by means of infra-red rays. As a result of this investiga- tion, a prominent glass factory improved the infra-red transmis- sion of a red glass by eliminating the iron impurities which cause an absorption band at i^u. Protection of Moving-Picture Film from Heat of Lamp In certain military work it was desired to stop the raovement of the picture film in order to make an examination of the details of a single picture projected upon a screen. As ordinarily used, the moving-picture film is protected from injury by the intense infra-red rays from the lamp because it is 246 WAR WORK OP THE BUREAU OP STANDARDS exposed to these rays for only a very short period of time. Apply- ing the information obtained in the foregoing and in a previous investigation, the intensity of the infra-red rays was greatly reduced by using a dilute (i to 2 per cent) solution of cupric chlo- ride, which absorbs but little of the visible spectrum, while being practically opaque to infra-red rays. Glasses for Protecting the Eyes from Injurious Radiations Information on the protective properties of glasses used in oxyacetylene cutting and welding is very important. That this Bureau's efforts in obtaining such data are appreciated by manu- facturers is attested by the fact that the first issue of its techno- logic paper on this subject was exhausted within three months after publication, one shipbuilding firm calling for almost a hundred copies. In addition to research work on this subject, quite a number of ofiicial tests were made on the eye-protective properties of glasses submitted by dealers and by the construction departments of the navy yards at Washington and Philadelphia. During the past year extensive new data were obtained on the ultra-violet and infra-red radiations transmitted by various glasses used for spectacles. The sources of radiation used were typical of those to which one is exposed in occupational pursuits. For example, the quartz-mercury vapor lamp and the iron (magnesite) arc used in these tests were representative of the soiu-ces of ex- tremely intense ultra-violet radiation, while the gas-filled tungsten lamp and the sun represented high intensities in the visible and infra-red spectrum. These new data were published in a revised and enlarged (the third) edition of Technologic Paper No. 93, "Glasses for Protecting the Eyes from Injurious Radiations." Reflecting Power of Stellite and Magnalium Stellite is an alloy of chromium, cobalt, and molydbenum. It takes a high polish and appears to be quite permanent in ait. It reflects from 60 to 65 per cent of the visible rays. (See Scientific Paper No. 308.) From a military standpoint the use of stellite was under consideration for reflecting mirrors in periscopes and range finders. In response to a request from the Chief Signal Oflicer, War Department, for information on the reflecting power of magnalium (aluminum-magnesium alloy) mirrors, data were obtained on samples prepared by the division of metallvurgy. RADIOMETRY 247 It was found that the reflecting power of the best samples of magnalium increases uniformly from about 85 per cent in the blue to 90 per cent in the red. The average sample would probably reflect 85 per cent depending upon the polish. Instruments and Methods of Radiometry While the investigations of thermopiles, photoelectric cells, etc., were not undertaken with a, view to direct military applica- tion, it is known that the method of construction and general technique developed in this Bureau were useful to others who had military problems tmder investigation. For exemple, on the basis of Scientific Papers Nos. 188, 229, 282, and 319, the ther- mopiles were constructed for the purpose of detecting the presence of airplanes, etc., at night by the heat emitted by the engines. Apparatus was sent to France for testing out this device. Thermopiles made by the Bureau were lent for measuring the intensity of searchlights. One thermopile was lent to the United States Naval Experimental Station to determine its usefulness in detecting ships at night by the heat emitted from the smokestacks. Reports of the official tests on this device made by the National Research Council were very favorable, and it was suggested that further work be done. Lacquers for Protecting Tamishable Metallic Surfaces Owing to the scarcity of optical glass it was desirable to find suitable metal mirrors as a substitute for lenses used in range finder?, periscopes, airplane cameras, searchlights, etc. Silver- on-glass mirrors are easily prepared, but a lacquer covering is required to prevent the silver from tarnishing. From the experience gained in this laboratory, it appears that xmlacquered mirrors which are inclosed to the extent afforded in a camera will not tarnish appreciably in one to two years. This tarnish is quite soluble and easily removed by dipping the mirror in water or by polishing it with soft chamois skin and opticians' rouge. Tests were made of water-white silver lacquers, such as "Alba- line," "Zapon," etc. They were diluted, filtered, and poured upon the freshly polished silver. The excess material was allowed to drain off and then the mirror placed in a level position to dry. Theoretically, a lacquered-silver surface should reflect almost as raiich light as does an unlacquered surface. In practice, however, the reflecting power was found to vary greatly with the condition of the lacquered surface. A newly lacquered silver 248 WAR WORK Olf THE BXmEAU O^ STANDARDS mirror reflects from 70 to 75 per cent of the incident yellow light, while in the red it reflects from 80 to 85 per cent, depending upon the homogeneity of the lacquer. Uncoated or bare silver reflects from 90 to 95 per cent of these rays. In the near infra-red, where nitrocellulose is free from absorp- tion bands, the reflecting power increased gradually to, about 95 per cent. For reflecting heat rays of long wave lengths, which are absorbed by the lacquer, mirrors of nickel or gold (which reflect 95 per cent) should be used. At the request of the General Engineer Depot of the War Department for information on lacquers suitable for protecting, nietal searchlight roirrors an examination was made of the effect of ultra-violet light upon, the various lacquers commercially available. It was found that the photochemical action of ultra-violet light such as obtains in a searchlight Uberates sulphides and nitrous oxides which may be present as impurities or which are constituents of the lacquer and which then attack the silver. As a result the silver imder the lacquer, following exposure to a carbon arc or a quartz-merctiry vapor lamp, is quickly turned brown in color. For example, after 6 hours' exposure at a dis- tance of 0.5 from a small (15-ampere) carbon arc a lacquered silver-on-glass mirror had cha,nged to a brownish color, which strongly absorbed the yellow, green, and violet lights. As a result the reflecting power of the lacquered mirror for wave- length X=o.55M (greenish yellow) was decreased from 74 per cent to 68 per cent. Similarly, at X = 0.65^1 (orange-red color) the reflecting power was decreased from 82 to 79 per cent, and at wave length X=o.7ju (red light) the reflectivity was decreased from 85 per cent to 83 per cent. In a powerful searchlight such a lacquered silver mirror would be injured in a few hours. A bare nickel-plated glass mirror would be just as efficient. A gold mirror would be more satisfactory. Specifications and Methods of Test In connection with the preparation of standards for head-and- eye protection which is being carried out by the Bm-eau in con- jimction with the safety engineers of Federal industrial estab- lishments considerable experimental work was done on the pro- tective properties of glasses already in use in the navy yards.: This was done for the purpose of obtaining a glass of suitable RADIOMETRY 249 transparency for oxyacetylene cutting and welding, and at the same time of providing ample protection against injurious radia- tions. Specifications were then drawn defining the maximum transmission of injurious radiations permissible, especially for infra-red rays. Sources of radiation which are readily obtained by manufacturers were also investigated. A gas-filled tungsten lamp was found suitable for testing the protective properties of glasses for infra-red or so-called heat rays. For the General Engineer Depot tests were made and specifi- cations written for glasses for soldiers and engineers to provide protection against glare from snow, etc. Conferences, Consultations, Etc. Under this caption there are enumerated some of the military problems presented by Government btu-eaus and manufactm-ers. The subject most frequently presented related to thermal-radio- dynamic devices for secret signaling, and for detecting submarines or battleships in a fog by the heat radiated from the smokestacks, etc. Many of the inquirers did not seem to realize that for radia- tions of wave lengths such as would be emitted by a smokestack water vapor is the most opaque substance known, and that, con- sequently, dtiring a fog, when this sort of signaling device would be most needed, the radiations to be detected would be completely absorbed within a very short distance from the soxu-ce. Many of the questions submitted were so impractical that it was easier to advise what not to do than to suggest improvements. The selenium cell had its share of suggested applications, nearly all of which, as it turned out, seemed impractical. For example, in reply to inquiries from the Naval Proving Ground, Indian Head, Md., on the applicability of photoelectric devices (seleniimi cell) for registering the speed of projectiles, some laboratory tests were made which indicated that such devices are not sufficiently positive in action to use in such tests. Another inqtiirer desired to obtain a series of photographic ex- posures at frequent intervals throughout the day, and wished to use the selenitim cell (in some manner that he had not yet worked out) to regulate the diaphragm opening of a camera. The inappli- cability of the selemtun cell was indicated to him, and a series of tests was made which demonstrated the adaptability of the Cal- lendar stmshine receiver and the sliding mechanism of the recorder to the actuation of the diaphragm opening of a camera. By con- necting the diaphragm with the sUding arm of the recording mecha- 250 WAR WORK OP THE BUREAU OP STANDARDS nism, the to-and-fro motion of the latter regulates the diaphragm opening of the camera, and hence the amount of light reaching the photographic plate. Members of the National Research Council conferred on the applicability of the selenium cell as a means of communication for airplanes, etc. The potassium photoelectric cell deyice described elsewhere in this report was suggested as best adapted for such signaling. In reply to inquiries from the General Engineer Corps on the feasibility of using thermal radiations for destructive purposes, as, for example, igniting a captive balloon, it was shown that ter- restrial somrces of radiation were not powerful enough. Another inquirer thought sunlight could be used; but among other things, the method had not been thought out oi producing the concave mirror 20 to 40 feet in diameter which would be required in order to obtain sufficient intensity to ignite an object at a distance of 3 to 5 miles. To a manufacturer of quartz mercury vapor lamps, and to a paper manufacttirer using such lamps in fading tests, information and specific directions were given on making radiant power Ufe tests of mercury vapor lamps, and their applicability in fading tests of this kind. The feasibility of testing toxic gases by radiometric instruments was discussed with military officials from the Bureau of Mines, and selenium cells made at tlve Bureau were provided for experi- mental work in the development of tests for the presence of these gases. In response to a request from the National Research Council, a representative of the Bureau attended a conference at Columbia University, New York (August, 1918) , at which were presented the results of experiments on the detection of various objects such as airplanes, human beings, etc., which give oflf a certain amount of heat, by means of a thermopile (after tiuis Btureau's construction) placed at the focus of a 2 -foot concave mirror. By invitation from the special board, a representative of the radiometry section demonstrated the Bureau's thermal radiody- namic secret-signaling apparatus in September, 1918, at the naval experimental station. New London, Conn. In addition to the specific cases just cited, information was given on various radiometric matters such as the reflective properties of metals useful in searchlights, etc.; the thermal radiodynamic signal- ing devices; photo-electric substances; eye-protective glasses; the RADIUM 251 constancy and methods of standardization of sources of ultra- violet radiations of high intensity ; emissivity of paints to be used for preventing the heating of certain kinds of electrical machinery; construction of ironclad Thomson galvanometers; standard black- ening of radiometer receivers; and various other radiometric questions having a military application. Publications The following recent publications in radiometry contain data which have a military application, though, of course, some of the earlier investigations were not undertaken with that specific end in view : Technologic Paper No. 93 (3 ed., rev. and enlarged). Glasses for protecting the eyes from injurious radiations. Scientific Paper No. 229. Various modifications of bismuth-silver thermopiles having a continuous absorbing surface. Scientific Paper No. 282. Sensitivity and magnetic shielding tests of a Thomson galvanometer for use in radiometry. Scientific Paper No. 300. Emissivity of straight and helical filaments for tungsten. Scientific Paper No. 308. Reflecting power of tungsten and stellite. Scientific Paper No. 319. Instruments and methods used in radiometry, III — ^the photo-electric cell and other selective radiometers. Scientific Paper No. 322. Photo-electric sensitivity of bismuthinite and various other substances. Scientific Paper No. 325. Spectroradiometric investigation of the transmission of various substances. Scientific Paper No. 330. The decrease in ultra-violet and total radiation with usage of quartz-mercury vapor lamps. Scientific Paper No. 338. Some optical and photo-electrical properties of molyb- denite. Scientific Paper No. 344. The spectral photo-electric sensitivity of silver sulphide and several other substances. Thermal-Radiophonic Devices for Secret Signaling This subject is treated under the heading " Invisible Signaling." RADIUM Diiring the war self-luminous materials were very extensively used for the illumination of aircraft instruments of all kinds, of gunsights, of targets for indirect fire, of marching compasses, of watches, of navigation instruments; in short, they were used wherever an entirely self-contained light of low brightness was desired. In determining the suitability of these materials for given pur- poses and in drawing specifications for their use numerous ques- tions arose. Many of these questions related to points concermng which much misinformation existed, others to details that were easily and commonly overlooked, and others to features that 252 WAR WORK OP THE BUREAU OE STANDARDS were but little understood. The answers to some were to be found scattered through the literattue of physiological optics, of radioactivity, and of phosphorescence. The answers to others required a knowledge of the properties of commercial self-luminous materials. Up to June, 191 7, no article treating of the properties of such materials had been pubUshed, and but two bearing directiy upon (noncommercial) self-luminous materials were available. Nowhere was to be foimd in collected form information covering the many important practical points involved in the efl&cient use of these materials. , Early in 191 6 the Bureau of Standards began an investigation of these materials to collect information concerning them and to study the problem of the precise measurement of their bright- ness. Suitable apparatus was constructed (Fig. 29) and a long series of measurements of many specimens obtained from dif- ferent sources was begun in October, 1916; some of these materials are still under study. In the spring of 191 7 a study of the various problems connected with their practical use was begun. Consequentiy, when in the summer of 191 7 these materials acquired a military importance, the Bureau was in a position to answer many of the questions that arose and to tmdertake the investigation of various problems that presented themselves from time to time. It soon became the recognized source of in- formation on the subject. The Bureau's war activities in this field embraced four distinct lines: 1. The furnishing of information and recommendations. 2. The investigation of specific problems. 3. The preparation and standardization of equipment for the factory testing of dials. 4. The routine precise measurement of the brightness of the markings on articles submitted for test. Information and Recommendations Information and recommendations were disseminated by means of six typed circulars, many letter reports, and numerous confer- ences with representatives of those branches of the Army and the Navy who were especially interested in the subject and with repre- sentatives of the companies manufacturing and appl3ang these materials. Special reference should be made to the large number of conferences in the last half of 191 7 between representatives of the Bureau and representatives of the National Research- RADIUM 253 Council and of the science and research division, Signal Corps, U. S. A. A few of the questions that had to be answered and of the errors that had to be corrected will illustrate the nature of the information that was required. 1. Shall the dial plate be silvered? Dial plates illuminated before the summer of 191 7 were frequently silvered. These dials present a very pleasing appearance by day and in total darkness, but under certain moonlight conditions the contrast between the dial plate and the markings is so slight that the markings can not be seen. The dial plate should be black. 2. It was commonly believed that the effective life of radium luminous materials is comparable to that of radium (radium is half disintegrated in 1760 years). Actually the effective life of these materials is relatively short. 3. It was urged that the material should be specified by its radium content. Actually a specification of the radium content is not sufficient to determine the brightness or life of the material. 4. It was suggested by some that a high brightness should be specified in order to insure a long life. While increasing the initial brightness does increase the effective life, the increase in cost is out of all proportion to the increase in life. 5. What is the effect of replacing the radium by other radio- active material ? 6. Shall dry material or that mixed with varnish be used? How does the brightness differ in the two cases? 7. How soon after the material is made shall the illuminated article be inspected for acceptance ? 8. Should the dial contain many marks or few? 9. What brightness is required? How does the brightness required vary with the size and style of the luminous markings ? 10. Shall the Ituninous lines for rear sights be made very narrow ? It was found that under working conditions the apparent width of luminous lines for rear sights is practically the same whether the line be 0.5 or 1.7 mm in width. Throughout this range the apparent width is about 5 or 6 mm. Hence very narrow lines, requiring brighter material, are desirable. It was to answer and to furnish the information needed to enable one to discuss intelligently such questions as these that the six typed circulars were issued. They were rather hurriedly prepared to meet immediate needs, in some measure overlapped. 254 WAR WORK OP THE BUREAU OF STANDARDS and were in no sense intended to be regarded as finished products. They were issued in small editions and supplied to those most interested. They do not lend themselves to abstracting, but an idea of the groimd covered may be indicated by the following table of contents : RLC-i. Self-luminous materials contaiiiiiig radioactive excitants; a report from National Bureau of Standards, March lo, 1917; revised November 16, 1917. pp.17. General description of these materials; Properties of the phosphcarescent base; Ptoperties of the radioactive excitants — ^Radium, Mesothorium, Radiothorium, Polo- nitim; Growth and decay of brightness; Manufacturers, materials, prices; Applied material; Legibility; Specifications and testing. RLC-2. Supplement to the preceding report, April 10, 1917. pp. S- This treats mainly of the life of dry material contained in tubes. RLC-3. Specifications for sel?-ltmiinous dials, pointers, scales, etc. Proposed to science and research division. Signal Corps, U. S. A., September 24, 1917. pp. 3.- (Superseded by RLC-6,) RLC-4. Notes and suggestions concerning luminous instrumentSj November 16, 1917. pp.6. Size of characters. Spacing, Nuinber of marks. Positional reading, Sjfmbols, Bright- ness, Application, Pointers. . : ; ! ■; RXC-s- Self-luminous materials: Brief notes covering a few points of practical importance. January 8, 1918. pp. 14. Unapplied material — General remarks. Growth in brightness for 3 weeks, Decrease in brightness after 3 weeks. Effect of light. Brightness, Fineness of grain; Applied material — Brightness reduced on application. Initial growth after application. Ulti- mate decrease in brightness; Manufacturers and prices; Relative values of different products. RLC-6. Notes and recommendations regarding specifications for the illuminating of articles by means of self-luminous materials containing radioactive excitants. August 10, 1 918. pp. 24. Basis of specifications, Self-luminoixs materials, Unit of brightness. Total light, Brightness of markings. Summary, A form of specifications. Appendix. An illustrated report (9 pp., 4 illus.), describing in detail tiie equipment used by the Bureau of Aircraft Production for the factory inspection of dials, was prepared February, 1919, for and at the request of the New York Navy Yard. Investigation of Specific Problems Some of the problems that have been investigated especially for the ptUTJose of obtaining information of military value are the following: Variation of brightness with the thickness of the material. Variation of brightness with the width of line. Effect of immersion in kerosene. Methods for applying the luminous material. Brightness required for specific dials, gun sights, targets, indi- cators, compasses, etc. In dealing with the last-named problems the Bureau cooperated closely with the parties interested. Usually the Btu-eau prepared apparatus by means of which an article of the desired type could RUBBER 255 be seen with markings of variable or various brightnesses. Then the parties interested examined the article at night and in con- ference with representatives of the Bureau. In this way the brightness that seemed most desirable was determined. Certain stencils were designed to use as the rear lights for trucks. This work was undertaken by request for the purpose of determining what form will give the driver of the following truck the most information regarding his distance from the pre- ceding one. The parties requesting the work lost interest in it after the armistice was signed, and the work was not completed. Equipment for Factory Inspection of Dials Apparatus was prepared and standardized for the use of the Bureau of Aircraft Production in the factory inspection of lumi- nous dials. A carefully cut stencil of the luminous dial under test is backed by a sheet of milk glass and illuminated from the rear. The lamp illuminating the stencil is moved until the open- ings in the stencil appear to be of the same brightness as the cor- responding markings on the dial. If such a match can be obtained only by moving the lamp farther from the stencil than a prede- termined distance, the dial is too faint and is rejected. Five of these outfits were adjusted and calibrated, and the nec- essary lamp renewals were provided. Routine Precison Measurements Routine precision measiwements of the brightness of selected articles were made for several purposes: (i) to enable a prospec- tive bidder to ascertain if a proposed material will meet the requirements; (2) to serve as a basis for the award of contract; (3) to serve as a check on the factory inspection; (4) to furnish information regarding the tmiformity of the work; and (5) to determine the permanence of the brightness. The study of these materials and problems is being continued, and it is expected that in the course of the year a circular treat- ing these materials will be issued by the Bureau. ROPE, MANILA This subject will be found under the title "Manila Rope." RUBBER Solid Tires Early in 191 7 the attention of the Bureau was called by the Quartermaster Corps of the War Department to the tmsatisfactory quality of solid motor truck tires which the Government was 256 WAR WORK OP THE BUREAU OP STANDARDS obtaining in the open market. Although these tires were bought from various manufacturers, they all appeared to be about the same grade, which was no better than that of the relatively poor solid tires with which American motor trucks had been equipped during their service on the Mexican border. At the suggestion of the Bureau samples of all these tires were secured for tests and analyses, and the results of the investiga- tion were correlated with the service rendered by the tires. The results of this combination of laboratory and field work were used as a basis for a series of conferences between tire manufac- tiurers, officials of the Quartermaster Corps, and the Bureau of Standards. After a number of such conferences a set of specifi- cations was drawn up jointly for solid-rubber tires for motor trucks and artillery wheels, and were first adopted in November, 191 7, followed by revised editions on December 5 of that year. These specifications were adopted subsequently by the Motor Transport Corps and finally as standard specifications by the pturchase, storage, and traffic division of the War Department. The specifications call for physical tests of the rubber com- pounds and fabric, chemical analyses of the rubber compound, and physical and chemical tests of the steel base bands. The Biureau of Standards' methods are specified for making these tests and analyses. Dturing the period of the war following the adoption of these specifications, inspectors were stationed at the" various plants. These inspectors took i tire from every 500 and cut out appropriate samples for test. The physical tests of the rubber compotmd were made at the plant, whUe the samples for analyses of the rubber and tests of the base bands were forwarded to the Btu-eau. Difficulties were encountered in the anylysis of the rubber stock in the tires, because two relatively new com- poimding ingredients, glue and gasblack, were being used. Meth- ods, therefore, had to be devised for the quantitative determina- tion of these two ingredients. It was found that by making a modified Kjeldahl determination to arrive at the quantity of nitrogen present, the percentage of glue in the stock could be estimated quite closely. The determination of gasblack was much more difficult, but a procedure was soon arrived at for this deter- mination, which with slight modifications and improvements was used during all of the tests. The research work on this deter- mination forms the subject of technologic papers No. 136. Although much of the routine physical testing at the various RUBBER 257 plants was done by local inspectors, samples were sent regu- larly to the Bureau as a means of checking their results. During the year beginning with March, 1918, over 500 complete rubber analyses were made, representing the purchases of 250 000 solid tires having a value of about $20 000 000. It will thus be seen that the War Department through the Bureau's aid was enabled to buy this material on rigid specifications and to have all purchases tested for compliance with these specifications. The tires had to pass tests made at the Bureau before they were finally accepted by the "War Department. The result of this coop- erative work was a vast improvement in the quality of solid tires fiuniished the War Department. Through the specifications the department was able to buy goods of standard quality on the basis of competitive bids. Pneumatic Tires and Tubes Experience in connection with the development of specifications for and the testing of solid tires as described above led to a similar procedure for pneumatic tires. The results of the solid-tire test- ing had been so satisfactory that it was deemed advisable to buy pneumatic tires on the same basis. This, however, was a much more complicated and difiicult matter, since pneumatic tires are made of so many different tjrpes and qualities of materials. Speci- fications would be required for each particular type, and a deci- sion as to what constituted a good tire was a very difficult one to reach. Manufacturers differ greatly as to the grade of materials they use and as to the methods of manufacture. A great many conferences were held concerning these specifications, and although the work was started in November, 191 7, its tremendous quan- tity and the great differences of opinion involved halted the work temporarily vmtil the outcome of that on solid-tire specifications should be known. The work on pneumatic tires was again taken up during the summer of 191 8, and in the following October speci- fications were issued as a result of the cooperative efforts of the Motor Transport Corps, the War Industries Board, the Rubber Association of America, and the Brn-eau of Standards. The details of these tests are in accordance with the Bureau's methods, and references are made in the specifications to the effect that tests and analyses shall be in accordance with the methods in current use at the Bureau of Standards. The tests to be per- formed on a pneumatic tire are numerous and of a very diverse 26035°— 21' 17 258 WAR WORK OP THE BUREAU OP STANDARDS natiire. The dimensions of the tire must fall within definite limits, the fiber must be of a certain quality^ and the different rubber stocks must be subjected to physical tests and chemical analyses. Determination of the adhesion or " friction " between the different layers of rubber and fabric must also be made. As it was neces- sary to reduce this testing as far as possible, while still retaining a good check on the quality of tires delivered, it was agreed that one tire in every thousand or less should be tested. As soon as the specifications were issued manufacturers conmienced to make tires in conformity with them. Before placing a manufacturer on the list for competitive bidding, his tire had to be tested and ap- proved by the Bureau, and, as a result of this requirement, practi- cally all of the larger manufacturers sent two or more tires to the Btireau of Standards for test. The increase in work thus brought about required the use of much more extensive quarters for the rubber and chemical laboratories. Because a large majority of these tires were produced in the district near Akron, Ohio, and since there was no ftirther room for expansion at the Bureau, it seemed desirable to locate a branch laboratory at Akron. As the result of an agreement between the University of Akron and the Bureau, suitable rooms for this work were fitted up in their chemi- cal laboratory. The fxmds for the equipment were made avail- ble by the Motor Transport Corps, and the resulting laboratory was well equipped to handle the routine analytical work necessary. This work was so largely curtailed by the singing of the armistice that the laboratory was never used to its full capacity. It was expectedjthat should the war continue, the Government would buy about $5 000 000 worth of pneumatic tires a month, necessi- tating a laboratory force of 1 5 chemists to do the analytical work called for by the specifications. As in the case of solid tires it was the intention to station inspectors at the factories to secure sam- ples and do the physical testing, their results to be regularly checked by figures obtained on duplicate samples sent to the Bureau. Supplementing the specifications for pneumatic tires were speci- fications for inner tubes. The tests of tubes were handled in all respects in the same way as pneumatic tires, in nearly all cases manufacturers making both tubes and tires. Since the specification plan for pneumatic tires was put into effect at about the time of the signing of the armistice, the full re- sults expected were not attained. There is every indication that RUBBBR 259 had hostilities continued, the results from this work would have been as satisfactory as those obtained in connection with solid tires. Gasoline Hose The preparation of specifications for a large number of parts used in the construction of airplanes formed some of the earliest work of the standards committee of the Society of Automotive Engineers in cooperation with the Aircraft Production Board. Among other things were included specifications for rubber tubing to conduct gasoline from the tank to the engine of an aircraft. Before specifications could be written for such a tube, considerable experimental work had to be done to determine whether rubber was suitable for this ptirpose, and if so, what kind of rubber com- pound and what type of tube construction would be most satisfac- tory. The Btireau cooperated with both the Aircraft Production Board and the Society of Automotive Engineers in carrying out this work. The Bureau's part consisted in the laboratory testing of various tjrpes of tubes as they were submitted by manufactiurers. The t5T)e developed as a result of this work was made up of a com- bination of rubber and fabric containing a metal spiral reinforce- ment in the center of the tube, which prevented the tube from closing because of the distortion of the inner rubber by absorp- tion of gasoline. Specifications drawn up for this type of rubber gasoline hose were used as a standard by the Aircraft Production Board. Oil-Proof Packing The Ordnance Department needed a special oil-proof rubber packing for the recoil mechanism of a particular type of gun. The French had been using a rubber packing called " Dermatine " in a similar gim, and this proved to be fairly satisfactory. As the Ordnance Department was not satisfied with the packings originally submitted by manufacturers samples of Dermatine packing were obtained from France and were tested at the Bureau. Samples of Dermatine obtained through the Ordnance Depart- ment, together with the Bureau's analysis of the material, were furnished to various manufacturers, who then carried out research work of their own and submitted samples for test. These were tested in a practical way by the Ordnance Department, while a check chemical analysis was made in each case by the Bmreau. As a result of these tests a certain product was picked out as the most satisfactory, and this was then supplied to the department, the 26o WAR WORK OP THE BUREAU OF STANDARDS individual deliveries of packing being checked by chemical analyses made at the Bureau of Standards. Insulated Wire Throughout the period of the war the Biweau made physical tests and chemical analyses for controlling the quality of insulated wire and cable, particularly submarine cable which was being purchased under specifications issued by the Panama Canal and the torpedo depots of the Coast Artillery Corps of the Army. A joint committee on the analysis of rubber insulation, comprising members of the American Institute of Electrical Engineers, the American Society for Testing Materials, and the Bureau of Standi ards, was requested by the Navy Department to furnish a set of specifications with details of test methods for a 40 per cent rubber compound for insulated wire. The specifications recom- mended by this committee were adopted by the Navy Department and later by the Signal Corps. Mechanical Rubber Goods Rubber hose of different kinds, including that used for steam, water, air, etc., together with miscellaneous rubber goods, such as belting, valves, and packings, commonly classed as "mechanical rubber goods," were investigated by the Bureau at the request of the "War Department, with the object of developing standard specifications for materials of this sort to be used as a basis of award in the purchase of supplies for military purposes. The methods followed in investigating and correlating the physical and chemical properties of the various kinds of rubber goods had for the most part previously been developed at the Bureau in connection with similar work for other Government departments. In addition to chemical analyses to determine the percentage of rubber present and its quality, physical tests were made to deter- mine the quality and strength of fabric and the strength, stretch, and elasticity of the rubber as an indication of the correctness of vulcanization. A comparative study was made of various samples in each of the numerous classes as a means of judging the excellence of the fabrication and when practicable the materials were subjected to conditions approximating those of service, and in some cases accelerated aging tests were carried out to estimate the relative life of the rubber compounds used. As a result of these laboratory investigations and of numerous conferences with manufacturers and representatives of the War Department, standard specifications have been adopted for practically all SAFETY STANDARDS FOR INDUSTRIAL ESTABLISHMENTS 261 miscellaneous rubber goods used for military purposes, and these specifications have been issued by the purchase, storage, and trafl&c division of the War Department as one of their catalogues. In all of the above-mentioned work the complete laboratory equip- ment at the Bureau, such as the small rubber mill (Fig, 30), was of great assistance in the prompt handling of the investigations. Rubber Jar Rings for Canning It is perhaps not generally known that a very important item in connection with the matter of food conservation is the subject of rubber jar rings. At the request of the States Relations Service of the Department of Agriculture the Bureau cooperated in the production of standard specifications for jar rings in order to establish a standard of quality for use by the public and to eliminate the many inferior brands of rings which have been the cause of extravagant waste as a result of food imperfectly pre- served. The method adopted by the Bureau was to investigate the physical and chemical properties of the various grades of jar rings on the market and to compare the results obtained with the known behavior of rings in service. It was found that of the many different grades of commercial rings obtainable, those having certain physical properties that are easily measurable may be depended upon to give satisfactory results. Specifications including a series of simple tests and measurements were written by the Bureau and are now published by the Department of Agriculture and recommended for use by the public. Druggists' Supplies At the request of the General Purchasing Ofl&ce of the Field Medical Supply Depot of the Army, an investigation was con- ducted on a number of rubber articles classed as druggists' sup- plies, such as rubber hospital sheeting, hot-water bottles, rubber tubing, etc. Physical and chemical tests were made in accord- ance "with the Bureau's standard procedure for such materials, and the award of contracts for hospital supplies was made on the basis of results obtained. SAFETY STANDARDS FOR MILITARY INDUSTRIAL ESTABLISHMENTS In 191 7 a field inspection of the military and naval industrial establishments was made through the initiative of the United States Employees' Compensation Commission. This field inspec- tion resulted in the submission of a number of recommendations 262 WAR WORK OP THE BUREAU OP STANDARDS for increasing the safety in the arsenals and navy yards. Among these was a recommendation that a safety engineer be appointed for each industrial establishment in the Army and Navy. As a result safety engineers were appointed .by the War and Navy Departments and stationed at various arsenals and navy yards for the purpose of assisting in accident prevention. At the time of appointment of the safety engineers an organized effort was made in the Federal industrial establishments to reduce accidents, but no standards for the methods of guarding and for material specifications for such guards were available in either the War or Navy Departments. It was accordingly necessary to effect an organization by which this work of drawing up safety rules could be accomplished. The safety engineers held conferences at Washington on November 14 and 15, 1917; at Brooklyn on January 15, 1918; and at Norfolk on March 14 and 15, 191 8, and invited representa- tives of the Bureau of Standards to be present. Committees were appointed to draw up safety rules and the Biu-eau was asked to cooperate with these committees in perfecting and codifying the rules. For this purpose, representatives of the Watertown Arsenal and the Boston Navy Yard were detailed to the Bureau of Standards for several weeks in the spring of 191 8. Preliminary drafts of these safety standards were distributed for criticism and suggestions and after final correction were adopted by vote of the safety engineers. Illustrations were prepared and the whole submitted to the officials of the War and Navy Departments for formal approval and promulgation. These standards have not been published by the Bureau of Standards. The safety rules cover the following subjects: Building construc- tion, crane construction, elevators, fire appliances and equipment, toilets, wash rooms and locker rooms, power plants and prime movers, power-transmission apparatus, machine guarding, remote- control apparatus, head-and-eye protection. Each of the codes represents the collection in Codified form of the best existing methods and practices then in vogue arranged in a form to afford easy reference. Nothing original is claimed for them, although considerable new material has been added. Arrangements were made whereby the safety engineers were to prepare and submit to the Bureau suitable illustrations for each of the codes. At the time the armistice was signed, illustrations for five of the codes had been received. Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 30. — Small experimental rubber mill ,. This mill permits the Bureau to study the processes employed in rubber making. During the war rubber was employed on a vast scale for tires, hospital supplies, etc. - L '**jtP 'z V -^ tI '_£Zf!^ Fig. 31. — Sound-ranging apparatus used to determine the position of a gun by meas- uring the time at which the sound wave caused by the gun's discharge reaches suc- cessive points Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 32. — Exterior and interior views of plant for the recovery of toluol from the manufacture of illuminating gas Large Quantities of toluol were needed in the mantifacture of explosives, and the Bureau investi- gated the problem of recoverinE toluol without impairing the gas furnished SEARCHLIGHTS 263 The preparation of these safety rules involved in some cases tests and investigations of the devices on the market and of the properties of the material employed for protection. This was notably the case in connection with protection of eyes of workmen from flying particles and from injurious radiation, where it was necessary to investigate the transmission of the different kinds of glass, especially for ultra-violet radiation, and to test glass of various kinds with respect to its ability to resist blows. A widespread demand that these Federal standards be made available for general use has resulted in the Bureau undertaking to elaborate them and more fully perfect them from the stand- point of general application and mandatory enforcement by vState and municipal officials. It is expected that this work will result in the development of a series of national industrial safety codes. SEARCHLIGHTS The high-power search lamps have acquired tremendous impor- tance in the last few years. Few data, however, are on record regarding some of the fundamental characteristics of these search lamps or bearing upon some of the more important problems of searchlight illumination. An investigation in cooperation with the Engineer Corps of the Army was carried out, using the combined facilities of the colorimetry section of the optics division and the photometry section of the electrical division of the Bureau. This work is briefly described below. From data collected by the Bureau, two distinct types of distri- bution appear: That in the medium-intensity arcs resembles a "rose" distribution; that in the high-intensity a "cardioid" dis- tribution. The development of high-intensity arcs by the use of specially prepared carbons has resulted in the production of a light which is very blue relative to artificial incandescent sources. In a general investigation of searchlights, including the effectiveness and suitability of their illumination for various purposes, it appeared desirable to have at least approximate quantitative data on the color of this light. An investigation was made by spectrophotometric and special colorimetric methods, and a report was issued to the Engineer Corps and others, January 23, 1919. Another report was made to the American Physical Society, April 25, 1919, and wns published in the Physical Review. 264 WAR WORK OF THE BUREAU OK STANDARDS The result of these investigations may be roughly summarized by saying that the color of the light from these arcs is approxi- mately equivalent to the light of the noon siui in "Washington, although relatively more intense in the blue-violet. The diffusion or scattering of light along the path of a searchlight beam is very important from several standpoints. The diffusion of the light out of the beam decreases the intensity of light falling upon the target. This diffused light limits the visibility of the target for an observer near the search lamp who must look through and along the beam. On the other hand, this diffusion of the light along the beam makes possible the use of the searchlights as land- marks for military and other purposes, and also facilitates the train- ing of the beam in a given direction, particularly when combined action of two or more search lamps is desired. Measurements from various directions have been made of the brightness of the path of the beam due to the diffusion of the light by the atmosphere. The percentage of polarized light in the dif- fused light has also been determined. In order to make any conclusions as to the size and distance of a given target that may be seen in a searchlight beam, there is demanded some knowledge as to what is the minimum difference of brightness that the eye can detect. Laboratory experiments were made to collect data upon the contrast sensibility of the eye imder conditions closely simulating those under actual searchlight illumi- nation. Very definite results have been obtained. In connection with the testing and use of search lamps, the ques- tion of how much light is lost as the beam penetrates the atmos- phere must be considered. It is important to know what the transmission of the atmosphere is when tests are made at great distance from the search lamp. Very meager definite data exist upon the transmission of the atmosphere under all weather conditions — clear, hazy, foggy, rainy, etc. It is important also to have a knowledge of the relative trans- mission for different-colored lights. A method has been devised to procure relative data on the trans- mission of the atmosphere throughout the spectrum, and some observations have been made for a clear atmosphere, for one of rather high humidity, and for rains. These data are not only valuable for the searchlight engineer, but are important also in aerial and in other photography and to the illuminating engineer interested in automobile head lamps, etc. SOUND-RANGING APPARATUS 265 A chart has been made showing where all losses of Ught in the search lamp take place, and their magnitude from the moment the current flows through the arc to the time that the beam impinges upon the target. A physical photometer was constructed for use in coimection with the photometry of search-lamp arcs. On account of the high intensity of the light from arcs, they are admirably adapted to the application of such an instrument. SOUND-RANGING APPARATUS Introduction The process of locating a distant or concealed piece of artillery by means of the sound waves produced when the gun is fired is a new and very important development of the present war. I^et us assume that a gun is fired at a position which we may call i. This gives rise to a soxmd wave that passes out in all directions. If now, we have an instrument similar to a telephone transmitter located at each of three different stations 2,3, and 4, which may be selected more or less at random, but whose positions are known^ it will be evident that the soiuid wave will, in general, arrive at these stations and affect telephone transmitters at different times. If wires are brought from the telephone transmitters to the cen- tral station located at any convenient point and connected to a telephone receiver, the sound of the gim would be heard suc- cessively as the wave front arrived at each station and the inter- vals of time between sounds would be a measure of the differ- ences in distance from the gtm of the various stations. It has been found possible to provide apparatus which will record the instant of arrival of the wave at each station and so afford a meas- ure of the time intervals between the arrival of the sound at the several stations. The principle is, therefore, very similar to that of a telephone system except that a recording apparatus is used instead of the usual telephone receiver. The time intervals having been measured and the locations of the stations 2,3, and 4 being definitely known, it is possible that any one of a. variety of meth- ods will locate definitely the position of the gun. This is the fundamental principle of sound ranging which has been worked out in several practical embodiments used by the French and British services. The principle of sound ranging can be easily imderstood if we consider that a sornid wave passing from a gun has a wave front which is circular in form. In due time this will reach the first 266 WAR WORK OP THE BUREAU OF STANDARDS listening Station, which we have called 2, A short time later the sound wave will reach station 3, and still later station 4. If now, we draw a circle about point 3, tangent to the circle of the wave front through station 2, the radius of this circle di willbe the distance traveled' by sound in the time interval between the ar- rival of the sound wave at stations 2 and 3. Similarly, the radius ^2 represents the distance traveled by sotmd in the interval be- tween the arrival of the sound at 2 and 4. Since the velocity of sound in air tmder given conditions is definitely known, if we have determined the relative time of arrival of the sotmd wave at the three stations, the; distance d, and d^ can be directly calculated. As an illustration of the method of determining the gun position, let us consider that the distances d^ and d^ have been determined as above stated. Then let us draw a circle about the point 3 with radius d^ and another about the point 4 with the radius d,. It is evident that the problem is now the familiar geometrical one of finding the center of a circle which passes through the given point 2 and is tangent to two given circles drawn about the points 3 and 4, and having the given radii dj and d^, respectively. The solution of this problem has been worked out in a variety of ways both graphically and analjrticaliy. In actual practice the determination of the position of the gim is by no means as simple as outlined above. The velocity of sound varies with the temperature and humidity of the air, and the position of the wave front at any time is affected by the direc- tion and velocity of the wind. All these variable factors have to be observed at frequent intervals and corrections applied. Fur- ther, even if only one gun is fired, three distinct soimds are pro- duced as a rule; one due to the firing of the gun, another due to the passage of the projectile through the air which is very pro- nounced when the velocity of the projectile is much greater than that of soimd, and the third which may be produced by the explo- sion of the shells. These sounds, in general, do not reach the recorder in the order mentioned above. It is necessary to dis- tinguish between these if a correct solution of the problem is to be had. Further, it will often happen that a number of guns are fired so nearly at the same time that the number of sounds pro- duced by all of them will be recorded on the same portion of the chart, and it often becomes necessary for one to distinguish be- tween the record produced by firing the gun and those produced by the shell, but it is also necessary to determine which portions SOUND-RANGING APPARATUS 267 of the record received from the different stations belong to any particular gun. It will thus appear that in actual practice sound ranging involves a rather complex practice. However, the appa- ratus and methods of use have now been developed to such a high degree that it is possible imder average practical conditions to determine all the information referred to above and make the necessary corrections for the variable factors involved, and deter- mine quite closely the position of the gim within one to two min- utes after the gun is fired. Upon the entrance of the United States into the war, the matter of sound ranging was brought to the attention of this Govern- ment by the French Scientific Commission. This commission strongly urged that an investigation of the matter be taken up at once for the purpose of determining upon a system to be used by the American Army, either by selecting from the systems already in use the one best adapted to the purpose of sound ranging, or developing a new and improved system. The portion of this investigational work which has been done at the Bureau of Stand- ards forms the subject of the present report. The work of the Bureau of Standards on sovmd ranging was taken up following a conference held in the office of the Council of National Defense. This conference was attended by the officer who had been placed in charge of the sotmd-ranging work, a member of the National Research Council, two officers represent- ing the French Scientific Commission, two members of the staff of the Bureau of Standards, and others. At this conference the sound-ranging problem was discussed in its general aspects and an agreement was reached whereby the War Department's staff, then located at Princeton, would take up an investigation of the Dufour and Bull systems, and the Bureau of Standards would make a study of the Cotton- Weiss system and the so-called T-M system. The work of the Bureau in this connection was preceded by a careful study of various descriptions of sound-ranging apparatus and of reports bearing on its operation which had been placed at the Bureau's disposal by the French Scientific Commission. Two sets of apparatus were designed and constructed, one embodying the Cotton-Weiss system and the other the T-M system, and after a little preliminary work in the laboratories of the Biu-eau of Standards these two sets of apparatus were taken to the Sandy Hook Proving Groimd for field tests. 268 WAR WORK OP THE BUREAU OF STANDARDS Preliminary Work The first of the two types of apparatus investigated by the Bureau, the so-called Cotton-Weiss system, is a nonrecording system which was used to a considerable extent on the French front quite early in the war, but, according to the Bureau's infor- mation, was after a short time used only in a few special cases. In this system two sets of apparatus are used at a distance of a kilometer or more apart, each set having two receiving stations located perhaps lOO m distant from each other. An indicating device on each set shows the difference in time of arrival of sound at the two listening stations. These differences in time deter- mined at the two separate locations afford sufficient data for calculating the approximate position of the gun. In the second type of apparatus, the so-caUed T-M system, three or more receiving stations are used, each of which is con- nected by a pair of wires to a central recording station. At the recording station there is an oscillograph or galvanometer actuated by changes in current on the lines to the receiving stations, and the pointers of the oscillograph trace a line on a smoked paper tape run at uniform speed. The difference in time of arrival of the sound from the gtm at the various receiving stations is deter- mined by measurements of distance on the smoked paper, and these differences in time determined for three or more stations provide data for calculating the position of the gun. Complete sets of both Cotton- Weiss and T-M t3TJes of apparatus were constructed and taken to Sandy Hook, where several days were spent in making tests. During the same period the War Department's staff at Princeton made similar tests on both the Dufour and Bull types of apparatus. As a result of these early tests it was decided at a conference between representatives of Princeton and the Bureau to abandon for the present fiu-ther development work on the Cotton- Weiss system, chiefly because it appeared that the recording types of apparatus presented marked advantages over the nonrecording Cotton- Weiss system. The first series of tests with the T-M apparatus did not yield satisfactory results, due partly to troubles with the microphones and partly to a lack of sufficient sensitivity in the oscillographs. A sufficient number of tests were made to bring out fully the weaknesses in the preliminary design, after which the apparatus was brought back to the Bureau of Standards and remodeled. Improved microphones were constructed, and a new tj^e of oscil- SOUND-RANGING APPARATUS 269 lograph which will be described in detail later was developed, and the apparatus was again taken to Sandy Hook for test. During this series of tests the entire apparatus worked in a very satisfac- tory manner and three receiving stations were laid out roughly on rather short base lines, and some actual ranging tests were made. These ranging tests, although only partially satisfactory, were very instructive and served to bring out certain minor defects in the ap- paratus which required elimination. This second series of tests revealed the desirability of conducting rather extensive studies of several phases of the sound-ranging problem, and a series of such investigations was planned and carried out by the Bureau prior to designing and building the improved sets, which were finally devel- oped. A brief account of these experiments is given below; de- tailed descriptions of the apparatus and methods developed and used by the Bureau will be given later. Brief Review of Investigations of Various Phases of the Sound- Ranging Problem The preliminary work showed that the original form of T-M microphone was not sensitive enough for effective practical use. Furthermore, the resistance of the microphone buttons is so low that a relatively low-line resistance is necessary in order that the latter will not be too large in comparison with the frame. In gen- eral, it is not practicable to secure this low-line resistance by using heavy wire because of the bulk and weight of line equipment as well as the cost of the wire. The Bvireau set to work on the prob- lem of developing a more sensitive microphone and one which could be used with a relatively high-resistance line wire. These objects were finally accomplished by using a four-diaphragm mi- crophone, each diaphragm having a separate microphone button attached to it. These four microphone buttons were connected electrically in series, thus giving a total line resistance of four times that of a single button. This arrangement was found to work very satisfactorily, and it was found possible to give all the sensitivity that was desired and still use a line resistance of the order of 1000 ohms. Another phase of the microphone problem that was given special attention was that of making the microphones selective to the muzzle wave. It was foimd desirable to secxure records that are free from the effects of the bow wave, or so-called crack, due to the velocity of the projectile moving through the air. A record with these effects eliminated is greatly simplified and rendered much 270 WAR WORK OF THE BUREAU OP STANDARDS easier of accurate analysis. The Bureau finally developed a mi- crophone which contained iextremely high sensitivity to the muzzle wave, but which would not give any appreciable response to the crack even though the shell was so near the microphone that the crack of the bow wave was many times greater than that of the muzzle wave. The microphones can be very readily rendered sensitive to the crack for special work, as occasion requires, with- out any change whatever in construction of the microphone; that is, a given microphone can be changed in the field to one very sen- sitive to a crack or to one which is imresponsive thereto, as may be required. A microphone was finally developed which was f oxmd to be very much less sensitive to wind pressure than the original T-M micro- phone. The original T-M oscillograph was of the moving-iron type, and the Bureau's preliminary work indicated that it was not one adapted to this work. Accordingly, a moving-coil oscillograph was developed, using a very short pointer and an electromagnet which gave a surplus of sensitivity for all practical purposes. This oscillograph has proved very rugged and reliable under service conditions, and it is believed to fully meet the requirements of the sound-ranging service. In the original apparatus used by the French Army the smoked chart was fixed after the record had been removed from the re- corder by laying an additional strip of paper over it and passing it through a, number of rolls. This was fotmd to give rather unsatisfactory results, such records being poorly fixed and rather easily smudged. Another method used for some time consists of dipping the smoked record in a thin solution of shellac and then permitting it to dry. The objections to both these methods were chiefly the danger of losing the record in handling, due to acci- dental smudging, and also the time elapsing before the record was available for study. This was important, particularly in the case of the shellac fixing. After considerable experiment the Bureau developed a method of fixing by causing the tape to pass over a drum which applied melted paraffin to the inside of the tape, which promptly soaked through and fixed the record on the other side. This was made a process continuous with the making of the record itself, so that the record came out of the recorder all fixed and ready for use. The danger of losing the records due to smudging a,nd handling was thus entirely eliminated and the SOUNDS TRANSMITTED THROUGH THB EARTH 271 record was immediately available for analysis upon coming out of the recorder. This method of fixing has proved eminently satisfactory in practice. In addition to the above a large amount o. experjnental work was done in connection with the design of a suitable type of trans- former for the recorder, the development of timing devices for accurate recording of time on the record, and, particularly, much attention was given to methods of procedure in sound ranging- A number of methods of analyzing the charts, working up the data, and calculating the position of the gun, including the making of various corrections for wind, temperatture, and humidity, were developed in connection with this work. The final form of this apparatus (Fig. 31) was completed and went to France in the autumn of 1918, arriving there only the day before the armistice was signed, so that opportunity was not afforded for trying it out imder actual war conditions. The apparatus has been very thoroughly tested, however, at the Naval Proving Grotmd at Indianhead, Md., and its success was fully demon- strated. SOUNDS TRANSMITTED THROUGH THE EARTH When the French Scientific Commission visited this coimtry in the spring of 191 7, they brought with them various devices for the detection of sounds which were transmitted through the earth. These were particularly valuable in listenting for mining opera- tions. In October, 191 7, the Engineer Corps and the Bureau undertook to develop these tjT)es of apparatus for the use of our troops abroad. Two types of apparatus for detecting earth vibrations of high frequency (sound waves through the earth) have been studied. The first type of apparatus is entirely mechancial and is called thi geophone (Fig. 33) ; the second type is electrical, and is called a seis- momicrophone. Geophone The geophone consists of a heavy mass suspended by two diaphragms inside of the case with thin air spaces above the upper diaphragm and below the lower diaphragm. One of these air spaces is connected by rubber tubing to one of the ears of an observer. If the geophone is placed firmly in contact with the earth, an earth vibration causes the case to move with the earth, but the steady mass due to its inertia and its suspension tends to 272 WAR WORK OP THE BURBAXJ OP STANDARDS remain at rest, so there is a relative displacement between the case and the steady mass. This displacement sets up in the air space alternate condensation and rarefaction, which are transmitted through the rubber tubes to the ear, thus producing the sensation of sound. Two geophones are usually used, one being connected to each ear of the observer. Then, if the two geophones and the connect- ing tubes are identical, they become listening devices by which the ear can detect the direction of the sotmd through the earth, in the same way that the direction of sotmd coming through the air can be detected. The geophones which were brought by the French Commission were very sensitive, but it seemed worth while to see if a still more sensitive instrtiment could not be produced, especially one which would more faithfully reproduce the character of the soimd which was received. Accordingly, a considerable number of instniments were constructed, using different sizes of instnunent, methods of suspension, kinds of case, sizes of tube running to the ear, and practically every other factor which cotild be readily varied. The resultant instruments were tested in dugouts, which were con- structed to simulate as far as possible actual field conditions. A standard source of sotmd was used and the results carefully collected. The tests of the apparatus constructed as a restilt of these investigations showed that it was appreciably better than the French instrument on which it was based and was also more con- venient to use. Seismomicrophone The seismomicrophone or telegeophone consists of a microphone which is attached to a geophone. However, a ntunber of modifi- cations can readily be introduced, so that the finished instnunent does not resemble the geophone. As the microphone is a very sensitive instrument, it is not necessary to use as large a mass as in the case of the geophone. In fact, an ordinary microphone to which is attached a small weight gives quite satisfactory results. The seismomicrophone can not readily be used for determining the direction of sotmd, as the binaural effect is not satisfactorily transmitted through a microphone. However, it is a very useful instrument in minmg operations. The complete outfit which was developed for the U. S. Army was lighter and more compact than that used by the French. Miscellaneous Publications, Bureau of Standards, No. 45 Fig. ^:;^. — I nstrmnent for detecting sounds transmitted through the earth Listening apparatus of this kind was developed to siicli a point of perfection that toward the end of the war scarcely any mining operations could be carried on without detection Fig. 34. — Print of radiograph of a repaired shoe One of a series taken for the purpose of studying the clinching of nails by different types of soling machines Miscellaneous Publications, Bureau of Standards, No. 46 Fig. 35. — Steel artillery wheel tested to destruction in the large Emery precision testing machine This illustrates only one of the thousands of physical tests of materials which the Bureau carried, out for the Army and Navy. The design of the wheel used on the Army's "Class B" motor truck was based upon a complete wheel investigation conducted at the Bureau SUBMARINE DETECTION 273 It was reported from the front that the development of listening apparatiis on both sjdes reached such a high standard that mining operations could be carried out only with the greatest danger to those concerned. Sotmds of digging, picking, and the like, could be detected at such a distance as to give ample time for the development of counter measures. Hence, mining operations very largely ceased along most parts of the front before trench warfare was abandoned. In some places, however, it was still found possible to carry on mining operations where soil conditions were such as to prevent effective counter measures. The instruments described in this report were used very little on the front, though the fact that they were available had impor- tant bearing on the conduct of mining operations. SUBMARINE DETECTION Means for Visual Detection Immediately after the United States entered the war, the detection of the submarine was perhaps the most pressing of all problems. In the course of time various means of detection were developed, some of them of a highly technical and compUcated natinre. At the outset it appeared that the simplest means of detection — viz, sighting the submarine or its periscope — should not be neglected but developed to its highest possible efficiency. The National Research Council appointed a committee, on which the Biu-eau of Standards was represented, to consider and make recommendations on this subject, in particular in regard to aids to vision, methods of observation, and development of the lookout service. The report of this committee was prepared at the Bureau and transmitted to the National Research Council on July 23, 1917. A supplemental report was issued on October 6 of the same year. Among the recommendations contained in these reports the fol- lowing aids to vision were called to the particular attention of the council as being of tmdoubted value and capable of practical application: (i) Goggles to protect the eyes from top, bottom, and side lights; (2) orange glasses, such as that known as "Coming G-36;" (3) binoculars of low power, large field, and (for twilight and night) highest possible brightness. Recommendations were likewise made as to the organization of the lookout service, with a view to aiding in the prompt detection of the submarine or its periscope. 26035°— 21 18 274 WAR WORK OF THE BUREAU OF STANDARDS TELEPHONE PROBLEMS Study of the Telephone Situation in the District of Columbia The extraordinary situation which confronted the telephone company in Washington as a direct and indirect result of the war was called to the official attention of the Public Utilities Commission of the District of Columbia through the company's formal petition for relief, filed November 30, 191 7. The size and activities of the Government departments grew by leaps and bounds; many new bureaus were created; tens of thousands of war workers were required; and several military camps were established in and around the District. All this resulted in an enormous and rapid increase in the demand for telephone service and facilities, both local and long distance. In addition, the demand for war workers seriously handicapped the telephone company in maintaining its operating staff. The telephone company petitioned the commis- sion for both traffic and financial relief. The Federal Government was, of course, vitally interested in prompt, accurate, and adequate telephone service. Under date of December 15, 191 7, the Public Utilities Com- mission formally requested that this Bureau be represented at the hearings. These were held during December, 191 7, January, February, and July, 191 8. The efforts of the Bvireau's telephone engineers were primarily directed to determining the facts in the case and to devising measures for temporary as well as permanent relief. On March 15 the Postmaster General, through the Secretary of Commerce, requested the Bureau of Standards to prepare a report on the local telephone situation, to include : 1 . A digest and analysis of the telephone situation in Washington as brought out in the hearings held up to that time. 2. The relative advantages and limitations of the manual, the semiautomatic, and full automatic systems of telephone. 3. The relative adaptability of manual, semiautomatic, and automatic central-office installations to prompt and economic enlargement. 4. The practicability of "physical connection" and manual systems, especially with reference to long-distance working. 5. An analysis of the methods proposed for immediate and for permanent relief of the local telephone situation, including com- parative investment data, maintenarice, and operating expenses. TELEPHONE PROBLEMS 275 The Bureau's report, subsequently published as Part II of H. R. Report 379, accompanying H. R. 10337, entitled "Extension of the Telephone System in the District of Columbia," was sub- mitted to the Postmaster General in April, 191 8. More than half of the report was taken up by a rather detailed analysis of the lo;al telephone situation as developed in the hearings. The materials of this section. Part I, which included a number of tables, was grouped under the following headings: Line and station data, Service, Traffic, Adequacy of equipment. Availability of automatic appliances, Revenues, expenses, and return on fixed capital, and Distribution of charges between the Government and public. In Part II was discussed (i) the relative ability of the manual, semiautomatic, and full automatic systems to handle traffic; (2) the quality of telephone service that might be expected of each system from the standpoint of reliability, accuracy, speed, supervision, continuity, and secrecy; and (3) the relative number of employees required for operation and maintenance for the three systems. No detailed discussion of the relative adaptability of manual, semiautomatic, and automatic central-office installations to prompt and economical enlargement was possible in this report because of the laclc of reliable data and the brief time available for the preparation of the report. Part III was therefore neces- sarily very brief. Part IV was also brief, the practicability of physical connection between the several systems having been already fully established. In Part V was discussed the telephone company's proposed method for immediate and permanent relief as outlined in the hearings, and there was also set forth in this section a method of relief proposed by the engineers of the Bureau. The company's plan was in brief the development of its plant through additional equipment of the same type already in use and the corresponding enlargement of its operating force. The method proposed by the Bureau for meeting, the immediate needs of the Government contemplated the installation of automatic equipment in the Government departments and independent establishments, beginning with the War Department, where the telephonic demands were greatest, by providing for the handlhig of interdepartmental business through an automatic switchboard to be located in the main central office, and the enlargement of 276 WAR WORK OP THE BUREAU OP STANDARDS the latter so as to permit connecting selected local subscribers responsible for the bulk of the incoming and outgoing Government traffic. By this means the Government's requirements would have been adequately and promptly met through what might be described as a telephone system within the local system. It was fiuther proposed to undertake a piecemeal conversion of the company's central offices from manual to automatic opera- tion, beginning with those having the largest number of subscribers and the largest volume of traffic. At the July hearings the experts of the Bureau were on the stand for direct and cross examination relative to the data set forth in the above report and in other exhibits submitted at the request of the commission. The hearings were adjourned on July 31 without any decision by the commission, whose jurisdiction in the case was terminated, for the time being at least, when the Federal Government took over the telephone systems of the country. In the meantime the company had acquired control over the situation, and the imexpectedly early terrtiination of hostilities in the following November entirely changed the aspect of the problem. TESTING MACHINES, CALIBRATION OF this subject is treated under the title "Calibration of Testing Machines." TEXTILES Cotton Fabrics The assistance of the Bureau of Standards was requested by various purchasing branches of the war organization in the summer of 191 7 in the development of methods of dyeing and of testing the various olive-drab cotton fabrics needed. The American textile industry was at that time just recovering from the shortage of colors which had followed the cessation of foreign importations. Supplies of dyestufEs were limited. Many of the fast colors which now again are obtainable were not to be had. Those on the market were not standard either in strength or in quality. The situation favored the use of inferior colors and unsuitable methods of dyeiag. The work undertaken by this Bvureau at the instance of the War Department is described below tmder the following heads: (i) Development of test methods; (2) study of dyestuffs by tests of laboratory dyeings ; (3) study of mechanical processes by frequent tests of samples from representative dyehouses. TBXTII.BS 277 In 191 7 the War Department was using test methods which had been adopted when foreign colors were available. These tests were exceedingly severe. The chemicals used included many which are never encountered in either the wearing or the launder- ing of textile materials. In one test, for example, a sample of the fabric was soaked for several minutes in concentrated hydro- chloric acid, in another a sample was given alternate treatment with potassium permanganate and sodium bisulphite, and finally a sample was steeped in bleaching-powder solution. These tests, operated to enforce the use of a fast hydrosulphite vat color formerly sold in this country by a German dyestuff factory, and only recently offered by American manufacturers. This Biureau believed that, tests which do not reproduce the conditions of wear tend to limit the dyer in his choice of colors with- out advantage to the purchaser, and that moreover they restrict development on the part of the color makers. Accordingly, it was recommended that simpler tests be adopted which should reproduce service conditions and that the tests adopted be rigidly applied. The recommendations of the Bureau were largely accepted. Prior to the recommending of these tests it was necessary to assemble data as to laundering methods and the expected life of garments and equipment in service. At one stage of the investi- gation a number of representative fabrics were sent to a Wash- ington laundry to be put through the regular washing process several times. The results were compared with those obtained under the application of the laboratory test methods. The sulphtir colors, which ultimately were tmiversally used in this work, differ more in their properties than is generally recog- nized. Particularly noticeable is their variation in fastness to light. Certain sulphur colors are so extremely fugitive that dye- ings from them fade seriously upon a few hours' exposm-e to bright svmlight. The good sulphitr blacks, on the other hand, are very fast to light. The fastness of the browns and olives varies considerably. The variety of sulphur colors which appeared on the market made it desirable to include a light fastness test in the specifications. Exposmre to sunlight is the natural method of testing. Unfortimately such exposvure takes several weeks and can not be carried out in a manner which is both fair and repro- ducible. 278 WAR WORK OP THE BUREAU OP STANDARDS Data on the performance of various lights sold for the testing of dyestuffs were obtained from the radiometric section, whose investigations of the nature of the light produced are described in Bureau of Standards Scientific Paper No. 330, and obtained also by comparing the action of the lights with that of sunlight on representative fabrics. The incorporation of a light test in the specifications was followed by a noticeable decrease in the num- ber of samples on which it was necessary to make an unfavorable report because of deficiency in this respect. The wood colors, such as cutch, can be used to obtain the desired shade for Army fabrics. ' The shades so produced are quite fast to washing and to light. Unfortunately they are changed by perspiration. Accordingly, it was necessary to incorporate in the specifications tests with organic acids to duplicate the effect of perspiration and to prevent the use of wood colors. Acetic acid was recommended on account of its general availability. At the request of the office of the Quartermaster General and the War Industries Board, a collection of all the available colors for dyeing the olive-drab shade on cotton was made. From these, small strips of cloth were dyed by following closely on a laboratory scale the directions of the manufacturers. These dyeings were tested carefully, and the test samples were then mounted on cards for easier comparison. The samples were tested for the effect of stmlight and of the dye-fading artificial lights, of washing and rubbing, and of acid, both before ahd after treatnient with the various metallic salts used to increase the fastness of sulphur colors. The cards, containing about 4000 samples, were much used in conferences with representatives of the War Department. They were of special value in that they afforded a ready means of demonstrating the defects of particular dyestuffs. They also showed the necessity of adopting proper methods of after treatment, and were of considerable value in interpreting the results of tests of commercial dyeings made at the Bureau. At the time of the signing of the armistice they were being used to determine whether or not further limitation of the supply of toluene allotted to color makers could be made by the substitution of certain dyestuffs not containing toluene or its derivatives without leading to the production of unsuitable colors. Cotton military fabrics are commonly dyed in the piece in the continuous machine. This may be operated so as to produce TEXTURES 279 excellent dyeings with the sulphur colors, or to produce dyeings in which the color is only superficially fixed. Other mechanical dyeing processes can also be operated so as to reflect the skill and care of the dyer. To determine the effect of variations in the speed of operation of dyeing machinery, in the placing and posi- tion of squeeze rollers, in the temperattu-e and composition of the dye bath, and in methods of after treatment and finishing, samples from representative mills were frequently tested in cooperation with the inspection division of the cotton-goods branch of the Quartermaster General's Ofiice. Tests of these samples led to the recommendation of certain processes and the criticism of contractors who were producing faulty material. In general, the fabrics which were shown by tests in 1918 to be deficient in fastness failed because of defects in the mechanical process of dyeing rather than because of the use of inferior colors. While engaged in this work the Bureau was called upon for information as to a large number of related topics, such as methods of redyeing garments and the stripping of tailor's cUppings for paper making. Woolen and Mixed Fabrics and Felts The expansion of the Army following the entrance of the United States into the war was followed by a great increase in the amount and variety of woolen materials purchased. This led to a large increase in the testing work requested of the Bureau. The work included the following: (i) Testing fastness of color; (2) studies of methods of dyeing; (3) determining fiber composi- tion of mixed materials. Colors for the production of fast olive-drab shades on wool are easy to make in comparison with the colors used on cotton and were available at the time the demand arose. The properties of the colors were also well known to dyers, as the dyestuffs were identical in composition with those which had formerly been imported. The test methods in use in 191 7 had been employed by the War Department for several years. They consisted of boiling in dilute-soap solution, boiling in dilute sodium-carbonate solution, and exposure to light. In the summer and early fall of 191 8, as part of the general movement toward the standardization of tests and their better correlation with service conditions, the Bm-eau recommended that the use of the boiling solutions be discontinued, on the ground that they are not generally used in washing woolen material. It was suggested that the use of luke- 28o WAR WORK OIC THE BUREAU OF STANDARDS warm solutions be substituted, and tha:t in the washing tests emphasis should be placed on the rubbing. Boiling solutions were, however, generally used throughout the war. Their use affords a quick method of demonstrating the presence of inferior dyes. The light tests were made in a similar manner to those used for cotton fabrics. It was customary to make tests of the fastness to laundering on rather larger pieces than were used for the cotton- fabric tests and to determine simultaneously the amoimt of shrinkage. A collection of the dyes used in producing the military shade on woolen fabrics was made and used for the preparation of dyed fabrics. The latter were tested and mounted in the same way as the dyed cotton fabrics and were applied to the same uses. The bichromate used for mordanting or fixating the wool colors becoming scairce toward the end of the war, a number of experi- ments were undertaken to determine the feasibility of substituting other mordants. These experiments were not completed, but in general it was found that the bichromate is so superior to other readily available mordants that it would be more wise to attempt to increase the supply of chrome ores than to attempt to substi- tute another material. Wool being considerably higher in cost than cotton, it is ad- visable in the manufactvire of certain types of woolen fabrics and felts to mix with it such an amount of cotton as can be used with- out resulting in diminished usefulness of the finished article. For example, saddle blankets were made for military uses with a cot- ton warp and a woolen filling. In this way the strength of the cotton yam was obtained without sacrificing the desirable char- acteristics of wool for this particular use. Similarly, in the manu- facture of felt considerable saving in cost may be made without loss of utility. Quite a variety of such mixed materials were purchased by the military branches, and a large number of samples were submitted to this Bureau for determination of the fiber composition. After some experimentation a convenient method of manipulation of the customary process of making such analyses, which consists, essen- tially, in dissolving the wool in boiling sodium-hydroxide solution, was agreed upon and recommended to the War Department for the guidance of inspectors. TEXTILES 281 Waterproofed Canvas Waterproofed canvas was purchased in large quantities by vari- ous biranches of the War Department. The conditions of use were so severe that it was necessary to make a. very careful inspection of material for purchase. Practically every manufacturer had a different formula or mechanical process. Waterproofed canvas for military uses must not only be thoroughly waterproof, but must also be resistant to mildew, free from objectionable odor or pois- onous material, permanently pliable in cold weather and after pro- longed exposure to sunlight, and reasonably cheap. The War Department requested in the fall of 1917 that the Bureau cooperate with the purchasing agents in establishing methods of testing and in selecting the material for purchase. Most of the methods of testing suggested in the literature on the subject were found to be too lenient. Others did not reproduce service conditions. Two methods of testing for waterproof prop- .erties were eventually agreed upon. The first was designed to reproduce the penetrating effect of beating rain., In this test a sample of the fabric was stretched over a wide-mouthed bottle, fastened in place, and subjected to a continued spray at consider-: able pressiu-e for several hottrs. As tarpaulins are at times spread upon the ground to protect materials place?! upon them from water, another test was used to dupHcate this condition. A gen- erous sample of the fabric was formed into a bag, filled with heavy articles, and placed in a pail of water for several days. The^e tests were found to be quite severe. To determine the resistance to mildew,, strips of the various fabrics were prepared for the tensile-strength machine and then suspended for several weeks in loosely covered jars in which there was a growth of mildew. The jars were kept in a warm, dark, and damp place. After the exposure the strips were cleaned and, ex- cept in the cases where the fabric had completely disintegrated, were tested for tensile strength. It was found that the mildew growths could spread over the siuface of some fabrics without injuring the fabric itself. The effect of the probable extremes of temperature was deter- mined by careful inspection of the material after heating for sev- eral hours at 45° C, and again while immersed in a freezing mixture. 282 WAR WORK OP THE BUREAU OP STANDARDS A number of representative fabrics were exposed to sunlight on the roof for several weeks. It was found that in some cases the waterproofing material was decomposed by the action of light. It was realized that the construction of the canvas used had an important bearing on the wearing qualities of the finished fabric and a number of samples were subjected^ to 'the regular analysis of cotton fabrics carried out by the textile section. An analysis of the coaiting material was made on about 150 samples. The materials found included asphaltum, wood tar, dry- ing oils, paraffin, resin, glue, petrolatum, starch, inorganic and organic antiseptics, and various inorganic weighting materials and pigments. Several ready-mixed waterproofing preparations were offered by manufacturers for use in recoating fabrics.. Many of these were analyzed and tested by appljring the usual tests to fabrics coated in the laboratory. An interesting feature of the study of waterproofed canvas was the examination of a number of French, EngUsh, and German fabrics obtained through the War Department. Few of these were as good as the average American canvas. Airplane Fabrics The Btireau conducted a series of experiments which led to the choice of mercerized cotton fabric of a certain construction as a substitute for the Hnen fabric employed as a covering for airplane wings used before the war and later unobtainable. In the routine examination of samples submitted for criticism by manufacturers, the Bureau cooperated by making the tests for amount of sizing and freedom from acidity or alkalinity, which were incorporated in the specifications. It was found iii the course of tensile strength determinations and oiher physical tests that the mercerized yams produced by various mills differed more than was to' be expected from varia- tions in their constriaction. It was believed that these differences were due to the various mercerizing processes employed. A study was undertaken of the effect of certain changes in the customary process of mercerizitig on the physical properties of mercerized yams. The results of this study will be published. Cordage Manila fiber is the best material of its class for the manufacture of ropes. Other fibers which are considerably cheaper do not resist rotting as well as manila and are not as strong. Conse- TEXTILES 283 quently, the mixture of these fibers with manila in the manufac- ture of rope is not only fraudulent adulteration, but may lead to serious accidents or the premature lessening of the utility of articles or appliances in which the rope is used. Because of its longer life, manila rope is economical in spite of its higher cost. Prior to 191 7 the use of the inferior fibers was rather common. In that year a test was devised elsewhere for distinguishing be- tween manila and the other fibers. The accuracy of this test was determined by making many analyses, under different conditions, of known mixtures of manila and nonmanila fibers. It was eventually incorporated in the standard specifications for manila rope issued by this Bureau for the guidance of manufacturers working on Government cordage contracts. Reports of the Bureau of Standards certifying to the presence of inferior fibers in rope sold as manila have been successfully upheld in the courts. A ntimber of rope samples were examined chemically for military organizations. (See article on "Miscellaneous Physical Tests.") Military Textile Equipment, TJoiforms, Blankets, etc. When this country entered the war the facilities of the military departments for the inspection and testing of military textile equipment were rather limited. The facilities of the Biureau were, therefore, made available to the Quartermaster Corps and the Ordnance Department. All of the materials ptu"chased by the last-named department were tested in the Bureau's laboratories, and this work constitutes a very large part of that carried out on militaiy textiles. The old Army fabrics and specifications covering them were entirely satisfactory, but the shortage of raw materials and manufacturing equipment made the use of substitutes necessary. It was definitely known that the substitution of structure alone would not necessarily duplicate an article, and the Bureau was called upon to devise test methods which could be used to predict the performance of proposed fabrics. The general procedure adopted was to study the old satisfactory fabrics to determine their performance and to compare the submitted fabrics with the results. The tests included the heat-retaining properties, resist- ance to abrasion, and resistance to repeated stress. As was the case in the design of airplane fabrics, the textile manufactiurers in this coimtry and abroad had never constructed clothing materials to have definite properties, and the Bureau experienced consider- able difficulty in determining the characteristics of raw stock and 284 WAR WORK OF THE BUREAU OF STANDARDS of mantif acturing which governed the exact properties previously mentioned. The opposition to the manufacturing of fabrics to meet definite requirements was so great that the Bureau installed a small experimental woolen manufacturing plant to demonstrate the value of such work. A great deal of valuable information was obtained in deter- mining the heat-retaining properties of materials, and it was con- clusively demonstrated that the old theory that cotton is not so warm as wool was entirely wrong and that the heat-retaining properties of any textile material depend entirely upon the man- ner in which the fibers are arranged. Many lines of possible research were thus thrown open to determine the best combinations of fibers and structures to make Army clothing lighter and more satisfactory than that commonly used; The study of the wearing properties of Army clothing was cen- tered about an investigation of the resistance of certain textiles to abrasion and to repeated stresses. In one or two cases labora- tory methods were checked up by actual performance tests, and here again much valuable information was obtained fpruse in constructing more satisfactory Army clothing materials. The information obtained through these military tests may be applied equally well to civilian clothing, and it seems reasonably certain that methods are available which will enable the manu- facturer as well as the Government to study more intelligently the characteristics of this class of goods.. The abrasion test is an attempt to reproduce by mechanical means the wear on fabrics received in everyday use. In this test the machine appears to gradually cut, break, and eliminate the fibers from tiie surface of the fabric until the loss greatly lessens the strength of the material. One of the chief difficulties pre- sented in this test is the lack of an abrader which will oiffer an even and hornogaieous abrading smiace at all times, so that each sample will receive the same treatment. The abrasion is measured indirectly by the resulting change in some other property, such as lowering of the tensile strength, or tear resistance, deteriora- tion in appearance, etc. All of the investigations for the standardization committee of the Director of Pxurchase and Storage, Quartermaster Depart- ment, including chemical and physical testing, were carried out in the textile laboratories of the Btu-eau of Standards, and the results have been incorporated in the standard specifications. TEXTH,BS 285 The section was in direct touch with the manufacturers of American dyes and had available at all titnes information as to the best way to dye materials to obtain the best results. Records were kept at all times which showed the exact performance of any American dyestuff. Information of this character was frequently supplied the military purchasing departments, and it is believed that the service rendered was of considerable value. Dyestuff Chemistry In addition to the work in dyestuff s alfeady described, war work in dyestuff chemistry of quite a different character was undertaken imder the direction of the colorimetry section. This is also described in the report on "Chromatic Camouflage." It is well known that solutions of individual dyestuffs differ widely in light-absorption properties. The spectroscopic exami- nation of light which has passed through such solutions affords a method for the identificatioh of colors. During the war period con- siderable progress was made in the projected development of a com- prehensive method for the identification and analysis of dyes by spectroscopic measurements. To aid in this work the chemistry laboratory contributed about 30 carefully purified dyes, made from purified intermediates by selected processes. The Ught- absorptive properties of dye solutions are found also in dyed fab- rics, films, colored glasses, paint pigments, leaves, flowers, and other colored objects. In the case of dyed fabrics the effect of viewing materials tmder different lights is so great that the dyer in matching shades must consider carefully the nature of the light under which a fabric will be used. By extending and applying, the principles involved it was found to be feasible to apply the light-absorptive properties of colored bodies to military uses. An example is a method by which signals may be conveyed to aviators by the use of fabrics dyed the color of grass. These, if carefully made, can not be distinguished from grass by the unaided eye at relatively short distances. When viewed through an appropriate color screen, however, the fabric will appear to be of quite a different color from the grass and thus will be distinguished at considerable distances. To try combinations which were suggested by the measure- ments of the colorimetry laboratory, it was necessary to dye a num- ber of fabrics and films. For this purpose use was made of the collection of dyes maintained by the chemistry laboratory and, where necessary, dyestuffs were synthesized from the intermediate 286 WAR WORK OP THE BUREAU OP STANDARDS compounds. Several well-matched fabrics and films were obtained which when viewed through the proper color screen were easily- discriminated. The results of these experiments were reported in detail to the military organizations interested. TIMEPIECES Information and Specifications From the beginning of the war the Bureau of Standards m its work on timepieces performed a valuable service which it was per- haps better qualified to undertake than any other existing agency. For a number of years the Bureau had carried on the investiga- tion and testing of high-grade pocket watches, and the experience thus secured formed an admirable preparation for the drawing up of specifications involved in the ptirchase of all classes of time- pieces. So far as is knoirai, the Bureau was consulted at least as to cer- tain details of the specifications for practically all of the time- pieces which were purchased by the Government during the war. For one year after the United States declared war on Germany, one member of the Bureau's staff served in an advisory capacity to the National Research Council on questions relating to procur- ing an adequate supply of serviceable navigating instruments for the new Govenatnent-owned merchant marine and later gave up his position at the Bureau to join the Emergency Fleet Corpora- tion, so that he might be able to devote his entire time to this one problem. Chronometers A particularly important phase of this problem consisted in the securing of marine chronometers or, what was the final outcome of the problem, the obtaining of an adequate supply of acceptable substitutes. At the begiiming of the war there were no chro- nometer factories in America, and the only available domestic source of supply was the local jewelers scattered throughout the country, from whom an exceedingly small number of chronometers were obtainable. Some could likewise be obtained from the British Admiralty. The production of new chronometers is not the work of only a few months, but involves years. The total supply accumulated from all sources satisfied only about one-third of the demand, and the remainder was met by using high-grade American-made watches called "ship watches." These were all tested, and each watch accepted for this service had to pass the rigid performance test adopted by the Bureau. TIMEPIECES 287 The Begiiming of a New Industry As has been the case in many other lines, the necessity for the use of watches as substitutes for chronometers has called the attention of the American people to this need, with the result that one of the leading watch manufacturers in this country has com- menced the making of marine chronometers. In all probability these will not be placed upon the market for some time, but it is to be hoped that eventually this will become another industry in which the United States will not be dependent upon a sister nation in time of public emergency. Testing of Timepieces The amount of routine testing performed by the Bureau has been exceedingly large, since all of the timepieces purchased hf the Emergency Fleet Corporation were given their acceptance test here. The timepieces tested for this organization include not only ship watches but also comparing watches and desk clocks. Since many of these instriunents were to be used for a purpose requiring special accuracy and because large numbers had to be rejected, this routine work was considered well worth while, even though each test covered about a month and the ntunber of timepieces tested amounted to approximately 15 000. Stop Watches The most important work of this section of the Bureau for the War Department was the aid furnished to the Ordnance Depart- ment in obtaining a supply of satisfactory stop watches. As stop watches are not made in America, the only ones available had to be imported, a!nd the errors of these instruments in practical use either influenced the accuracy of scientific data or jeopardized the safety of troops. The inspection or acceptance testing of these watches was conducted at the Bureau. The value and necessity of this work will be appreciated when one considers that even with specifications having only moderately close tolerances, the ntunber of rejections for the best types of watches was as large as 7 per cent, while for the poorest type it ran as high as 67 per cent. Airplane Clocks When the construction of airplanes on a large scale was first undertaken, practically the only available kinds of timepieces at all suitable for mounting on the instrument boards of airplanes were large>, heavy automobile clocks which were not sufficiently accurate to meet the requirements of this class of service. The Bureau, with the active cooperation of one of the leading watch 288 WAR WORK OP THE BUREAU OF STANDARDS manufacttirers, developed a suitable type 'of instrument having all the accuracy that was required even under the most adverse circumstances and for which the weight was only a small fraction of that of all known previously existing types. This new design was adopted by the Signal Corps, and, so far as is known by this Bureau, was used to the end of the war without a chaiige in any detail. A peculiar feature of this problem which was admirably taken care of by the manufacturer was that of the proper lubrica- tion of the movement so that the docks would not stop when exposed to the extremely low temperattue met with- at high altitudes. TOLUOL RECOVERY Preliminary Work, Conferences, etc. On April ii, 191 7, there was presented before the Pennsylvania Gas Association an article urging the attention of the gas industry to the matter of the recovery of toluol from city gas supplies, in order that there; might be available for the Federal Government the necessary quantities of tuluol for the production of T.N.T., which is the. preferred high explosive for military uses. Discus- sions on this subject between members of the Bureau's staff and a member of the Gas Institute likewise took place, and the subject appeared to be of such importance that during the remainder of April, 191 7, one of the Btureau's engineers made a careful study of :^ritish gas journals tp obtain information of interest to the Ameri- can ihdustry. When this work was completed, it was found that the information obtainable, while valuable, was not particiilarly applicable to similar problems which confronted American gas companies, bn May 2,3, and 4, representatives of the Bureau also took up with the engineering, physical, and chemical divisions of the National Research Council the question of securing by cooper- ative effort the maximum amoimt of information on this subject. As a result of a conference between the chairman of the American Gas Institute's committee on "The Relation of the Gas Industry to the Military Needsof the Nation, ' ' andrepresentatives of the Bureau, it was suggested that a committee be organized including representa- tives of the gas industry, of the Federal Government, ajid of State and city authorities, to undertake a comprehensiveinvestigation that would make available to the Government the following facts : ( i ) To what ejttent toluol could be secured from the American gas works ; (2) what procedure would be necessary in order to make this toluol availa- ble; (3) what changes in standards would be required in order that TOLUOL RECOVERY 289 the tuluol be removed from the gas without violation of local re- quirements; (4) what fmancial arrangements could be considered acceptable and proper to compensate the gas companies for the service rendered; and (5) what companies would be most available to undertake this work at an early date. As a result of this decision, the Bureau conferred with a number of agencies and endeavored in every way to render assistance look- ing toward the necessary readjustments of standards of gas service, elimination of high-candlepower standards, and substitution of heating values in general well below 600 Btu per cubic foot. Beginning in the latter part of May the Bureau received from several State commissions, city authorities, and gas company ofl&cials numerous inquiries on questions relating to toluol recovery and to the adjustment of standards under the existing abnormal conditions of supply of labor and materials. After consultation with representatives of all the parties inter- ested it was decided to hold a conference on the subject of toluol recovery on July 31 and August i and 2 at the Bureau. The three questions for discussion at the conference, as stated in advance of the meeting, were as follows : 1. ,The recovery of toluol from gas, including the methods and equipment required for such recovery, and various questions of operation, including standards, costs, 3aelds, and other interesting and pertinent information. 2. The extent to which the present costs of material and labor are affecting the gas-company operation, this to demonstrate to what extent the present situation is making necessary changes in existing standards or existing rates for gas apart from the question of toluol recovery. 3. The relative advantage of changing standards instead of changing rates for gas. During the meeting a general presentation of the need for some relief to gas companies was offered, and some suggestions were made upon the relative desirability of changing the standards for service instead of changing the price of gas to the customers. How- ever, no definite conclusions were reached at this session. A great deal of time was devoted to the discussion of the need for toluol and suggestions as to the best means for accomplishing its recovery. On the last day of the conference the State and city represent- atives, meeting by themselves, voted unanimously to adopt reso- 26035°— 21 19 290 WAR WORK OP THE BUREAU OP STANDARDS lutions appointing a joint committee to further investigate the subject, the chairman of the committee being a member of the Bureau's staff. ' The Bureau of Standards then undertook to inform itself more fully as to the current practices in the production of toluol from city gas supplies. For this piupose representatives of the Bureau visited a considerable number of plants where this work was being done and seciu-ed the maximum amount of information possible in the limited time available. Information was also sought by correspondence with others familiar with toluol practice who could not be met in conference. The information secured served as the basis for the report on "Recovery of Light Oils and the Refining of Toluol." The preliminary draft of this report was submitted to the mem- bers of the special gas committee on September 19, and thie first meeting to discuss this matter was held September 24. With^the assistance of the committee's suggestions the Bureau revised this repoi-t and issued it in mimeographed form on October 15. Report on Toluol Recovery This report covered the subject of toluol recovery in a general way and was divided into three parts: Part i, The technical rela- tion of the gas industry to the military needs of the nation; part 2, Pirinciples tinderlying benzol aiid toliiol recovery; part 3, Con- struction and operation of light-oil recovery plants. Part 1 gave a general discussion of the problem and described briefly the vari- ous methods used in manufacturing gas; jpart 2 described briefly the methods used in the recovery of toluol, while part 3 presented in detail the apparatus and equipment of one i:ype of toluol- recoyery plant installation. The adaptability of light-oil recbver^ by small gas plants was also discussed, and the cost Of installation and operation of light-oil recovery plants was touched upon. This paper was published in the Gas Age, of November 15, 191 7, the Gas Record, of November 14, 191 7, and the Journal of Indus- trial aiid Engineering Chemistry, ' of Jianuary, 1918. Through these publications the paper reached a large number Of persons interested in the work, and in addition the Bureau distributed ihany reprints to gas dompahies, State commissions, and individ- uals. A great many were asked fOr by the Ordnance Department of the Army and used in connection with its work. Additional Committee Work Following the publication of this paper a second meeting of the above-mentioned committee was held on October 22, at which TOIvUOL RECOVERY 29 1 time it was expected to discuss at considerable length plans for further work and particularly questions relating to the technical features of the proposed toluol contracts between the Govern- ment and gas companies. Preceding this committee meeting the Bureau had held a number of conferences with members of the Ordnance Department of the Army, suggesting certain principles bearing upon the question of toluol contracts, and the Bureau had forwarded to the department as a confirmation of these confer- ences a memorandum outlining certain principles which it sug- gested as worthy of consideration by the Ordnance Dejiartment in planning these contracts. On the same day (Oct. 20) a member of the Ordnance De- partment handed to a representative of the Bureau a copy of the outline of the plans of the former organization relating to the con- tracts for toluol and expressed his willingness that these matters be discussed by the committee two days later. These suggestions emananting from the Ordnance Department were therefore laid before the joint committee at its meeting on October 22. In connection, with the toluol-contract work done by the Ord- nance Department the Bureau of Standards was able to be of great assistance. Several members of the gas-engineering staff spent periods varying from a few days to six months acting as technical advisers of the Ordnance Department. Publications on Toluol Recovery and Gas Standards During: the early part of January, 1918, the need of a publica- tion giving in some detail the effect upon standards of gas quality of the removal of light oil became apparent. Accordingly, a brief r&iun^ was prepared of the standards of gas service then ia force in thoSe cities which would probably be considered for the erection of toluol recovery plants. The method of estimating influence of toluol recovery upon gas quality was described, and the following conclusions set forth. These conclusions repre- sented substantially the attitude of the Bureau on toluol recovery. As a stunmary of these points, the following suggestions are offered: I. Eliminate all candlepower requirements now in force, except for the cities where 18 candles or higher has been supplied, in which localities reach an understanding that at least 12 candles will be maintained for a period, say a year, during which time read- justments of appliances and substitution of mantle lamps would be accomplished to such an extent as to justify complete elimina- tion of candlepower regulations. 292 WAR WORK OF THE BUREAU OP STANDARDS 2. For plants making coal gas (or practically only coal gas) let the heating- value stand^d be from 550 to 570 British thermal units. 3. For plants making water gas, either alone or as a major constituent, let the heating- value standard be 570 to 600 British thermal units monthly average total heating value, the adjust- ment being made between these limits according to the economic conditions of operation. In order to make the conclusions clear and as far as possible, specifically applicable to all the cases in question, there was given in the paper a list of the localities where such changes of standards as are suggested would probably be necessary. This publication also had a wide circulation and was reprinted in the Gas Age, of January 15, 1918, the Gas Record, of January 9, 191&, and the Journal of Industrial and Engineering Chemistry of February, 191 8. In addition to reaching about 20 000 per- sons, a ntunber of reprints were distributed. Previous Publications Combined in Technologic Paper The supply of both of these publications was exhausted early in 191 8, so they were combined and published as Technologic Paper No. 117 of the Btureau of Standards, entitled "Toluol Recovery." Much additional matter was gathered, and changes which had been suggested were incorporated. This paper is in fotur parts. Part i is the introduction which deals with the need for toluol and the scope of the paper; part 2 is practically the otigiaal paper on "The Recovery of Light Oil and Refining of Toluol," previously described, but, in addition, contains a number of tables and other data assembled from va- rious sources and presented through the courtesy of the authors; part 3 is a discussion of the relation of toluol recovery to standards fpr gas service, and is the paper previously referred to as a separate publication; part 4 gives a typical form of contract which the Ordnance Department made with various companies for the opo'ation of toluol-recovery plants in coimection with city gas works. Increased Production of Toluol in the United States Notwithstanding the great importance of high explosives in the war the necessity for recovery from city-gas supplies of toluol, the base of trinitrotoluol (T. N. T.) was not appreicated in the United States until this coxmtry entered the conflict in 191 7. Pre- vious to 1914 the production of toluol in the United States prob- WHEEI.S, INVESTIGATION OF ARTILLERY, TRUCK, AND AIRPLANE 293 ably did not exceed 500 000 to 750 000 gallons per year. Most of this toluol came from coal-tar distillation, but smaller amotmts were prepared by refining the light oils which were removed from coke-oven gas in a few plants. After beginning hostilities the demand for toluol greatly increased, and it was evident that the common sources of this material would not meet the new commer- cial demands. As a consequence a large increase in the recovery of toluol from coke-oven gas immediately followed, and later the recovery of this constituent from city-gas processes was under- taken as mentioned in the preceding paragraphs. To give an approximate idea of the magnitude of the industry thus developed, one need only compare the small output of pure toluol in 19 1 2 and 191 3 of approximately 500 000 gallons with the probable output for the year 191 8. If the war had continued for the entire year, it was expected that the output would have exceeded 20 000 000 gallons, and by the end of the year 191 9 it was hoped that the rate of production would reach 30 000 000 gallons. At the time of the signing of the armistice there were 2 1 Govern- ment-owned toluol plants in operation in city gas works, a typical plant being illustrated in Fig. 32, and a number of other plants were nearing completion or were actually ready to begin operation. Practically all of the increase was made possible because of the recovery of toluol from coke ovens and city-gas supplies'. WHEELS, INVESTIGATION OF ARTILLERY, TRUCK, AND AIRPLANE The Bureau's superior testing facilities resulted in its being called on repeatedly in connection with development and stand- ardization work of various kinds. Among the investigations caixied out were three series of tests on wheels, namely, (i) artil- lery or gun-carriage wheels in cooperation with the Ordnance Department, (2) truck wheels in cbnjunction with the Motor Transport Corps, and (3) airplane wheels with the Bureau of Aircraft Production. In all cases the wheel manufactiurers worked hand in hand with this Bureau in the development work, and it may not be out of place to remark that this cooperation of private interests with the Government was typical of that experi- enced by the Bureau during the war. Wheel-Strength Requirements On all types of wheels the basic requirements are essentially the same. Such vital requirements are (i) resiliency or elasticity. 294 WAR WORK OP THE BUREAU OP STANDARDS (2) strenglii, (3) durability and economy, and (4) lightness. In the war emergency the military authorities found themselves without any reliable comparative data and there was no time for service tests. The Biureau of Standards was therefore called upon to make laboratory tests. : Types of Laboratory Tests Each type of wheel, for whatever purpose it was designed, was subjected to two tests. The first of these is the radial com- pression test in which the wheel is supported on a block and stressed radially by applying load through an axle passing through the hub. The second one is the side-thrust test in which the wheel is supported on an axle which is held rigidly with its axis parallel to the line of application of the applied load. The load is applied to the side of the rim of the wheel thi^ough a suitable bearing block. The parallelism of these two tests to the stresses encoimtered by the wheel in service is easily seen. The radial-compression test sub- jects the wheel to the same form of loading it receives in service when supporting the loaded truck at rest. The side-thrust test parallels the stresses introduced by the skidding or turning of the truck. Value of Tests The value of the radial-compression and side-thrust test in deter- mining the efficiency of the wheel in question, on the basis of the requirements outlined in paragraph 2 of this section, is readily apparent. The resiliency or elasticity is determined by plotting the deformations or deflections against corresponding loads and determining the area under the resultant curve. The elastic resiliency of the wheel reported in inch-pounds is taken as the area imder the part of the curve cut off by a vertical line through the proportional limit. The strength of the wheel is determined by carrying the tests outlined above to a point where fracture occurs, or, in some cases, only to a load which represents an adequate factor of safety when there is no sign of failure up to that point. Dura- bility and economy can not, of course, be determined in a laboratory test, except in so far as the minimum required weight represents the minimum cost for metal. Lightness, with strength, constitutes the final basis for comparison of various wheels. Elastic resilien- cies and ultimate strengths are all reduced for comparative pur- pose to a unit-weight basis. WHEELS, INVESTIGATION OF ARTILLERY, TRUCK, AND AIRPLANE 295 Artillery Wheels The Bureau subjected about six artillery or gun-carriage wheels to a radial-compression and side-thrust test (Fig. 35). These tests were conducted on the Bureau's 2 300 000 pound Emery hydraulic testing machine, as were the tests of the motor-truck wheels described below. Load was applied to a section of the rim mid- way between two spokes. These wheels were of cast steel, and most of them were about 155 mm tread by 1350 mm diameter, or 6 by 53 inches in size. The Bureau's tests on these wheels were not so definitely a part of a development program as were the motor- truck-wheel test series. However, certain elements of weakness, including defective welded joints, were discovered and comparisons between various manufacturers' products were furnished the Ordnance Department. Truck Wheels In designing the class B military truck, the Army foxmd it desirable to develop a metal wheel having the good quahties of all the motor-truck wheels on the market. The laboratory tests of these wheels were conducted at the Bureau. The wheels tested were made of cast steel, pressed steel, wrought iron, and wood, and by various manufacturers. The metal wheels included both solid and built-up designs. The series included examples of U spokes, tubular spokes, solid spokes, and disk wheels. The wood wheel included in the series had seen some service, but most of the metal wheels came directly to the Bureau from the factory or fotmdry. Radial-compression and side-thrust tests were made on each type of wheel, and the elastic resiliency, strength, and deforma- tions were determined. Using the ioformation obtained from these tests, a hollow spoke and rim cast-steel wheel was designed for the Quartermaster's Department (motor transport section), which was called the "composite wheel." This type of wheel demonstrated the value of the laboratory tests by giving satisfac- tory results in service. Technologic Paper on Wheels The restdts of the motor-truck wheel test series have been pub- lished as Technologic Paper No. 150. The results of tests on 21 wheels are given. The deformations of the wheels in the radial- compression tests were recorded for fotur poiats on the rim spaced 90° apart. Load-deformation graphs plotted from these data 296 WAR WORK O? THE BUREAU OP STANDARDS are given, from which the proportional limit and elastic resiliencies of the wheel are computed. The proportional limits are then divided by the weight of the wheel to furnish unit-weight figiu-es for comparison. The same program was followed in the side- thrust tests. The Bureau publication gives a photograph of each wheel and a chemical analysis of the material for each metal wheel. Valuable suggestions for use in the design of truck wheels to afford maximum striength and resilience for minimum weight are included. Unique Stress Analysis of Wheel In addition to the data determined for each type of wheel, as outlined above, a complete strain-gage analysis was made of one cast-steel wheel with hollow spokes and rim in a radial-compres- sion test. A 2-inch Berry strain gage was used. Deformations were measured in gage lines established on the rim, on the spokes, and on the hub. After the radial-compression test had been finished, specimfens were cut from the wheel and tested in tension. The average modulus of elasticity as determined from these tests (28 500 000 pounds per square iiich)' was u^ed to reduce the strain-gage data to corresponding stresses. From these data graphs (or stress contours) were plotted showing the stresses in the wheel at 48 gage lengths for loads on the wheel up to 60 000 pounds. Due to the short gage length (2 inches), results of great precision were not to be expected, but in this case the results were very satisfactory. Analysis Aids Efficient Design Compressive stresses were found in all parts of the rim except the section between the spokes adjacent to the loading block. That section was subjected to tensile stresses by the behavior of the spokes. The only excessive stress in the hub was at the junction of the two spokes adjacent to the loading block, where the tensile stress reached the proportional limit of the material for an applied load of 50 000 poimds. The stresses in the spokes were all less than the proportional limit of the material, except for two gage lines on one spoke adjacent to the hub. There is a possibility that these two gage lines were located at a thin section of the lasting. To the best of the Bureau's knowledge, this stress analysis by use of the strain gage is the first of its kind ever attempted in connectibri with wheel design. It is believed to afford a profitable line of study in connection with further wheel design. Only WHEELS, INVESTIGATION OK ARTILLERY, TRUCK, AND AIRPLANE 297 by such a stress analysis can the most satisfactory distribution of the material in the wheel be obtained. Such data allow the modi- fication of a wheel so as to avoid excessive stresses which cause failure in service tests. Airplane Wheels One or more each of eight types of shock-absorbing wheels for airplane landing gears were subjected to radial-compression and side-thrust tests in the manner heretofore described. As these wheels are, of course, designed for smaller loads than are the artil- lery and truck wheels, the tests were conducted on the 230 000- pound emery hydraulic testing machine. Palmer Wire-Spoke Wheel Most of the wheels tested were of the wire-spoke type. There were also included samples of disk-type wheels with spring unit rods and wheels with S-spring steel spokes. Partly on th^ basis of these tests the Bureau of Aircraft Production adopted the Pal- mer swedge-spoke type of wheel as the standard construction. These wheels, size about 750 mm diameter by 125 mm tread, weight 8.7 to 9.2 pounds, averaged about 3000 pounds ultimate strength in side thrust and 7000 pounds in radial compression. These figures are based on tests conducted at the Bureau on about a dozen Palmer tjpe wheels from four manufacturers. The average results of the above tests enable the designer to establish a fair criterion for judging the performance of any other wheel under similar test conditions. The ultimate strength in side thrust for unit weight averaged 355 for five Palmer type wheels tested, and the corresponding ratio of strength to weight in radial compression was 810. These two values can be taken as fair criterions for judging the value of other types of wheels for this service. Comparison with Early American Wheel By way of comparison it may be noted that the strength-weight ratio for S-spring spoke wheels was found to be 160 in side thrust and 380 in radial compression; for the spring unit rod spokes with sheet-steel casing, 40 in side thrust and 430 in radial compression; average of two early American type wire-spoke wheels, 255 in side thrust and 820 in radial compression. The Palmer type wheel has a hub which is not placed symmetrically in the plane of the rim, but which gives added strength to resist the side thrust encoimtered by the wheels and chassis when the plane lands. Elasdc-resiUence comparisons also showed the relative superiority 298 WAR WORK OP THE BUREAU OP STANDARDS of the Palmer type wheel. This wheel was extensively used for equipping airplanes manufactured in England before these tests demonstrated its superiority over other designs and resulted in its general adoption for the use of American-made aircraft. X-RAYS Owing to the fact that the X-ray work of the Bureau was not initiated until the summer of 1 9 1 7 , the Bureau was not in a position to be of as much assistance to the military auth9rities in this field as it would have been had the work beeri started at an earlier date. The work done falls under two heads: (i) The study and testing of protective materials; (2) the radiography of various objects. The work included under the second head was strictly of a cooperative nature, the radiographs being taketi for the purpose of furnishing to the respective branches of the War Department ihforniation regarding the internal structure of specific articles and tne applicability of radiography to the examination or inspec- tioti of supplies. Protective Materials Lead-impregnated materials (glass, rubber, etc.) are widely used for protecting the roentgenologis,t and the patient from harmful exposure to X-rays. The protection of the roentgenol- ogist, whp under war conditions had to work many hours a day in close proximity to the X-ray tube, was especially important. A study of these commercial materials dempnstrated not only that they differed widely in the protection afforded, but that some pf them, even the most expensive, were practically worthless, and that all might be imprpyed in quahty. Radiographs were takeij of variou? pieces of lead glass. In one case it was found that of two pieces tested, the one casting the denser shadow, while but 3 mm thick and weighing only 2 pounds per square foot, was of a good qujality of lead glass, the other glass, 25.4 mm tlnck and weigh- ing 12)^ poimds per square foot, was practically worthless for X-ray protection. This piece was, however, representative of some of the most expensive glass that was bought by the Bureau in the open market in 191 7. It was also representative of some of the glass that was offered to the Surgeon General's Office. The Bureau advised the Surgeon General's Office and the larger dealers in X-ray supplies of the presence of this fraudulent material, and urged manufacttu-ers to produce material of a tetter quality than any that was on the market in the spring of 1917. The X-RAYS 299 response of the manufacturers was most gratifying, and materials of superior quality are now available. In connection with this work protective materials were tested for the Surgeon General's Office, inspectors of that office were furnished with data by means of which approximate information concerning the quality of these materials could be determined expeditiously by the simple process of measuring and weighing the specimens, and improved supplies were brought to the atten- tion of military authorities Radiography At the request of the reclamation division of the Quartermaster Corps of the Army repaired shoes (Fig. 34) were radiographed for the purpose of obtaining information regarding the driving and clinching of nails by different types of soling machines. Aluminum propellers for airplane apparatus were examined for the presence of internal flaws. Thin specimens of steel were similarly examined for the Ord- nance Department of the Army. At the request of the Bureau of Aircraft Production the prac- ticability of applying radiography to the inspection of wood was investigated in some detail. Although tmder suitable conditions radiographs can give much information regarding internal struc- ture, it appears that in the case of xmfabricated wood it can reveal but little of importance that can not be detected by a trained inspector. In the case of finished articles, however, it can in many cases yield valuable information regarding the fabrication, information that can be obtained in no other way. For example, in one piece of wooden construction examined it was evident that certain wooden pins had not been driven home. The radiography of a number of articles of an impromising nature, such as censored manuscripts, charcoal, etc., was attempted at the request of the military authorities interested. Although the possibility of successfully radiographing these articles appeared slight, the Biureau studied each case thoroughly, using various techniques before giving up the problem as hopeless. Somewhat similar to the preceding was the attempted radio- graphy of thick specimens of steel. No satisfactory result was obtained either by the use of X-rays or of radium.