pend upon the conditions under which it is to be used. It can be readily understood that it is not good engineering to insist upon a cement conforming to certain standards in all cases when at one time, for instance, it may be used as a foundation for a street pavement in dry work, and at another be laid in running water. In one instance a quick-setting cement is absolutely necessary, and in the other one that is moderately slow in taking its initial set is better. What should be done is to ascertain what the requirements of the work are and then use a cement that, as it is generally manufactured, comes the nearest to meeting these requirements. Tests should be continually made to ascertain if it is being kept up to its standard. One principle should be strictly adhered to in making tests of any kind of material: have the conditions governing the tests conform as closely as may be to those under which the material is to be used. Eliminate as much theory and uncertainty as possible, and spend neither time nor money in attaining a requirement that will never be of any benefit to the work. In actual construction cement is almost never used neat. It is first mixed with sand and is then called mortar. The common proportion for a natural cement is one part cement to two parts of sand by volume. This is, of course, purely arbitrary, but it seems to have come into general use from the fact that this mixture seems to be strong enough for the more common uses to which cement mortar is put. When a greater or an immediate strength is wanted a brand of Portland is adopted with varying proportion of sand. Some engineers indeed think that Portlands run more evenly than the naturals, and that where only a moderate strength is required the latter should be used, reducing the expense by increasing the proportion of sand. As it is the mortar that is to be used, whether in regular masonry or concrete, it is important and necessary to know the resulting volume from the mixing of cement and sand in different proportions. It should be specified, also, whether the cement is to be measured as originally packed or as poured loosely into the measuring-box. Tables Nos. 25 and 26 give the results of experiments made by L. C. Sabin, U. S. Assistant Engineer, to ascertain the amount of sand and cement required to make a cubic yard of mortar under different conditions. TABLE NO. 25. Barrels Parts of Sand to one of cu. ft. Loose Sand. Cement. Cement cu. ft. Cement packed packed 3.73 cu. ft. 3.73 cu. ft. per bbl. per bbl. Cement. per bbl. 71 lbs. per Cubic Yards 75 lbs. per Cubic Yards 80 lbs. per cu. ft. Cement packed 3.75 cu ft. All Sand weighed 100 lbs. per cu. ft., voids 37% percent. Cubic Yards Loose Sand. Loose Sand. per bbl. The above, while being very valuable as showing actual amounts of mortar to be obtained from the different mixtures of cement and sand, also emphasizes the importance of the unit to be used; as, taking the barrel of cement at 265 lbs. and the proportion of one part cement to two of sand, the tables give the following weights of cement for a cubic yard of mortar by each of the different methods: By volume, cement loose. Pounds. 660 750 970 The second method requires 13 per cent and the third almost 50 per cent more cement than the first. The plain and true inference is that the only sure way of knowing just how much cement is being used is to determine proportions by weight, or to specify that a cubic yard of mortar shall receive so many pounds of cement. This is particularly important now when so many manufacturers deliver their cement in bags by weight, and allowing a certain number of pounds for a barrel. When heavier cements, as the Portlands, are used, it is evident that there will not be so much difference in the methods employed. Cement mortar is often used in sea-water, and in preparing it considerable extra expense would be incurred in providing fresh water for the mixture. Quite a number of experiments have been made at various times and by different persons to determine the action of salt water, if used in mixing, and also when the mortar is immersed in it. Gen. Gillmore made some rectangular parallelopipeds of mortar 2×2×8 inches in vertical moulds under a pressure of 32 pounds per square inch until set. These were broken on supports from a pressure from above midway between the supports. The specimens were kept in a damp place for twenty-four hours, when they were placed in sea-water, where they remained ninety-four days, till broken. Table No. 27 gives his results. Cement 1, sand 2, by volume mixed with fresh water. Cement 1, sand 2, by volume mixed with sea-water, concentrated by heat 25 per cent.. In the report of Mr. E. C. Clarke previously referred to Table. No. 28 is given, showing the results of his investigations on this question. Mr. A. S. Cooper in a paper published in the Journal of the Franklin Institute, October, 1899, details some experiments made by him, shown in Table No. 29, to determine the effect of salt water. The briquettes were the American Society of Civil Engineers' forms, the proportions being determined by weight. They were stored in moist air for twenty-four hours and then in an immersion-tank till broken. The figures represent tensile strength per square inch in pounds. 7 days... 266 320 237 298 147 28 days.. 310 331 4 months. 6 months. 177 134 188 65 92 87 110 107 150 120 164 228 230 232 237 140 1 year.... While the actual figures given by Mr. Clarke and Mr. Cooper vary much as to the actual strength, owing doubtless to the character of the cement and the method of manipulation, they are relatively the same, there being a marked decline whenever the briquettes are immersed in salt water, especially the long-time tests with the Portland cements. Where the mixing is done with salt water and the immersing in fresh, the difference is not so striking. Although these tests show that cement mortar is weakened by the action of salt water, works have been carried on of sufficient time and extent to make it certain that the deterioration is not dangerThis becomes important in studying the action of frost on mortars, as it is customary to add salt to the water for mortar-mixing, when it must be used at low temperatures. ous. Mr. James J. R. Croes gives as a rule: "Dissolve 1 pound of rock salt in 18 gallons of water when the temperature is at 32° F., and add 3 ounces for every 3 degrees of temperature." He adds that masonry laid with such mortar stood well and showed no signs of having been affected by the frost. Mr. Alfred Noble states that a pier was built on the Northern Pacific Railroad near Duluth at a temperature varying from 0 to 20°. Portland cement was used for the mortar in proportions of 1 to 1 for face stone and 1 to 2 for backing. Salt was dissolved in the water, and the sand was warmed. The mortar froze very quickly, and several months afterwards was found to have perfectly set and to be in as good condition as that laid in milder weather. Table No. 30 gives the result of some of his experiments to determine the effect of salt upon the mortar, and Table No. 31 the combined effect of salt and freezing. The amount of salt seems to make no material difference, although the figures are slightly less for the greater quantities, and, as in the previous tables, the salt water gives poorer results than the fresh. These figures show some gain when salt water is used for the mixture and the briquettes immersed in fresh, and decided increase when they were frozen for six days and immersed in water long enough to thaw, but not a sufficient time to gain an additional set. The table would be of more value if it extended over a longer period of time. Table No. 32 is taken from a paper read before the Canadian Society of Civil Engineers in February, 1895, by Prof. Cecil B. Smith of McGill University. Set No. 1 was submerged, after 24 hours, in water of laboratory tank. |