This section is from the book "Cyclopedia Of Architecture, Carpentry, And Building", by James C. et al. Also available from Amazon: Cyclopedia Of Architecture, Carpentry And Building.
Concrete is composed of a mixture of cement, sand, and crushed stone or gravel, which, after being mixed with water, soon sets and obtains a hardness and strength equal to that of a good building stone. These properties, together with its adaptability to monolithic construction, combined with its cheapness, render concrete very useful as a building material.
Theoretically the proportioning of the sand and cementing material should be done by weight. It is always done in this way in laboratory testing. The volume of a given weight of cement is quite variable according as it is packed or loosely thrown in a pile. The same statement is true of sand. Since a barrel of Portland cement will increase in volume from 10 to 40 per cent by being merely dumped loosely in a pile and then shoveled into a measuring box, a contractor will frequently attempt to take advantage of this expansion by measuring the cement loose rather than by using the proportions as indicated by the original volume in the packed barrels. To a less extent the same uncertainty exists regarding the condition of the sand. Loose, dry sand occupies a considerably larger volume than wet sand, and this is still more the case when the sand is very fine.
The general principle to be adopted is that the amount of water should be just sufficient to supply that needed for crystallization of the cement paste; that the amount of paste should be just sufficient to fill the voids between the particles of sand; that the mortar thus produced should be just sufficient to fill the voids between the broken stones. If this ideal could be realized, the total volume of the mixed concrete would be no greater than that of the broken stone. But no matter how thoroughly and carefully the ingredients are mixed and rammed, the particles of cement will get between the grains of sand and thus cause the volume of the mortar to be greater than that of the sand; the grains of sand will get between the smaller stones and separate them; and the smaller stones will get between the larger stones and separate them. Experiments by Prof. I. O. Baker have shown that, even when the volume of the mortar was only 70 per cent of the volume of the voids in the broken stone, the volume of the rammed concrete was 5 per cent more than that of the broken stone. When the theoretical amount of mortar was added, the volume was 7.5 per cent in excess, which shows that it is practically impossible to ram such concrete and wholly prevent voids. When mortar amounting to 140 per cent of the voids was used, all voids were apparently filled, but the volume of the concrete was 114 per cent of that of the broken stone. Therefore, on account of the impracticability of securing perfect mixing, the amount of water used is always somewhat in excess (which will do no harm); the cement paste is generally made somewhat in excess of that required to fill the particles in the sand (except in those cases where, for economy, the mortar is purposely made very lean); and the amount of mortar is usually considerably in excess of that required to fill the voids in the stone. Even when we allow some excess in the above particulars, there is so much variation in the percentage of voids in the sand and broken stone, that the best work not only requires an experimental determination of the voids in the sand and stone which are being used; but, on account of the liability to variation in those percentages, even in materials from the same source of supply, the best work requires a constant testing and revision of the proportions as the work proceeds. For less careful work, the proportions ordinarily adopted in practice are considered sufficiently accurate.
On the general principle that the voids in ordinary broken stone are somewhat less than half of the volume, it is a very common practice to use one-half as much sand as the volume of the broken stone. The proportion of cement is then varied according to the strength required in the structure, and according to the desire to economize. On this principle we have the familiar ratios 1:2:4, 1:2 1/2:5, 1:3:6, and 1:4:8. It should be noted that in each of these cases, in which the numbers give the relative proportions of the cement, sand, and stone respectively, the ratio of the sand to the broken stone is a constant, and the ratio of the cement is alone variable, for it would be just as correct to express the ratios as follows: 1:2:4; 0.8:2:4; 0.67:2:4; 0.5:2:4.
VANDEVENTER BUILDING, KNOXVILLE, TENN. Leon Beaver, Architect. Reinforcing bars furnished by Expanded Metal and Corrugated Bar Co.