\^ .t ^. \\ ■i^^ ^ ^ IMAGE EVALUATION TEST TARGET (MT-3) /. A. 1.0 1.1 2.5 mWL mil :? i£o 12.0 Uft ^liL^ ^ 6" » Hiotograiiiic Sdences Corporation 23 WIST MAIN STRiiT VVIBSTfR,N,Y. 145M (716)a7!l-4S03 '^ ■i CIHM/ICMH Series. CIHM/ICMH Collection de microfiches. Canadian Inatituta for Hiatorical Microraproductions / Inatitut Canadian da microraproductiona iiiatoriquaa Tcehnieai and Bibliographie Not«a/NotM taehniquaa m bIbliographiquM Th« Inttitut* hat anamptad to obtain tha boat original copy availabia for fliming. Paaturaa of ttiia copy whieh may ba bibliographieally uniquo. which may altar any of tha imagaa in tha raproduction. or which may aignlfieantly changa tha uaual mathod of filming, ara chacicad balow. □ Colourad covara/ Couvartura da coulaur r~n Covoft damagad/ D Couvartura andommagia Covara raatorad and/or laminatad/ Couvartura raatauria at/ou palliculAa □ Covar titia mlaaing/ La titra da couvartura manqua □ Colourad mapa/ Cartaa giographiquaa m* coulaur □ Colourad inic (i.a. othar than biua or black)/ Encra da coulaur (i.a. autra qua biaua ou noirai r~n Colourad plataa and/or illuatratlona/ D D D D Ptanehaa at/ou illuatratlona 1% coulaur Bound with othar matorial/ Rali* avac d'autraa documanta Tight binding may cauaa ahadowa or diatortion along intarior margin/ Laraliura sarrta paut cauaar da i'ombro ou da la diatorrion la long da la marga intiriaura Blank laavaa addad during raatoration may appaar within tha taxt. Whanavar poaaibla. thaaa hava baan omittad from filming/ II aa paut qua cartainaa pagaa blanchaa ajoutiaa lora d'una raatauration appar a la a a n t dana la taxta. mala, loraqua cala 4tait poaaibla. eaa pagaa n'ont paa AtA f llmAaa. Additional eommanta:/ Commantairaa tuppMmantairaa: L'Inatitut a microfilm* la maiilaur axamplaira qu'il lui a iti poaaibla da m procurar. Laa dAtaila da cat axamplaira qui sont paut-4tra uniquaa du point da vua bibliographiqua. qui pauvant modif iar una imaga raproduita, ou qui pauvant axigar una modification dana la mAthoda normala da filmaga aont indiquto ci-daaaoua. Tha tot Tha D D Thia itam it filmad at tha raduction ratio chackad balow/ Ca document tat film* au taux da rMuction indiquA ci^daaaoua. 10X 14X 18X 22X Colourad pagaa/ Pagaa da coulaur □ Pagaa damagad/ Pagaa Pagaa andommagiaa Pagaa raatorad and/01 Pagaa raatauriaa at/ou palliculAaa Pagaa diacolourad. stainad or foxa« Pagaa dAcoloriaa. tachatiaa ou piquios Pagaa datachad/ Pagaa ditachtea Showthrough/ Tranaparanea Quality of prir Qualiti inigala da i'impraaaion Inciudaa aupplamantary matarii Comprand du matAiial suppMmantaira Only adition availabia/ Sauia Mition diaponibia □ Pagaa raatorad and/or laminatad/ Pagaa Pagaa diacolourad. stainad or foxad/ PaSaa pn Pagaa datachad/ rri Showthrough/ r~~| Quality of print variaa/ pn Inciudaa aupplamantary matariai/ r~n Only adition availabia/ oft film Orif bag tha aion othc first aion ori Tha ahal TINI whi( Map diffa antir bagii right raqu mati Pagaa wholly or partially obscured by errata slips, tissuaa. etc.. hava been refilmed to eneure the best possible image/ Lee pegee totalement ou partieilement obecurciee par un feuiilet d'errata. una pelure. etc.. ont it* filmies i nouveau da fapon A obtanir la mailleure imaga possible. 26X 30X y 12X 1«X aox a4x 28X 32X Th« copy filmtd h«r« has bMn r«produc«d thanks to tho gonorotity of: Library of the Public Arcliivas of Canada L'axamplaira fiim4 f ut roprodult grioa A la gAnArotit* da: La bibiiothAqua das Archives publiquas du Canada Tha imagas appearing hara ara ttia bast quality possibia considaring tha condition and lagibility of tha original copy and in kaaping wKh tha filming contract spacifications. Original coplas in printad papar covars ara fllmad baginning with tha front covar and anding on tha last paga with a printad or illuatratad impras- slon. or ths back covar whan appropriate. All othar original coplas ara filmed beginning on the first page with e printed or illuatratad impres- sion, and ending on the last page with a printed or illustrated impression. The last recorded frame on each microfiche shall contain the symbol — ^> (meening "CON- TINUED"), or the symbol ▼ (meaning "END"), whichever applies. Les images suhrantee ont 4tA reproduKaa avac la plus grand soin, compta tenu de la condition at da la nattetA de rexemplaire filmA, et en conformitA evec les conditions du contrat de filmage. Les exemplalres originaux dont la couverture en papier est ImprimAe sent filmAs en commenpsnt par la premier plat et en termlnant salt par la darnlAre paga qui comporta una empreinte d'impreasion ou d'lliustration, solt par la second plat, salon la cas. Tous les autres exemplalres originaux sont filmAs en commen^ent par la premiAre page qui comporte une empreinte d'Imprasslon ou d'lliustration et en termlnant par la darnlAre page qui comporte une telle empreinte. Un dea symboles suhrants apparaftra sur la darnlAre image de cheque microfiche, selon le ces: le symbols — »• signifle "A SUIVRE", le symbols y signifle "FIN". IMaps, plates, charts, stc, may be filmed at different reduction ratios. Those too large to be entirely Included In one exposure are filmed beginning In the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams Illustrate the method: Les cartes, planches, tableeux, etc., peuvent Atre filmAs A des taux de rAductlon diff Arents. Lorsque le document est trop grand pour Atre reproduit en un seul clichA, 11 est filmA A pertir de Tangia supArlsur gauche, de geuche A drolte, et de haut en bes, en prenent le nombre d'images nAcessaira. iLes diagrammes suivants lllustrent la mAthode. 1 2 3 32X 1 2 3 4 5 6 (All righti reierved). ADVANCK PROOP-(Snhjeet to remaion). This proof is sent to you for discussion only, and on tho express undorstandiDg that it is not to be used for any other purposo what- ever.— (.See Sec. 47 of the UonitituHon). Jociefi| of ^ioil ^ngineens, INCORPORATED 1887. TBANSAcnons. N.B— Tills Society, as a body, does not hold itself responsible for the facts and opinions stated in any of its publications. EXPERIMENTS ON CONCRETE MADE AT MoGILL UNIVERSITY. By Messrs. Theo. Denis, G. G. Hare and Carl Reinhardt, Students, Can. Soc. C E. Read Thursday, \9th November, 1896. Of late, monolithic works of great importance have been carried out, and every day concrete, as building material, is creeping to a foremost place. Although cement testing proper has been subjected to elaborate, .scien- tific and practical investigations, very few researches, and especially normally conducted researches, have been made on the strength and behaviour of concretes and bctons. This probably is due to the fact that fot' such experiments heavy and costly apparatus is needed. Investiga- tions on small specimens would be useless, and conditions approaching as nearly as possible to practice have to be followed. The following arc the results obtnined from a series of experiments made by students of McGill University, 1895-96 : — The object of this first series of experiments is to determine the effect of different per cents of water on the strength of the concrete. The limits were 16 and 30 per cent, of water, by weight of cement and sand, which are beyond the extremes of practice on both sides. cement. The cement used was, of course, the same brand throughout the series. It was a German Portland of good quality, .slow setting, on which separate sand tests were iniide in connection with this series. The results are tabulated below. sand. This was clean, coarse, angular dry sand of good quality, of slightly higher grade than usual practice. STONE, This was broken limestone of such size that the pieces would have passed through a ring 1^ inciies diameter. They were unscreened, and just as they came out of the breaker. Consequently a slight amount of dust was mixed with them. They had to be broken a little smaller than in actual practice. The blocks of concrete being only one cubic foot, it was thought that more accurate results would be obtained in this way. MOULDS. The moulds were made off inch plank, lined with sheet zinc. They were 5 feet long, 1 foot high and 1 foot wide, divided into four com- partments, which would mould four cubes at once, of dimensions 1x1 X 1 feet, forming spec^imens large cnougi) to investigate seriously upon. These were removed by unscrewing one side of the box and sliding them out. Care had to be taken to oil the sides of tho moulds slightly before ramming the mixture in them, to avoid trouble in getting them out. CONDITIONS OF MIXTURE AND PROPORTIONS. The proportions adopted for t\m series were ono of eeiuent, two of sand, uiid four of stones, by wcii^ht, the proportion of watorboinn bused on the wei^l.l of sand uml cenu'iit. Tlie cement und the sund were first l)ioroii<^hly mixed dry, then tlie water added gruduuliy. The stones wrrc then tlirowu on this mortar, spread out, and the whole vigourously and very thorouglily mixed. The fresh ooDCictc was thin plticod into tlie moulds und ramiucd in H to 2 inch liiyers. R.^M.MINO. The ranimcr was a block of hard wood 2 feet long by 2 x 2 inches, with a lathe turned liandle. It wns not very easy with this to ram uniformly, even throughout one block, and tliis is one of the main sources of di.sorcpancics in this scries of experiments. It was thought that a reduction of the breaking loads to a !>tandard weight of the blocks would be only fair, and would slightly improve the results. OROITPINd OF TESTS. The tests were made at one week, four weeks, and two months, and the results grouped accordingly, that is to say, the one week tests, with diiferent per cent, of water, compare between themselves, four weekg and two months likewise. Parallels between the results, at different ages, cannot be drawn on account of some specimens having been pre- pared under widely different conditions, t^or instance, the results at two months are exceedingly low as conipurcd with those obtained at one and four weeks. This is due to the fact that these two months speci- mens were the first prepared of all, and this before the cemented trough in which they were to be immersed was completed. Consequently they were kept 8 to 10 days longer than the others in the dry air of the laboratory, which seems to have had a di.^astrous effect on them. But in spite of these slight drawbacks, the annexed table shows that up to 24 per cent., the percentage of water has not a very great effect on the strength. This is an important point, for below 20 per cent, the mor- tar obtained is rather dry and very difficult to handle. But beyond this limit of 24 per cent, a greater proportion of water seems to weaken the concrete considerably. This limit is very sharply defined in the adjoining table, where an additional 2 per cent, of water, from 24 to 26 per cent., weakens the concrete by almost one-half for the one-week te-sts. It is, however, in- teresting to notice that strength is almost completely recovered with time, the four-week tests showing the weakening limit to be between 26 and 28 per cent., and the two months' between 28 and 30 per cent. So that if immediate strength be not required of the concrete structure, 28 per cent, of water will not affect the ultimate resistance if allowed to stand two months. In the parallel sand and cement tests the weak line is not so sharply defined, but yet it is sufficiently so to show that the same statement ap- plies. The tests in this case show a marked weakening between 14 and 16 per cent, of water for the one week, which strength is ultimately recovered, as is'shown by the four weeks' and two months' tests. The low limit of 14 per cent,, as compared with 24 for the concrete, is probably due to the fact that the stones of the concrete, on account of their porosity, absorb a part of the water. The table shows that the greatest density is obtained with 16 and 18 per cent. The weights of the cubes beyond this decrease up to 24 and 26 per cent., where they are again nearly equal in density to the 16 and 18 per cent, of water. Therefore this 24 and 26 per cent, seems to be the point where the best practical results are obtained, because 16 and 18 per cent, make up too dry a concrete to allow of easy handling. Another point incidentally comes up. Attention has been drawn to the poor results obtained by the same tests and reason of long exposure to dry air given. This shows up a very important point, namely, the 2 ncoesbity of covering up carefully all oonoreto and comont works ox- posed for any length of time to dry air and sun. The bad effect of these agents is plainly demonstrated, nnd it is doubtful whether much strength would ultimately have been recovered. It is also interesting to notice the results obtained by the concretes made of 1 part of cement, 2 of sand and 5 of stones, and 1 oemcot, 2 sand and 6 of stones. Tiie specimens of these compositions gave results equal to oonorctes, 1 , 2, 4, showing that for strength they are as good as the ones containing a less proportion of .stones, while being much more economical. These experiments are as yet very incomplete. But it is hoped that the researches in this subjcot will be continued, and thnt valuable inform- ation for the engineer in practice derived from them. CONORETi Pro{)ortionH by waigi Crushing Per cent, of water by weight of ce- 1 week, ment and sand. comp. tests. 16 792 18 653 20 746 22 620 24 679 i TfcbTS — COl t : 1 part ce strength per 1 weeks. 677 679 626 616 542 546 MPKESS meat, square 1 ION. 2 8ai neb. 2moE 382 fl07 007 67( S59 600 823 d, 4 stone. Average weight of sp. per c. r. 141.6 143.0 1.39.5 1.39.6 141.6 '"26 362 141.2 28 326 340 ; 138.0 30 245 Proportion by weigh 20 1 cer 20 CEHEN Proportions : 1 cemen 10 825 12 800 14 750 331 t : 1 cement, 2 703 nent, 2 sand, 6 728 T AND SAND TE t, 2 eand. 8 13 10 sat 8t0 STS. 00 11 00 361 id, 5 ne. m.5 stone. 1822 1666 1100 *16 475 1389 1777 18 396 20 400 1110 913 844 1266 1633 22 380 24 388 26 1233 1230 1000 • Line of weakness due to excess of water. McGill University, April, 1896. 3 I