‘vm BRO Hi Hoe — OP REN irs ca ee a ee a ee ee Sey Se ee ata soa Rah Oe - — ee{ ae a Neda a ete) THESIS OC grr aern — Paolo | - THESIS lll THESIS. -: te: + ty 9a rs ep) lo [3 Ia {tes (te [r= [ra By ? : ie eae , , . ‘ l oe G.i. CAPDOZO and A’. C. DODGE. THESIS TeSiS OF CLiuENT. The object of this set of tests was vrimarily to determine the effect, if any, of the repeated aprlication of loads -- well within the ultimate breaking linit -- on cement briqvetts,and incidentally to deternine the effect of saturation with water on the strength of bri- quests end cubes, The briavetts used in tnese tests were cf tne stan- <8) Gard shape with the smallest cross sectional aren esual cne (1) inch. The cubes were three (3) inches cn a side. The moulds used fur the bricguetts were the recular cast iron movlds used by the department, each movic hold- ins six (6) briaqivetts, For the cubes, @ special movld was constmcted. The sides and ends were of cast iron, anc at intervals cf three (3) inches vartical slots were WwW cut in each sige into which sheet steel squares three (3) inches on a side and three-sixteenths (3/16) inches thick were inserted, thus forming eight cubes in the mould. One side was screwed on with but one screw at one end and was held with an ordinary wooden clamp at the other, ‘thus Allowing for the easy removal of the cement blocks. Two different brands of cement were used (Atlas and Aetna). The Aetna being a local cement manufactured at Fenton, Michiszan, and the Atlas a well known standard brand. We found little or no difference in the quality 94507 of these cements thro -mhout the test. There were two (2) mixtures of each cement tosted, viz. a one-to-three (1 to 3) and a neat mixture. The cenont in every case was sifted throuch a hundred (100) gauze sieve, and the prerortions of water and sand were made by weignt. The send used was the standard crushed vartz testins sand, and was sifted threutnh a nunhber twenty (2) sieve. 5 Tne machine constructed for avnnlyins the interrittant ¢ aver te Which at cre f-4 load consist:d essentially of a end was attached a counte: weight and a clanp for hclding a briquette, while on the other a load was arrltied by reans of tension in & spring balance. This load was made intermittant through the revolution of 4 chaft, to the end of which an arm was fastened at a risht angle. A stud was screwed inte this arm and the balance hochked inte a groove in ti.1ls stud. Tnen the arm was in & Ver- tical position upward, the load was due simply to the weirht of the balance ard the longer arm of the lever; this load was accounted for by the counter weisnt on the other end makins the actual strain on the briquette: equal to zero (0). As the shaft revolved, the load began at once to be applied, reachinzg a maximum after a revolution through one hundred anc eighty (180) derrees, and then decreasing to zero (0) on the completion of the turn. The shaft made thirty-eight (38) revolutions per minute, thus making a corresponding number cf load ap- plications in the same time. A counter was attached to record the revolutions. For details sce drawing. For other tensile tests, a regular standard Fairbanks shot machine was used. The intermittant loads applied were determined by means of the sprins balance at the point of application cf the load, ard as a check,readings of load at berinning and end of test were taken by means of sprins balance put in place of the briquette. As the ratio of lever arns was four and seven-tenths (4.7) to one (1), multi- plyins the balance reading at tne point of application by four and seven tenths (4.7) pave the load on the bri- quette. This computed load on the briquette never dif- Yered from the load as actually measured with the balance in place of the briquette by more than two 2) or three (3) pounds. the first briquette placed in the machine was put in wet and given a load of one hundred and twenty pounds (120 lbs.), breaking on the first application. Accord-~ insly the second briquette was given a load of fifty- eight (58) povnds, which was thought to be about half its ultimate strength. After five thousand revolutions the briqvette, havine arnnarently suffered no change, was taken out and broken on the Fairbanks machine. The breaking load vas one hundred and eishty-two (182) pounds. The briquette had apparently grown stronger. Another briquette was accordingly taken from the water and tested on the Fairbanks machina, breaking at one hundred and thirty one (131) pounds. The intermittant load was now made eighty-seven (87) pounds and fifty thousand (50,000) applications were siven. Durins the aprlication of this load the machine was shut dcwn twice owing to the closing cf the shop for the night. After fifty thousand (50,000) explications, the oeiquette bein anparently vninjured, it was broken on tre Falrbeanks machine. Its ultimate ob was found te be tro hundred and ninety-two (292) povnds. $+ The arperent difforence in strensthn oetiwoen the briquettcs after leavine the machine as comrared with the strength i on being placed in the machine lea te the conclusion that either the intermittant load or the evaroration of the water which the briquette had absorbed strengtinened it. As the first conclision seems manifestly absurd, the second must account for the increase in strenztn. Accoré- invly a sories of tests wes uncertakon to verify this conclusion, and, as far as we were able to deternine with a limited number of briqiettes, there is a variation of about twenty-five (25) per cent in strength between wet and dry one‘l) to three (3) cement mixtures. The neat cement mixtures apparently suffered no change in strength whether wet or dry. This leads to the conclu- sion that for masonry work usin=s cement mortar or for concrete work which is exposed under water, the tensile or compressive streneth of the concrete and of tne joints of masonry wovld be creatly increased if faced over with a thin coat of neat ceirent. After this investigation, all bricuettes were dried before placing then in thc intermittant load machine. For results cof tnese tests see table II. Owins to the high tensile strees shown by tne neat cement and to the veakness of the eprings usec in the construction of the sprint palance, it Was linbo:clbie to cer CG make tes for fatigue with neat cement bricquettes. The tests for Pfatisuse show that for losds thet are Less than eighty (30) or ninety (SO) per cent of the ultimate Strengtn tne nunber of applications recuired for oréaking . is comparatively large. At about ninety (90) per cent of the final streos, there is a rather sudden weakening of the briquettes and only a few applications of the load ara cequired. For results of tne fatigue toats see tabis I. The compression test was conevucted on tire lasce Tinius Olsen machine. The cubes were placed between two cast iron surface plates in order that they might have a perfectly smooth bearing. Owing to the fact that the breaking, strength of most of the neat cement cubes was greater than the capacity of the machine, we were able to break but one neat cube, this breaking at a pres- gure cf fifty-one thousand, four hundred (51,400) pounds. The one (1) to three (3) cubes were broken botn saturated and dry. For results of these tests see table III. Briefly, the conclusions arrived at by these tests p- were, lst, that cement suffers little from repeated af plications of lcad until about ninety ($6) per cent of its breakin: ws load is reached, and that the saturation wf, of one (1) to thees (3) cement mixtures with water weakens thom about twenty-five (25) te thirty (290) per cert, while 4-9 4 = . = 4 . “~ ao < &. see : 4 % ~ y + eae + 4 saturation of neat Cemerzt has no efrfeer after cettinr. No. 10 VAX. Stress. 5 8# s7H + Miin. Stress. oOo OF O O TABLE I. No. of ap- plications of lodd. 5,000 50,000 25 ,O00 19,000 10,500 141 6 ,O0CO 152 Ultimate Strength. Lee! 29 aH 210" Affe. 28 30 2 da. Character o- Briquette. 1 1 i 1 tied j—~ a es to to to tc Lo to to Astna. Atlas. Aetna. Atlas. Aetna. Counter stopréed on this one ané consequently the number of applicaticns vec computed frem time. This seens Q@lse tc be an exception to the senreral mn cf results. No. PrP wo ND FE on ¢ AV. TABLE II. Average age S82 days. 1 to 3. Neat. dry. wet. NO. ary. wet. 310# 1943 L 650# 5338 2534 214: 220# 1924 265# 1944 1&7? 14.53; Z200# 154# 236; 153% 240! 153/ 2398 1665# Z 4. 5 6 AVerage Tne wet is equal to 71% cf 7 o mixtures. Thc wet Ls eaual GO bre ay AL 4SO## 69 557.5:; 5457 cary in strength for for neat mixtures. TABLE III, 8" cubes, average ace 36 days. lL to 3. Neat. Noe et. bry. 1 8, 000%; 11 , 800% 51,4007 2 8,250; 9,150% 3 7,000." 11,150." AV .7,750 1c, 700. ras Tre web is oqual to 72% of the dry in strength for - n Sop & woh te O nbMuUric eds 2 aL Fe =a Le gee, i EES TE te Va aS OTA aes Needs | ei GNVCARDOGZO “ AC bo006E4 i ae wre 2c - . rr ———, = _. pO