.:^ona:l a/ja ; . 
 
 'a. ■ . 
 
 •;■■’. :/w'jtm : / : : :AA;AA'; ': 
 
 M )p%M 
 
 : . ../USED DY ' A. A • A' A 'A; A A';’ 
 
 M : MfAEmwr : 
 
 mrnmm 
 
FraNK^iN Institute Liibrary 
 
 r 
 
 PHILADELPHIA 
 
 Class &<o<o- SBook Un 3 . 3 . Accession 
 
 REFERENCE 
 
 Given by the United States Government 
 


 
 
 

 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
No. 28. 
 
 PROFESSIONAL PAPERS 
 
 OF THE 
 
 CORPS OF ENGINEERS OF THE UNITED STATES ARMY. 
 
 PUBLISHED BY AUTHORITY OF THE SECRETARY OF WAR. 
 
 Headquarters, Corps of Engineers. 
 
 1901 . 
 
PROFESSIONAL PAPERS OF THE CORPS OF ENGINEERS, U. S. ARMY. 
 
 No. 28. 
 
 REPORT 
 
 BOARD OF ENGINEER OFFICERS 
 
 ON 
 
 TESTING HYDRAULIC CEMENTS, 
 
 WITH 
 
 SPECIFICATIONS FOR THE SEVERAL CLASSES 
 
 USED BY' 
 
 THE ENGINEER DEPARTMENT. 
 
 Major WILLIAM L. MARSHALL, 
 
 Major SMITH S. LEACH, 
 
 Captain SPENCER COSBY, 
 
 Corps of Engineers, U. S. Army, Members of the Board. 
 
 WASHINGTON: 
 
 GOVERNMENT PRINTING OFFICE. 
 
War Department, 
 Document No. 144. 
 
 Office of the Chief of Engineers. 
 
Office of the Chief of Engineers, 
 United States Army, 
 
 Washington , June 11 , 1901. 
 
 Sir: I have the honor to transmit herewith a report of a 
 Board of Officers, convened under your authorit} 7 ^ by Special 
 Orders, No. 56, Headquarters Corps of Engineers, 1900, to 
 consider and report upon the general subject of manipulating 
 and testing hydraulic cements used in the public works car- 
 ried on under the supervision of this office. 
 
 The Board has given the subject careful study, and its 
 report, which is accompanied by a proposed draft of specifi- 
 cations, is recommended for approval with a view of securing 
 uniform practice as far as possible throughout the Engineer 
 Department. 
 
 As the report contains much information that is valuable, 
 not only to the Corps of Engineers, but also to the engi- 
 neering profession at large, I further recommend that 1,000 
 copies be printed at the Government Printing Office and fur- 
 nished for the use of this office on the customary requisition. 
 
 Very respectfull} T , your obedient servant, 
 
 G. L. Gillespie, 
 
 Brig. Gen ., Chief of Engineers, 
 
 U. S. Army. 
 
 Hon. Elihu Root, 
 
 Secretary of War. 
 
 [First indorsement.]. 
 
 Approved as within recommended. 
 
 By order of the Secretary of War: 
 
 John C. Scofield, 
 
 Chief Clerk. 
 
 War Department, June 19, 1901. 
 
 5 
 
LETTER OF INSTRUCTIONS. 
 
 Headquarters, Corps of Engineers, 
 United States Army, 
 
 Washington , December 17 , 1900. 
 
 Major: In view of the lack of uniformity that exists in the 
 specifications for cements submitted to this office for approval, 
 it is considered desirable that a uniform set of specifications 
 for each class of cement should be prescribed for use in works 
 under the supervision of this office. 
 
 The Board will therefore consider the general subject of 
 testing cements and make such recommendations on that sub- 
 ject as will be applicable under the conditions prevailing on 
 the works carried on under the supervision of officers of the 
 Corps of Engineers. 
 
 The Board in its report will follow in a general way the 
 report of the Board on Electrical Installations, recently sub- 
 mitted, and, after its remarks and recommendations, will 
 include as appendixes to the report sets of standard specifica- 
 tions for both Rosendale and American Portland cements 
 which shall embrace all the requirements necessary to secure 
 a suitable cement. 
 
 It is not expected that the Board will introduce unnecessary 
 refinements in the tests required, but will bear in mind the 
 actual conditions that obtain in the Engineer Department. It 
 is desired, however, that the report be complete and authori- 
 tative along the line suggested. 
 
 The subject of slag cements has recently been reported on 
 by a Board of Engineer Officers, but without any recommen- 
 dations as to the conditions under which they may safely be 
 used. The Board is therefore requested to consider this sub- 
 ject, and if in its judgment its use is considered admissible, 
 it is requested to state such conditions and to prepare a set of 
 standard specifications for slag cements. 
 
8 
 
 LETTER OF INSTRUCTIONS. 
 
 In view of the possible econom3 r resulting to the Govern- 
 ment from its grinding together cement and silica sand, it is 
 requested that this subject be also given consideration by the 
 Board. 
 
 The Board is authorized to call upon the District engineer 
 officers for such records of tests or such other information as 
 may be available. 
 
 Very respectfully, 
 
 John M. Wilson, 
 
 Brig. Gen., Chief of Engineers, 
 
 IT. S. Army. 
 
 Maj. W. L. Marshall, 
 
 Corps of Engineers, 
 
 President of Board on Specifications for Cements, 
 
 New York City. 
 
REPORT. 
 
 United States Engineer Office, 
 
 New York City , June 6 y 1901. 
 
 General: The Board of Officers of the Corps of Engi- 
 neers, constituted by Special Orders, No. 56, Headquarters 
 Corps of Engineers, December 17, 1900, has the honor to sub- 
 mit the following report: 
 
 The Board convened at the Army Building, in New York 
 City, upon the call of the senior member, at 10 a. m., Jan- 
 uary 8, 1901, all the members of the Board being present. 
 At sessions held on that and the following da} T the Board 
 arrived at a substantial agreement upon the leading points of 
 the subject referred to it, and upon the general wording of 
 the specifications for the several classes of cement, and then 
 adjourned to meet at the call of the senior member for the 
 completion of the report. 
 
 Under the authority of Special Orders, No. 3, Headquarters 
 Corps of Engineers, January 12, 1901, the Board reconvened in 
 New York City, at the call of the senior member, at 10 a. m. , 
 March 28, 1901. During a session of that day the views of the 
 several members, presented in writing, were carefully consid- 
 ered, and the Board agreed upon a report and specifications. 
 On April 2 these were sent to the Chief of Engineers, but 
 were returned to the Board with the request that it consider 
 the additional subjects of sampling the cement, amount of 
 testing desirable depending on the importance of the work 
 and the size of the order, and the records that should be kept. 
 
 At the call of the senior member, and b} T authority of Spe- 
 cial Orders, No. 20, Headquarters Corps of Engineers, May 
 15, 1901, the Board reassembled in New York City at 10 a. m., 
 May 29, and formulated its report upon the additional sub- 
 jects referred to it, the members having previously communi- 
 
 9 
 
10 
 
 TESTING HYDRAULIC CEMENTS. 
 
 cated their views to one another in writing. The additional 
 subjects have been incorporated in the following report, which 
 is accompanied by three sets of proposed specifications: 
 
 GENERAL CONSIDERATIONS. 
 
 The constructing engineer is confronted by no problem 
 more difficult than to decide whether a certain cement, when 
 placed in a work, will behave in a predetermined way. This 
 is especially true of Portlands. Other cements are much 
 more reliable under conditions of exposure for which they are 
 suited. 
 
 The difficulties arise from the fact that tests for acceptance 
 or rejection must be made on a product not in its final stage. 
 A cement, when incorporated in masonry, undergoes for 
 months chemical changes in the process of setting, so that the 
 material subjected to strains in the work is not the material 
 tested, but a derivative of it. The object of tests is to estab- 
 lish two probabilities: First, that the product of the given 
 cement will develop the desired strength and hardness soon 
 enough to enable it to bear the stresses designed for it; sec- 
 ond, that it will never thereafter fall below that strength 
 and hardness. Up to the present time it appears that the 
 relation between the chemical and physical properties of raw 
 cement and of its partially indurated derivatives, determined 
 by tests, and the physical properties of the same cement or 
 its derivatives, after complete hydration and induration in the 
 work, can be stated only within rather wide limits. 
 
 The most useful tests of cements are those, first, which con- 
 nect themselves definitely with some serious defect to which 
 cements &re subject, or with some merit which they should 
 possess; second, which can be made with the least apparatus 
 and manipulation, and which give their indications in the 
 shortest time; and, third, which are freest from personal 
 equation and from influences of local surroundings. These 
 criteria, applied to the customary tests of cements, give indi- 
 cations as to their relative value and the best methods of 
 making them. 
 
 TEST OF GRINDING. 
 
 This test derives importance from the fact, apparently well 
 established, that, other things being equal, the finer the 
 
TESTING HYDRAULIC CEMENTS. 
 
 11 
 
 cement the greater will be its sand-carrying capacity; that is, 
 it will show greater strength with the same charge of sand, 
 or equal strength with a greater charge. According to the 
 best information the Board can obtain, the cementitious value 
 of this material is believed to reside principally, if not 
 wholly, in the very fine part. It follows that a grinding test 
 should be directed to determining the proportion which is 
 very fine rather than the residue above a certain size. The 
 Board does not propose any change in the accepted grinding 
 test of Portland cement, but favors for Natural cement the 
 use of the same size screen as for Portland, No. 100, with the 
 requirement that 80 per cent shall pass through it. The 
 screen should be frequently examined, magnified, if practi- 
 cable, to see that no wires are displaced leaving apertures 
 larger than the normal. 
 
 TEST FOR SPECIFIC GRAVITY. 
 
 This test is made with simple appliances, and its result is 
 immediately known. It appears to connect itself quite defi- 
 nitely with the degree of calcination which the cement has 
 received. The higher the burning, short of vitrification, the 
 better the cement and the higher the specific gravity. 
 
 This test has another value, in that the adulterations of 
 Portland cement most likely to be practiced and most to be 
 feared are made with materials which reduce the specific 
 gravity. The test is therefore of value in determining a 
 properly burned, nonadulterated Portland. If underburned, 
 the specific gravity may fall below 3; it may reach 3.5 if the 
 cement has been overburned. No other hydraulic cement is 
 so heavy in proportion to volume, Natural cement having a 
 specific gravity of about 2.5 to 2.8 and Puzzolan (slag) of 
 about 2.7 to 2.8. Properly burned Portland, adulterated with 
 slag, will fall below 3.1. 
 
 TEST OF ACTIVITY. 
 
 This test, made by gauging the cement with water and 
 observing the times of initial and permanent set, is partly 
 direct and partly indirect. It is direct in so far as its limits 
 relate to the time necessary to get the cement in place after 
 mixing, which must not be greater than the time of initial 
 set, and to the time within which the cement product must 
 
12 
 
 TESTING HYDRAULIC CEMENTS. 
 
 take its load, which must not be less ,than the time of perma- 
 nent set. It is indirect in so far as its limits relate to the 
 probable final strength, elasticity, and hardness of the cement 
 mixtures. In the latter respect it appears to be reasonably 
 well established that cements exhibiting great activity give, 
 after long periods, results inferior to those with action less 
 rapid. 
 
 The test for activity is easily made with simple appliances, 
 and its results are known in a few hours at most. Variable 
 results in the test are caused by different local conditions of 
 moisture and temperature and by the different judgments 
 of observers as to whether the needles penetrate or not. Gen- 
 erally speaking, both periods of set are lengthened by increase 
 of moisture and shortened b} T increase of temperature. Some 
 manufacturers claim that their cements show their best results 
 when gauged with particular percentages of water. It is not 
 considered good policy to encourage these peculiarities at the 
 expense of the uniformity of tests which is so greatl} T desired. 
 It is better to adopt a definite proportion of water for gauging 
 and require all cements of the same class to stand or fall on 
 their showing when so gauged. Such a percentage, adopted 
 and known, will probably be used by manufacturers in testing 
 goods sold to the Engineer Department, and a greater har- 
 mony between mill and field tests of the same cement will 
 result. 
 
 In gauging Portland cement in damp weather the samples 
 should be thoroughh T dried before adding water. This pre- 
 caution is not deemed necessaiy with Natural cement. Suf- 
 ficient uniformity of temperature will result if the testing 
 room be comfortably warmed in winter and the specimens be 
 kept out of the sun in a cool room in summer, and under a 
 damp cloth until set. 
 
 TEST FOR CONSTANCY OF VOLUME. 
 
 This test results from observations made on the pats or 
 cakes used in the setting test. It derives its value from its 
 connection with the quantit} r of expansives in the cement. 
 
 The test is easy to make, and its results are relative^ free 
 from personal error, though there is room for a difference of 
 judgment as to the appearance of the cakes. As they may 
 be preserved and the decision reviewed at any time on the 
 original data, such differences are immaterial. 
 
TESTING HYDRAULIC CEMENTS. 
 
 13 
 
 TESTS OF STRENGTH. 
 
 These may be subdivided into compressive and tensile tests, 
 the latter including the transverse test made by breaking a 
 beam of the cement. The compressive test need not be fur- 
 ther considered, as it is less easily made than the tensile test 
 and gives no surer indications. The ratio of compressive to 
 tensile strength of the same class of cements is quite uniform. 
 
 Of the tensile tests the direct pull is preferable to the 
 flexure test. 
 
 The tensile test is theoretically a perfect index of the quality 
 of the cement at the periods of test, and a comparison at dif- 
 ferent periods gives the best obtainable indication of what its 
 subsequent conduct will be. In the opinion of the Board the 
 two periods most generally adopted, seven and twenty-eight 
 days after mixing, are, on the whole, the best. The one-day 
 test, though of some value in a discriminating sense, should 
 not be placed in the same category as the other periods named. 
 
 The apparatus for tensile tests is somewhat elaborate and 
 delicate, but is of standard manufacture and readily obtainable 
 at relatively small cost. 
 
 In respect of uncertainties due to the personal equation of 
 the tester and to the influence of local conditions this test pre- 
 sents greater difficulties than any of the others considered. 
 The most scrupulous care must be observed in the manipula- 
 tions, and the tester should possess natural aptitude for such 
 work. The object is to determine the greatest stress per 
 square inch which the cement can be made to stand under 
 given conditions without rupture. If the conditions have 
 been carefully observed and several discrepant results are 
 obtained, the highest may be right, but the others are cer- 
 tainly wrong. No averaging should be done. 
 
 The remarks made above under the activity test as to the 
 relation between early hydraulic intensity and the final excel- 
 lence of a cement product are equally applicable to the indica- 
 tions from tensile tests. A cement which tests moderately 
 high at seven days and shows a substantial increase to twenty- 
 eight days is more likely to reach the maximum strength 
 slowly and retain it indefinitely with a low modulus of elas- 
 ticity than a cement which tests abnormally high at seven 
 days with little or no increase at twenty-eight days. 
 
14 
 
 TESTING HYDRAULIC CEMENTS. 
 
 ACCELERATED TESTS. 
 
 The rules recommended by the committee of the American 
 Society of Civil Engineers in 1885 have been substantially 
 accepted here and abroad as to tests of setting qualities and 
 soundness; more rapid tests for soundness are, however, pro- 
 posed and practiced, though no accelerated test has been gen- 
 erally accepted. 
 
 Accelerated tests proposed for the speedy detection of the 
 presence of expansives in cement usually consist in the appli- 
 cation, after gauging, of dry heat or of immersion in warm or 
 boiling water or steam. The immersion tests are most in 
 vogue. Thej' vary from immersing freshly gauged pats on 
 glass plates in water at 115° F. for twenty-four hours, or at 
 higher temperatures for various periods, to steaming or boil- 
 ing cakes or cylinders of the material to be tested at 212° F. 
 for varying times. 
 
 In France and Germany the swelling or expansion of boiled 
 cylinders is measured directly by calibration. Usually change 
 of volume not accompanied by visible evidences of it — i. e. , dis- 
 tortion or disruption — is not observed in American tests pre- 
 scribed in specifications for the reception of cements. Of all 
 these tests the boiling test is the simplest , requires onl} r appa- 
 ratus everywhere available, and is recommended b} r the Board. 
 It has been the experience that this test detects material that 
 is unsound by reason of the presence of active expansives; 
 but in some cases it rejects material that would give satisfac- 
 tory results in actual work and will reject material that would 
 stand this test after air slaking. 
 
 The great value of the test lies in its short-time indications 
 and in at once directing attention to weak points in the cement 
 to be further observed or guarded against. Of two or more 
 cements offered for use or on hand, the cements that stand the 
 boiling tests are to be taken preferably; it should be con- 
 stantly applied on the work among other simple tests to be 
 noted, for although the boiling test sometimes rejects suitable 
 material, it is believed that it will always reject a material un- 
 sound by reason of the existence of active expansives. Sul- 
 phate of lime, while enabling cements to pass the boiling tests, 
 introduces an element of danger. 
 
TESTING HYDRAULIC CEMENTS. 
 
 15 
 
 This test is proposed as suggestive or discriminative only. 
 Except for works of unusual importance it is not recommended 
 that a cement passing the other tests proposed shall be rejected 
 on the boiling test. 
 
 TESTS TO BE MADE. 
 
 For selecting Portland and Puzzolan cement from among 
 the brands offered, the Board recommends that the following 
 tests be made: 
 
 1. For fineness of grinding. 
 
 2. For specific gravity. 
 
 3. For soundness, or constancy of volume in setting. 
 
 1. For time of setting. 
 
 5. For tensile strength. 
 
 For Natural cement we recommend the omission of the 
 specific-gravity and soundness tests. 
 
 On the works the Board recommends simple tests when the 
 more elaborate tests can not well be made. 
 
 In determining the minimum requirements for cements 
 given in the subjoined specifications we recognize that many 
 cements that attain only fair strength neat and with sand in a 
 short time and show marked gains of strength on further 
 time will fulfill the requirements of the service, and that unusu- 
 ally high tensile strength attained in a few days after gaug- 
 ing is often coupled with a small or negative increase in 
 strength in further short intervals. Unusually high tests in 
 a short time after gauging should be regarded with suspicion, 
 although some well-known brands of American cements show 
 great strength in short-time tests and, so far as observed, are 
 reliable in air and fresh water. Cements offered under such 
 known brands should show their characteristic strength and 
 other qualities or be suspected as spurious or adulterated, if 
 not rejected, even though the minimum requirements of the 
 specifications are met. The practice of offering a bonus or 
 free gift of money in addition to the contract price for cement 
 testing above a fixed high point should be prohibited as un- 
 necessary, for cements so obtained are likely to be unsound 
 in a manner not easily detected in the time usually available 
 in testing. 
 
16 
 
 TESTING HYDRAULIC CEMENTS. 
 
 It is believed that most of the very high testing Portland 
 cements have lime in excess, the effect of which is tempo- 
 rarily masked by the use of sulphate of lime. Overlimed 
 cements so treated are unfit for use in sea water. For such 
 uses a chemical analysis should be required, and the quantity 
 of sulphuric acid, as well as magnesia, be limited to a low per- 
 centage. It is not yet known that sulphate of lime in quan- 
 tity less than 2 per cent is injurious to cements to be used in 
 fresh water or in air. It masks expansives .that might ulti- 
 mately cause the destruction of the work, but it is not known 
 whether this effect is permanent. Its addition is now deemed 
 necessary to control time of setting. It makes a quick-setting 
 cement slow setting, at the same time increasing tensile 
 strength acquired in short time. 
 
 MANIPULATION OF CEMENTS FOR TESTS. 
 
 I. Fineness. 
 
 Place 100 parts (denominations determined by subdivisions 
 of the weighing machine used) by weight on a sieve with 100 
 holes to the linear inch, woven from brass wire No. 40, Stubb’s 
 wire gauge; sift by hand or mechanical shaker until cement 
 ceases to pass through. 
 
 The weight of the material passing the sieve plus the weight 
 of the dust lost in air, expressed in hundredths of the original 
 weight, will express the percentage of fineness. In order to 
 determine this percentage the residue on the sieve should be 
 weighed. 
 
 It is only the impalpable dust that possesses cementitious 
 value. Fineness of grinding is therefore an essential quality 
 in cements to be mixed with sand. The residue on a sieve of 
 100 meshes to the inch is of no cementitious value, and even 
 the grit retained on a sieve of 40,000 openings to the square 
 inch is of small value. The degree of fineness prescribed in 
 these specifications (92 per cent) for Portland through a sieve 
 of 10,000 meshes to the square inch is quite commonly attained 
 in high-grade American cements, but rarely in imported 
 brands. On the Pacific coast, where foreign cements only 
 are in the market, this requirement may be lowered for the 
 present to 87 per cent on No. 100 sieve. 
 
TESTING HYDRAULIC CEMENTS. 
 
 IT 
 
 . II. Specific gravity. 
 
 The standard temperature for specific gravity determina- 
 tions is 62° F., but for cement testing temperatures may 
 vary between 60° and 80° F. without affecting results more 
 than the probable error in the observation. 
 
 Use any approved form of volumenometer or specific grav- 
 ity bottle, graduated to cubic centimeters with decimal sub- 
 divisions. Fill instrument to zero of the scale with benzine, 
 turpentine, or some other liquid having no action upon cements. 
 
 Take 100 grams of sifted cement that has been previously 
 dried by exposure on a metal plate for twenty minutes to a 
 diy heat of 212° F. , and allow it to pass slowly into the fluid 
 of the volumenometer, taking care that the powder does not 
 stick to the sides of the graduated tube above the fluid and 
 that the funnel through which it is introduced does not touch 
 the fluid. 
 
 Read carefully the volume of the displaced fluid to the near- 
 est fraction of a cubic centimeter. Then the approximate 
 specific gravity will be represented by 100 divided by the 
 displacement in cubic centimeters. 
 
 The operation requires care. 
 
 III. Setting qualities and soundness. 
 
 The quantity of water and the temperature of water and 
 air affect the time of setting. The specifications contemplate 
 a temperature varying not more than 10° from 62° F., and 
 quantities of water given herein : 
 
 For Portland cements use 20 per cent of water. 
 
 For Puzzolan cements use 18 per cent of water. 
 
 For Natural cements use 30 per cent of water. 
 
 Mix thoroughly for five minutes, vigorously rubbing the 
 mixture under pressure; time to be estimated from moment 
 of adding water and to be considered of importance. 
 
 Make on glass plates two cakes from the mixture about 
 three inches in diameter, half an inch thick at middle, and 
 drawn to thin edges, and cover them with a damp cloth or 
 place them in a tight box not exposed to currents of dry air. 
 At the end of the time specified for initial set apply the needle 
 
 1593—01 2 
 
18 
 
 TESTING HYDKAULIC CEMENTS. 
 
 one-twelfth of an inch diameter weighted to one-fourth of a 
 pound to one of the cakes. If. an indentation is made the 
 cement passes the requirement for initial setting, if no inden- 
 tation is made by the needle it is too quick setting. At the 
 end of the time specified for “ final set” apply the needle one 
 twenty-fourth of an inch diameter loaded to one pound. The 
 cement cake should not be indented. 
 
 Expose the two cakes to air under damp cloth for twenty - 
 four hours. Place one of the cakes, still attached to its plate, 
 in water for twenty-eight days; the other cake immerse in 
 water at about 70° temperature supported in a rack above the 
 bottom of the receptacle; raise the water gradually to the 
 boiling point and maintain this temperature for six hours and 
 then let the water with cake immersed cool. Examine the 
 cakes at the proper time for evidences of expansion and dis- 
 tortion. Should the boiled cake become detached from the 
 plate hy twisting and warping or show expansion cracks the 
 cement may be rejected, or it may await the result of twenty- 
 eight days in water. If the fresh-water cake shows no evi- 
 dences of swelling, the cement may be used in ordinary work 
 in air or fresh water for lean mixtures. If distortion or 
 expansion cracks are shown on the fresh -water cake, the 
 cement should be rejected. 
 
 Of two or more cements offered, all of which will stand the 
 fresh-water cake test for soundness, the cements that will 
 stand the boiling tests also are to be preferred. 
 
 IV. Tensile strength. 
 
 Neat test's . — -Use unsifted cements. Place the amount to be 
 mixed on a smooth, nonabsorbent slab; make a crater in the 
 middle sufficient to hold the water; add nearly all the water 
 at once, the remainder as needed; mix thoroughly by turning 
 with the trowel, and vigorously rub or work the cement for 
 five minutes. 
 
 Place the mold on a glass or slate slab. Fill the mold with 
 consecutive layers of cement, each when rammed to be one- 
 fourth of an inch thick. Tap each layer 30 taps with a soft 
 brass or copper rammer weighing 1 pound and having a face 
 three-fourths of an inch diameter or seven-tenths of an inch 
 square with rounded corners. The tapping or ramming is to 
 
TESTING HYDRAULIC CEMENTS. 
 
 19 
 
 be done as follows: while holding the forearm and wrist at a 
 constant level, raise the rammer with the thumb and fore- 
 finger about half an inch and then let it fall freely, repeating 
 the operation until the layer is uniformly compacted by 30 
 taps. 
 
 This method is intended to compact the material in a man- 
 ner similar to actual practice in construction, when a metal 
 rammer is used weighing 30 pounds, with circular head 5 
 inches in diameter falling about 8 inches upon layers of mortar 
 or concrete 3 inches thick. The method permits comparable 
 results to be obtained by different observers. 
 
 After filling the mold and ramming the last layer, strike 
 smooth with the trowel, tap the mold lightly in a direction 
 parallel to the base plate to prevent adhesion to the plate, and 
 cover for twenty-four hours with a damp cloth. Then remove 
 the briquette from the mold and immerse it in fresh water, 
 which should be renewed twice a week for the specified time 
 if running water is not available for a slow current. If molds 
 are not available fortwenty-four hours, remove from the molds 
 after final set, replacing the damp cloth over the briquettes. 
 In removing briquettes before hard-set great care should be 
 exercised. Hold the mold in the left hand and, after loosening 
 the latch, tap gently the sides of the mold until they fall apart. 
 Place the briquettes face down in the water trough. 
 
 For neat tests of Portland cement use 20 per cent of water 
 by weight. 
 
 For neat tests of Puzzolan cement use 18 per cent of water 
 by weight. 
 
 For neat tests of Natural cement use 30 per cent of water 
 by weight. 
 
 Nearly all this water is retained by Portland cement, whereas 
 only about one-third of the gauging water is retained by Puz- 
 zolan or Natural cements; from this it follows that an apparent 
 condition of plasticity or fluidity that ultimately little injures 
 Portland paste, very seriously injures Puzzolan or Natural 
 mortars and concretes by leaving a porous texture on the 
 evaporation of the surplus w T ater. 
 
 Sand tests . — The proportions 1 cement to 3 sand are to be 
 used in tests of Puzzolan and Portland, and 1 cement to 1 sand 
 in tests of Natural or Rosendale cements. Crushed quartz 
 
20 
 
 TESTING HYDRAULIC CEMENTS. 
 
 sand, sifted to pass a standard sieve with 20 meshes per linear 
 inch and to be retained on a standard sieve with 30 meshes to 
 the inch, is to be used. 
 
 After weighing carefully, mix dry the cement and sand 
 until the mixture is uniform, add the water as in neat mix- 
 tures, and mix for five minutes by triturating or rubbing 
 together the constituents of the mortar. This may be done 
 under pressure with a trowel or by rubbing between the fin- 
 gers, using rubber gloves. The rubbing together seems nec- 
 essary to coat thoroughly the facets of the sand with the 
 cement paste. 
 
 It is found that prolonged rubbing, when not carried beyond 
 the time of initial set, results in higher tests. Five minutes 
 is the time of mixing quite generally adopted in European 
 specifications. The briquettes are to be made as prescribed 
 for neat mixtures. 
 
 Portland cements require water from 11 to 12^ per cent by 
 weight of constituent sand and cement for maximum strength 
 in tested briquettes. 
 
 Puzzolan, about 9 to 10 per cent. 
 
 Natural, about 15 to 17 per cent. 
 
 Mixtures that at first appear too dry for testing purposes 
 often become more plastic under the prolonged working 
 required herein. 
 
 In general, about four briquettes constitute the maximum 
 number that may be made well within the time required for 
 initial setting of moderately slow setting cements. 
 
 Three such batches of sand mixtures should be made, and 
 one briquette of each batch may be broken at seven and 
 twenty-eight days, giving three tests at each period. At 
 least one batch of neat cement briquettes should be made. 
 
 If the first briquette broken at each date fulfills the mini- 
 mum requirement of these specifications it is not necessary to 
 break others which may be reserved for long-time tests. 
 
 If the first briquette does not pass the test for tensile 
 strength, then briquettes may be broken until six briquettes* 
 two from each batch, have been broken at seven days, and the 
 remaining six reserved for twenty-eight-day tests. The high- 
 est result from any sample is to be taken as the strength of 
 the sample when the break is at the least section of briquette. 
 
 If, on the twenty-eight-day tests, the cement not only more 
 
TESTING HYDRAULIC CEMENTS. 
 
 21 
 
 than fulfills the minimum requirements of these specifications, 
 but also shows unusual gain in strength, it may still be accepted 
 if the other tests are satisfactory, notwithstanding a low seven- 
 day test, if early strength is not a matter of importance. Such 
 cements are likety to be permanent. 
 
 For a batch of four briquettes, the following quantities are 
 suggested as in accord with these specifications. Water is 
 measured b} T fluid-ounce volumes, not by weight, temperature 
 varying not more than 10° from 62° F. 
 
 PORTLAND CEMENT. 
 
 Neat: 20 ounces of cement, 1 ounces of water. Mix wet 
 five minutes. 
 
 Sand: 15 ounces sand, 5 ounces cement, 2-J- ounces water. 
 Mix thoroughly dry; then mix wet five minutes. 
 
 PUZZOLAN CEMENT. 
 
 Neat: 20 ounces cement, 3f ounces water. Mix wet five 
 minutes. 
 
 Sand: 15 ounces sand, 5 ounces cement, 2 ounces water. 
 Mix thoroughly dry; then mix wet five minutes. 
 
 NATURAL CEMENT. 
 
 Neat: 20 ounces cement, 6 ounces water. Mix wet five 
 minutes. 
 
 Sand: 10 ounces cement, 10 ounces sand, 3£ ounces water. 
 Mix dry; then wet for five minutes. 
 
 For measuring tensile strength, a machine that applies the 
 stress automatically at a uniform rate is preferable to one 
 controlled entirely by hand. 
 
 These specifications for tensile strength contemplate the 
 application of stress at the rate of 400 pounds per minute to 
 briquettes made as prescribed herein. A rate so rapid as to 
 approximate a blow or so slow as to approximate a continued 
 stress will give very different results. 
 
 The tests for tensile strength are to be made immediately 
 after taking from the water or while the briquettes are still 
 wet. The temperature of the water during immersion should 
 be maintained as nearly constant as practicable; not less than 
 50° nor more than 70° F. 
 
22 
 
 TESTING HYDRAULIC CEMENTS. 
 
 The tests are to be made upon briquettes 1 inch square at 
 place of rupture. The specifications contemplate the use of 
 the form of briquette recommended by the committee of the 
 American Society of Civil Engineers, held when tested by 
 close-fitting metal clips, without rubber or other jdelding 
 contacts. The breaks considered in the tests are to be those 
 occurring at the smallest section, 1 inch square. 
 
 SIMPLE TESTS. 
 
 Tests of cement received upon a work in progress must 
 often be of much simpler character than prescribed herein. 
 
 Tests on the work are mainly to ascertain whether the 
 article supplied is genuine cement, of a brand previously 
 tested and accepted, and whether it is a reasonably sound and 
 active cement that will set hard in the desired time, and give 
 a good, hard mortar. Simple tests may give this information, 
 and such should be multiplied whether or not more elaborate 
 tests be made. Pats and balls of cement and mortar from 
 the storehouse and mixing platform or machine should be 
 frequently made. The setting or hardening qualities, as de- 
 termined roughly by estimating time and by pressure of the 
 thumbnail, should be observed; the hardness of the set and 
 strength, by cracking the hardened pats or cakes between the 
 fingers, and by dropping the balls from the height of the arm 
 upon a pavement or stone and observing the result of the 
 impact. 
 
 By placing the pats in water as soon as hardened sufficiently 
 and raising the temperature to the boiling point for a few 
 hours and observing the character and color of the fracture 
 after sufficient immersion, information as to the character of 
 the material, whether hydraulic, a Portland or Puzzolan, 
 whether too fresh or possibly “ blowy,” may be speedily and 
 quite well ascertained without measuring instruments. 
 
 Many engineers and users of cements regard such simple 
 tests, taken in connection with the weight and fineness of the 
 cement and the apparent texture and hardness of the mortars 
 and concretes in the work, sufficient field tests of a material 
 of known repute. The more elaborate tests, described above, 
 should be made in well-equipped laboratories by skilled cement 
 testers. 
 
TESTING HYDRAULIC CEMENTS. 
 
 2a 
 
 CLASSIFICATION OF TESTS. 
 
 The tests to be made are two classes. 
 
 (1) Purchase tests on samples furnished by bidders to as- 
 certain whether the bidder may be held on the sample to the 
 delivery of suitable material, should his offer be accepted. 
 
 (2) Acceptance tests on samples taken at random from 
 deliveries, to ascertain whether the material supplied accords 
 with the purchase sample, or is suitable for the purpose of 
 the work, as stated in the specifications for cement supplies. 
 
 (1) Purchase tests . — Under these specifications bids for Port- 
 land cements will be restricted to brands that have been ap- 
 proved af tey at least three years’ exposure in successful use 
 under similar conditions to those of the proposed work. This 
 specification limits proposals to manufacturers of cements of 
 established repute, and in so far lessens the dependence to be 
 placed upon tests of single samples of cement in determining 
 the probable quality of the cements offered, that sample pack- 
 ages may not be required with the proposals when the brand 
 is known to the purchaser. When the cement is not known 
 to the purchasing officer by previous use, a barrel of it should 
 be required as representing the quality of cement to be sup- 
 plied. A full set of tests should be made from this sample, 
 and subsequent deliveries be required to show quality at least 
 equal to the sample. 
 
 In this connection it is advisable in districts where well- 
 equipped laboratories have been established, that sample 
 packages of the cements in use in that territory, as sold in 
 the open market, be obtained and tested as occasion offers to 
 ascertain the characteristic qualities of the brands as commer- 
 cial articles, the information to be used in subsequent pur- 
 chases of cements. 
 
 When purchase samples are waived, acceptance tests should 
 be based upon the known qualities of the brand, as shown by 
 previous tests. 
 
 The sample barrel should not be broken further than to 
 take therefrom the necessary samples for testing. After- 
 wards it should be put away in a dry place and kept for fur- 
 ther testing, should the results obtained be disputed. 
 
 (2) Acceptance tests .— The tests to be made on cements 
 delivered under contract depend not only on the extent, 
 
24 
 
 TESTING HYDRAULIC CEMENTS, 
 
 character, and importance of the work itself, but also on the 
 time available between the delivery and the actual use of the 
 material. 
 
 (a) On very important and extensive works, equipped with 
 a testing laboratory and adequate storehouses, where cement 
 ma}^ be kept at least thirty days before being required for use, 
 full and elaborate tests should be made, keeping in view the 
 fact that careful tests of few samples are more valuable than 
 hurried tests of many samples. 
 
 (b) On active works of ordinary character, when time will 
 not permit full tests, and on small works where the expenses 
 of a laboratoiy are not justified, the tests must necessarity be 
 limited to such reasonable precautions against the acceptance 
 and use of unfit material as may be taken in the usualty short 
 interval between the receipt and use of the material. 
 
 Such conditions were in view in formulating the specifica- 
 tion that proposals will be received from manufacturers of 
 such cements only as have been proved by at least three }^ears’ 
 use under similar conditions of exposure. Of the tests named 
 in the specifications those for fineness, activity or hydraulicity, 
 specific gravity, weight of packages, and accelerated tests for 
 indications as to soundness, may be made within two da} T s 
 after the receipt of the material and with a very small 
 outlay for instruments. 
 
 Cement of established repute, shown b} T specific gravity 
 and fineness to be property burnt and ground, or normal for 
 the brand, that will set hard in reasonable time, the cakes 
 snapping with a clean fracture when broken between the 
 fingers, and standing the tests above named, may be accepted 
 and used with reasonable certainty of success. Nevertheless, 
 packages taken at random from the deliveries should occasion- 
 ally be set aside and samples taken therefrom sent to a testing 
 laboratoiy for the more elaborate tests for tensile strength 
 (and for soundness should the boiling tests not be conclusive). 
 The final acceptance and payment for such cement as ma} T not 
 have been actually placed in the work should, by agreement, 
 be made to depend upon such tests. 
 
 In all cases where cement has been long stored it should be 
 carefully tested before use to ascertain whether it has deterio- 
 rated in strength. 
 
TESTING HYDRAULIC CEMENTS. 
 
 25 
 
 Should the simple tests give unsatisfactory or suspicious 
 results, then a full series of tests should be carefulty made. 
 
 When Portland cement is in question the specific gravity 
 and fineness tests should be made to guard against adultera- 
 tion, and in all cases test weighings should be made to guard 
 against short weights. 
 
 In cases where the amount of cement or the importance of 
 the work will not justify the purchase of the simple apparatus 
 required for the specific gravity, fineness, and boiling tests, 
 the cement can be accepted on the informal tests mentioned 
 herein, which require no apparatus whatever, but in such 
 cases cements well known to the purchaser by previous use 
 should be selected, and purchased directly from the manu- 
 facturer or his selling agent in order that responsibility for 
 the cement may be fixed. 
 
 Certified tests by professional inspectors made as prescribed 
 herein on samples taken from the cement to be shipped to 
 the work, in a manner analogous to that customary among 
 engineers in the purchase of structural steel and iron, may 
 be required in such cases. 
 
 SAMPLING. 
 
 The entire package from parts of which tests are to be made 
 is to be regarded as the sample tested. It should be marked 
 with a distinctive mark that must also be applied to any" part 
 tested. The package should be set aside and protected against 
 deterioration until all results from tests made from it are 
 reached and accepted by both parties to the contract for 
 supplies. 
 
 Cement drawn from several sample packages should not be 
 mixed or mingled, but the individuality of each sample pack- 
 age should be preserved. 
 
 In testing it should be borne in mind that a few tests from 
 any sample, carefully made, are more valuable than many 
 made with less care. 
 
 The amount of material to be taken for formal tests is 
 indicated herein where weights of the constituents of four 
 briquettes are given, to which should be added the amount 
 necessary for the tests for specific gravity, activity, and 
 soundness. 
 
26 
 
 TESTING HYDRAULIC CEMENTS. 
 
 In extended tests the material should be taken from the 
 sample package from the heads and center of barrel, and 
 from the ends and center of bag, by such an instrument as 
 is used by inspectors of flour. All material taken from the 
 same sample package may be thoroughly mixed or mingled 
 and the tests be made therefrom as showing the true charac- 
 ter of the contents of the sample package. 
 
 In making formal tests at the work for acceptance of cement 
 sample packages should be taken at random from among sound 
 packages. The number taken must depend upon the impor- 
 tance and character of the work, the available time, and the 
 capacity of the permanent laboratory force. For tensile 
 strength the tests with sand are considered the more impor- 
 tant and should alwa}^s be made. Tests neat should be made 
 if time permits. 
 
 It is not necessary in any case on a large work to test more 
 than 10 per cent of the deliveries, even of doubtful cement, 
 and a much less number of samples may be taken should no 
 cause for distrust be revealed by the tests made. In very 
 important work of small extent each package may be tested. 
 A cement should be rejected if the samples show dangerous 
 variation in quality or lack of care in manufacture and result- 
 ing lack of uniformity in the product without regard to the 
 proportion of failures among samples tested. 
 
 In all cases in the use of cements the informal or simple 
 tests of the character named herein should be constantly car- 
 ried on. These constitute most valuable tests. Whenever 
 any faulty material is indicated by such tests, elaborate tests 
 should be at once instituted and should the fault be confirmed, 
 the cement delivered and not used should be rejected and the 
 use of the brand be discontinued. 
 
 TESTS FOR WEIGHT. 
 
 From time to time packages should be weighed in gross and 
 afterwards the weight of neat cement and tare of the packages 
 determined. If short weight of neat cement is indicated, a 
 sufficient number of packages should be weighed and the aver- 
 age net weight per package ascertained with sufficient cer- 
 tainty to afford a satisfactory basis of settlement. 
 
TESTING HYDRAULIC CEMENTS. 
 
 27 
 
 RECORDS. 
 
 For tests at professional laboratories no general require- 
 ments as to records seem to be necessary. Each laboratory 
 has its own blanks with certificate, and if a copy of the speci- 
 fications be sent with the samples, the record returned should 
 be sufficient. For records of formal tests on the work, or in 
 a district laboratory, blank forms should be used. It is desir- 
 able to have the specification requirements stated on the form. 
 Notations should be adopted to show for each test that the 
 cement passed or failed or that the test was not made. No 
 inference should be drawn from the lack of any entry other 
 than that the recorder has neglected his duty. 
 
 SILICA CEMENT OR SAND CEMENT. 
 
 This is a patented article manufactured by grinding together 
 silica or clean sand with Portland cement, by which process 
 the original cementing material is made extremely fine and its 
 capacity to cover surfaces of concrete aggregates is much in- 
 creased. The sand is an adulteration, but on account of the 
 extreme fineness of the product it serves to make mortar or 
 concrete containing a given proportion of pure cement much 
 more dense, the fine material being increased in volume. 
 
 The increase in cementing capacity due to the fine grinding 
 of the cement constituent offsets, in great degree, the effects 
 of the sand adulteration, so that sand cement made from equal 
 weights of cement and sand approximates in tensile strength 
 to the neat cement and the material is sold as cement. 
 
 The extreme fine grinding also improves cement that con- 
 tains expansives, but nevertheless sand cement should not be 
 purchased in the market, but should be made on the work from 
 approved materials, if used for other purposes than for grout- 
 ing, for which it is peculiarly adapted. 
 
 Whether this material should be used in important works 
 for mortar and concrete, the Board considers a question of 
 cost and expediency. 
 
 Over against the saving in cement may be placed the 
 royalty on a patented article, the cost of the plant and of 
 manufacture, the inconvenience of attaching a manufacturing 
 establishment to a work under construction, and other ele- 
 
28 
 
 TESTING HYDRAULIC CEMENTS. 
 
 merits bearing not only on first cost of cementing material 
 but also involving the element of time. When cement is 
 high priced, means of transportation limited, labor, sand, 
 and concrete materials cheap and abundant, the conditions 
 may justify the use of sand cement on economic grounds. In 
 any case, the cement from which the product is made should 
 be tested precisely as other cements. 
 
 SLAG CEMENT. 
 
 This term is applied to cement made by intimately mixing 
 by grinding together granulated blast-furnace slag of a cer- 
 tain quality and slaked lime, without calcination subsequent 
 to the mixing. This is the only cement of the Puzzolan class 
 to be found in our markets (often branded as Portland), and 
 as true Portland cement is now made having slag for its 
 hydraulic base, the term “slag cement” should be dropped 
 and the generic term Puzzolan be used in advertisements and 
 specifications for such cements. 
 
 Puzzolan cement made from slag is characterized physically 
 by its light lilac color; the absence of grit attending fine grind- 
 ing and the extreme subdivision of its slaked lime element; 
 its low specific gravity (2.6 to 2.8) compared with Portland 
 (3 to 3.5); and by the intense bluish green color in the fresh 
 fracture after long submersion in water, due to the presence 
 of sulphides, which color fades after exposure to dry air. 
 
 The oxidation of sulphides in dry air is destructive of Puz- 
 zolan cement mortars and concretes so exposed. Puzzolan is 
 usually very finely ground, and when not treated with soda 
 sets more slowly than Portland. It stands storage well, but 
 cements treated with soda to quicken setting become again 
 very slow setting from the carbonization of the soda (as well 
 as the lime) element after long storage. 
 
 Puzzolan cement properl}" made contains no free or anhy- 
 drous lime, does not warp or swell, but is liable to fail from 
 cracking and shrinking (at the surface only) in dry air. 
 
 Mortars and concretes made from Puzzolan approximate in 
 tensile strength similar mixtures of Portland cement, but 
 their resistance to crushing is less, the ratio of crushing to 
 tensile strength being’ about 6 or T to 1 for Puzzolan and 9 to 
 11 to 1 for Portland. On account of its extreme fine grind- 
 
TESTING HYDRAULIC CEMENTS. 
 
 29 
 
 ing Puzzolan often gives nearly as great tensile strength in 3 
 to 1 mixtures as neat. 
 
 Puzzolan permanently assimilates but little water compared 
 with Portland, its lime being already hydrated, It should be 
 used in comparatively dry mixtures well rammed, but while 
 requiring little water for chemical reactions, it requires for 
 permanency in the air constant or continuous moisture. 
 
 PROPER USES OF PUZZOLAN CEMENT. 
 
 Puzzolan cement never becomes extremely hard like Port- 
 land, but Puzzolan mortars and concretes are tougher or less 
 brittle than Portland. 
 
 The cement is well adapted for use in sea water, and gen- 
 erally in all positions where constantly exposed to moisture, 
 such as in foundations of buildings, sewers and drains, and 
 underground works generally, and in the interior of heavy 
 masses of masonry or concrete. 
 
 It is unfit for use when subjected to mechanical wear, attri- 
 tion, or blows. It should never be used where it may be ex- 
 posed for long periods to dry air, even after it has well set. It 
 will turn white and disintegrate, due to the oxidation of its 
 sulphides at the surface under such exposure. 
 
 Specifications for Portland, Natural, and Puzzolan cement 
 are appended hereto. 
 
 Kespectfully submitted. 
 
 W. L. Marshall, 
 
 Major , Corps of Engineers. 
 
 Smith S. Leach, 
 
 Major , Corps of Engineers. 
 
 Spencer Cosby, 
 
 Captain , Corps of Engineers. 
 
 Brig. Gen. G. L. Gillespie, 
 
 Corps of Engineers, JJ. S. Army , 
 
 Washington , D. C. 
 
30 
 
 TESTING HYDRAULIC CEMENTS. 
 
 SPECIFICATIONS FOR AMERICAN PORTLAND 
 CEMENT. 
 
 (1) The cement shall be an American Portland, dry and free 
 from lumps. By a Portland cement is meant the product 
 obtained from the heating or calcining up to incipient fusion 
 of intimate mixtures, either natural or artifical, of argillaceous 
 with calcareous substances, the calcined product to contain at 
 least 1.7 times as much of lime, by weight, as of the materials 
 which give the lime its hydraulic properties, and to be finely 
 pulverized after said calcination, and thereafter additions or 
 substitutions for the purpose only of regulating certain prop- 
 erties of technical importance to be allowable to not exceed- 
 ing 2 per cent of the calcined product. 
 
 (2) The cement shall be put up in strong, sound barrels well 
 lined with paper, so as to be reasonably protected against 
 moisture, or in stout cloth or canvas sacks. Each package 
 shall be plainly labeled with the name of the brand and of the 
 manufacturer. Any package broken or containing damaged 
 cement may be rejected or accepted as a fractional package, at 
 the option of the United States agent in local charge. 
 
 (3) Bidders will state the brand of cement which they pro- 
 pose to furnish. The right is reserved to reject a tender for 
 any brand which has not established itself as a high-grade 
 Portland cement and has not for three years or more given 
 satisfaction in use under climatic or other conditions of ex- 
 posure of at least equal severity to those of the work proposed. 
 
 (4) Tenders will be received only from manufacturers or 
 their authorized agents. 
 
 (The following paragraph will be substituted for paragraphs 
 3 and 4 above when cement is to be furnished and placed by 
 the contractor: 
 
 No cement will be allowed to be used except established 
 brands of high-grade Portland cement which have been made 
 by the same mill and in successful use under similar climatic 
 conditions to those of the proposed work for at least three 
 years. ) 
 
 (5) The average weight per barrel shall not be less than 375 
 pounds net. Pour sacks shall contain one barrel of cement. 
 If the weight, as determined b} r test weighings, is found to 
 be below 375 pounds per barrel, the cement may be rejected, 
 
TESTING HYDRAULIC CEMENTS. 
 
 31 
 
 or, at the option of the engineer officer in charge, the con- 
 tractor may be required to suppl} r , free of cost to the United 
 States, an additional amount of cement equal to the shortage. 
 
 (6) Tests may be made of the fineness, specific gravity, 
 soundness, time of setting, and tensile strength of the cement. 
 
 (7) Fineness. — Ninety-two per cent of the cement must pass 
 through a sieve made of No. 10 wire, Stubb’s gauge, having 
 10,000 openings per square inch. 
 
 (8) Specific gravity . — The specific gravity of the cement, as 
 determined from a sample which has been carefully dried, 
 shall be between 3.10 and 3.25. 
 
 (9) Soundness . — To test the soundness of the cement, at least 
 two pats of neat cement mixed for five minutes with 20 per 
 cent of water by weight shall be made on glass, each pat about 
 3 inches in diameter and one-half inch thick at the center, 
 tapering thence to a thin edge. The pats are to be kept 
 under a wet cloth until finall} 7 set, when one is to be placed in 
 fresh water for twenty-eight days. The second pat will be 
 placed in water which will be raised to the boiling point for six 
 hours, then allowed to cool. Neither should show distortion 
 or cracks. The boiling test may or may not reject at the op- 
 tion of the engineer officer in charge. 
 
 (10) Time of setting . — The cement shall not acquire its 
 initial set in less than forty-five minutes and must have 
 acquired its final set in ten hours. 
 
 (The following paragraph will be substituted for the above 
 in case a quick-setting cement is desired: 
 
 The cement shall not acquire its initial set in less than twenty 
 nor more than thirty minutes, and must have acquired its final 
 set in not less than forty-five minutes nor in more than two 
 and one-half hours.) 
 
 The pats made to test the soundness may be used in deter- 
 mining the time of setting. The cement is considered to have 
 acquired its initial set when the pat will bear, without being 
 appreciably indented, a wire one-twelfth inch in diameter 
 loaded to weigh one-fourth pound. The final set has been 
 acquired when the pat will bear, without being appreciably 
 indented, a wire one twenty-fourth inch in diameter loaded 
 to weigh 1 pound. 
 
 (11) Tensile strength . — Briquettes made of neat cement, 
 after being kept in air for twenty-four hours under a wet 
 
82 
 
 TESTING HYDKAULIC CEMENTS. 
 
 cloth and the balance of the time in water, shall develop ten- 
 sile strength per square inch as follows: 
 
 After seven days, 450 pounds; after twenty-eight days, 540 
 pounds. 
 
 Briquettes made of 1 part cement and 3 parts standard sand, 
 by weight, shall develop tensile strength per square inch as 
 follows : 
 
 After seven days, 140 pounds; after twenty-eight days, 220 
 pounds. 
 
 (In case quick-setting cement is desired, the following ten- 
 sile strengths shall be substituted for the above: 
 
 Neat briquettes: After seven days, 400 pounds; after twenty- 
 eight da} r s, 480 pounds. 
 
 Briquettes of 1 part cement to 3 parts standard sand: After 
 seven days, 120 pounds; after twenty-eight days, 180 pounds.) 
 
 (12) The highest result from each set of briquettes made at 
 any one time is to be considered the governing test. Any 
 cement not showing an increase of strength in the twenty- 
 eight-day tests over the seven-day tests will be rejected. 
 
 (13) When making briquettes neat cement will be mixed 
 with 20 per cent of water by weight, and sand and cement 
 with 12i per cent of water by weight. After being thor- 
 ough^ mixed and worked for live minutes, the cement or mor- 
 tar will be placed in the briquette mold in four equal layers, 
 and each layer rammed and compressed by thirty blows of a 
 soft brass or copper rammer three-quarters of an inch in 
 diameter (or seven-tenths of an inch square, with rounded cor- 
 ners), weighing 1 pound. It is to be allowed to drop on the 
 mixture from a height of about half an inch. When the ram- 
 ming has been completed, the surplus cement shall be struck 
 off and the final layer smoothed with a trowel held almost hori- 
 zontal and drawn back with sufficient pressure to make its edge 
 follow the surface of the mold. 
 
 (14) The above are to be considered the minimum require- 
 ments. Unless a cement has been recently used on work 
 under this office, bidders will deliver a sample barrel for test 
 before the opening of bids. If this sample shows higher tests 
 than those given above, the average of tests made on subse- 
 quent shipments must come up to those found with the sample. 
 
 (15) A cement may be rejected in case it fails to meet any 
 of the above requirements. An agent of the contractor may 
 
TESTING HYDKAULIC CEMENTS. 
 
 33 
 
 be present at the making of the tests, or, in case of the fail- 
 ure of any of them, they may be repeated in his presence. 
 If the contractor so desires, the engineer officer in charge 
 may, if he deem it to the interest of the United States, have 
 any or all of the tests made or repeated at some recognized 
 standard testing laboratory in the manner herein specified. 
 All expenses of such tests to be paid by the contractor. All 
 such tests shall be made on samples furnished by the engineer 
 officer from cement actually delivered to him. 
 
 SPECIFICATIONS FOR NATURAE CEMENT. 
 
 (1) The cement shall be a freshly packed natural or Rosen- 
 dale, dry, and free from lumps. By Natural cement is meant 
 one made by calcining natural rock at a heat below incipient 
 fusion, and grinding the product to powder. 
 
 (2) The cement shall be put up in strong, sound barrels, well 
 lined with paper so as to be reasonably protected against mois- 
 ture, or in stout doth or canvas sacks. Each package shall be 
 plainly labeled with the name of the brand and of the manu- 
 facturer. Any package broken or containing damaged cement 
 may be rejected, or accepted as a fractional package, at the 
 option of the United States agent in local charge. 
 
 (3) Bidders will state the brand of cement which they pro- 
 pose to furnish. The right is reserved to reject a tender for 
 any brand which has not given satisfaction in use under cli- 
 matic or other conditions of exposure of at least equal severity 
 to those of the work proposed. 
 
 (4) Tenders will be received only from manufacturers or 
 their authorized agents. 
 
 (The following paragraph will be substituted for paragraphs 
 3 and 4 above when cement is to be furnished and placed by 
 the contractor: 
 
 No cement will be allowed to be used except established 
 brands of high-grade natural cement which have been in suc- 
 cessful use under similar climatic conditions to those of the 
 proposed work.) 
 
 (5) The average net weight per barrel shall not be less than 
 
 300 pounds. (West of the Allegheny Mountains this may be 
 265 pounds.) . . . sacks of cement shall have the same 
 
 1593—01 3 
 
34 
 
 TESTING HYDRAULIC CEMENTS. 
 
 weight as 1 barrel. If the average net weight, as determined 
 by test weighings, is found to be below 300 pounds (265 
 pounds) per barrel, the cement may be rejected, or, at the 
 option of the engineer officer in charge, the contractor may 
 be required to supply free of cost to the United States an 
 additional amount of cement equal to the shortage. 
 
 (6) Tests may be made of the fineness, time of setting, and 
 tensile strength of the cement. 
 
 (7) Fineness . — At least 80 per cent of the cement must pass 
 through a sieve made of No. 40 wire, Stubb’s gauge, having 
 10,000 openings per square inch. 
 
 (8) Time of setting . — The cement shall not acquire its initial 
 set in less than twenty minutes and must have acquired its 
 final set in four hours. 
 
 (9) The time of setting is to be determined from a pat of 
 neat cement mixed for five minutes with 30 per cent of water 
 by weight and kept under a wet cloth until finally set. The 
 cement is considered to have acquired its initial set when the 
 pat will bear, without being appreciabty indented, a wire one- 
 twelfth inch in diameter loaded to weigh one-fourth pound. 
 The final set has been acquired when the pat will bear, with- 
 out being appreciably indented, a wire one twenty-fourth inch 
 in diameter loaded to weigh 1 pound. 
 
 (10) Tensile strength . — Briquettes made of neat cement 
 shall develop the following tensile strengths per square inch, 
 after having been kept in air for twenty-four hours under a 
 wet cloth and the balance of the time in water: 
 
 At the end of seven days, 90 pounds; at the end of twenty- 
 eight days, 200 pounds. 
 
 Briquettes made of one part cement and one part standard 
 sand by weight shall develop the following tensile strengths 
 per square inch: 
 
 After seven days, 60 pounds; after twenty-eight days, 150 
 pounds. 
 
 (11) The highest result from each set of briquettes made at 
 any one time is to be considered the governing test. Any 
 cement not showing an increase of strength in the twenty-eight 
 day tests over the seven-day' tests will be rejected. 
 
 (12) The neat cement for briquettes shall be mixed with 30 
 per cent of water by weight, and the sand and cement with 17 
 per cent of water by weight. After being thoroughly mixed 
 
TESTING HYDRAULIC CEMENTS. 
 
 35 
 
 and worked for five minutes the cement or mortar is to be 
 placed in the briquette mold in four equal layers, each of 
 which is to be rammed and compressed by thirty blows of 
 a soft brass or copper rammer three-fourths of an inch in 
 diameter (or seven-tenths of an inch square with rounded cor- 
 ners), weighing 1 pound. It is to be allowed to drop on the 
 mixture from a height of about half an inch. Upon the com- 
 pletion of the ramming the surplus cement shall be struck off 
 and the last layer smoothed with a trowel held nearly horizon- 
 tal and drawn back with sufficient pressure to make its edge 
 follow the surface of the mold. 
 
 (13) The above are to be considered the minimum require- 
 ments. Unless a cement has been recently used on work 
 under this office, bidders will deliver a sample barrel for test 
 before the opening of the bids. Any cement showing by 
 sample, higher tests than those given must maintain the 
 average so shown in subsequent deliveries. 
 
 (14) A cement may be rejected which fails to meet any of 
 the above requirements. An agent of the contractor may be 
 present at the making of the tests, or, in case of the failure 
 of any of them, they may be repeated in his presence. If 
 the contractor so desires, the engineer officer may, if he 
 deems it to the interest of the United States, have any or all 
 of the tests made or repeated at some recognized standard 
 testing laboratory in the manner above specified. All ex- 
 penses of such tests shall be paid by the contractor, and all 
 such tests shall be made on samples furnished by the engineer 
 officer from cement actually delivered to him. 
 
 SPECIFICATIONS FOR PIJZZOEAN CEMENT. 
 
 (1) The cement shall be a Puzzolan of uniform quality, 
 finely and freshly ground, dry, and free from lumps, made by 
 grinding together without subsequent calcination granulated 
 blast-furnace slag with slaked lime. 
 
 (2) The cement shall be put up in strong sound barrels well 
 lined with paper, so as to be reasonably protected against 
 moisture, or in stout cloth or canvas sacks. Each package 
 shall be plainty labeled with the name of the brand and of the 
 
36 
 
 TESTING HYDRAULIC CEMENTS. 
 
 manufacturer. An}^ package broken or containing damaged 
 cement may be rejected, or accepted as a fractional package, 
 at the option of the United States agent in local charge. 
 
 (3) Bidders will state the brand of cement which they pro- 
 pose to furnish. The right is reserved to reject a tender for 
 any brand which has not given satisfaction in use under cli- 
 matic or other conditions of exposure of at least equal severity 
 to those of the work proposed, and for any brand from cement 
 works that do not make and test the slag used in the cement. 
 
 (4) Tenders will be received only from manufacturers or 
 their authorized agents. 
 
 (The following paragraph will be substituted for paragraphs 
 3 and 4 above when cement is to be furnished and placed by 
 the contractor: 
 
 No cement will be allowed to be used except established 
 brands of high-grade Puzzolan cement which have been in 
 successful use under similar climatic conditions to those of 
 the proposed work and which come from cement works that 
 make the slag used in the cement.) 
 
 (5) The average weight per barrel shall not be less than 330 
 pounds net. Four sacks shall contain 1 barrel of cement. 
 If the weight as determined by test weighings is found to be 
 below 330 pounds per barrel, the cement may be rejected or, 
 at the option of the engineer officer in charge, the contractor 
 may be required to supply, free of cost to the United States, 
 an additional amount of cement equal to the shortage. 
 
 (6) Tests may be made of the lineness, specific gravity, 
 soundness, time of setting, and tensile strength of the cement. 
 
 (7) Fineness.— Ninety-seven per cent of the cement must 
 pass through a sieve made of No. 40 wire, Stubb’s gauge, 
 having 10,000 openings per square inch. 
 
 (8) Specific gravity . — The specific gravity of the cement, as 
 determined from a sample which has been carefully dried, 
 shall be between 2.7 and 2.8. 
 
 (9) Soundness . — To test the soundness of cement, pats of 
 neat cement mixed for five minutes with 18 per cent of water 
 by weight shall be made on glass, each pat about 3 inches in 
 diameter and one-half inch thick at the center, tapering thence 
 to a thin edge. The pats are to be kept under wet cloths until 
 finally set, when they are to be placed in fresh water. They 
 should not show distortion or cracks at the end of twenty-eight 
 days. 
 
TESTING HYDRAULIC CEMENTS. 
 
 37 
 
 (10) Time of setting . — The cement shall not acquire its ini- 
 tial set in less than forty-five minutes and shall acquire its 
 final set in ten hours. The pats made to test the soundness 
 ma} r be used in determining the time of setting. The cement 
 is considered to have acquired its initial set when the pat will 
 bear, without being appreciably indented, a wire one-twelfth 
 inch in diameter loaded to one-fourth pound weight. The 
 final set has been acquired when the pat will bear, without 
 being appreciably indented, a wire one twenty-fourth inch in 
 diameter .loaded to 1 pound weight. 
 
 (11) Tensile strength . — Briquettes made of neat cement, 
 after being kept in air under a wet cloth for twenty-four 
 hours and the balance of the time in water, shall develop ten- 
 sile strengths per square inch as follows: 
 
 After seven da}^s, 350 pounds; after twenty-eight days, 500 
 pounds. 
 
 Briquettes made of one part cement and three parts stand- 
 ard sand by weight shall develop tensile strength per square 
 inch as follows: 
 
 After seven da 3 ^s, 140 pounds; after twenty-eight, days, 220 
 pounds. 
 
 (12) The highest result from each set of briquettes made at 
 any one time is to be considered the governing test. Any 
 cement not showing an increase of strength in the twenty- 
 eight-day tests over the seven-day tests will be rejected. 
 
 (13) When making briquettes neat cement will be mixed 
 with 18 per cent of water by weight, and sand and cement 
 with 10 per cent of water by weight. After being thoroughly 
 mixed and worked for five minutes the cement or mortar 
 will be placed in the briquette mold in four equal layers and 
 each layer rammed and compressed by thirty blows of a soft 
 brass or copper rammer, three-quarters of an inch in diameter 
 or seven-tenths of an inch square, with rounded corners, 
 weighing 1 pound. It is to be allowed to drop on the mixture 
 from a height of about half an inch. When the ramming has 
 been completed the surplus cement shall be struck off and the 
 final layer smoothed with a trowel held almost horizontal and 
 drawn back with sufficient pressure to make its edge follow 
 the surface of the mold. 
 
 (14) The above are to be considered the minimum require- 
 ments. Unless a cement has been recently used on work 
 
38 
 
 TESTING HYDRAULIC CEMENTS. 
 
 under this office, bidders will deliver a sample barrel for test 
 before the opening of bids. If this sample shows higher tests 
 than those given above, the average of tests made on subse- 
 quent shipments must come up to those found with the sample. 
 
 (15) A cement may be rejected in case it fails to meet any 
 of the above requirements. An agent of the contractor may 
 be present at the making of the tests, or, in case of the failure 
 of any of them, they may be repeated in his presence. If the 
 contractor so desires the engineer officer in charge may, if 
 he deems it to the interest of the United States, have an} 7 or 
 all of the tests made or repeated at some recognized testing 
 laboratory in the manner herein specified, all expenses of such 
 tests to be paid by the contractor. All such tests shall be 
 made on samples furnished by the engineer officer from 
 cement actually delivered to him. 
 
 O 
 

 
 
 
 
 
 
 
I