In building walls, or other masses of concrete, large pieces of stone, old bricks, chalk, etc., are often packed in for the sake of economy.

Care should be taken that the lumps thus inserted do not touch one another. They should be so far apart, and clear of the face, that the concrete may be well rammed around them.

Where chalk or lumps of absorbent material are used, care must be taken that they are not exposed so as to absorb wet or moisture, otherwise they will be liable to the attacks of frost, and may become a source of destruction to the wall.

Proportion Of Ingredients

The materials to form concrete for ordinary work are generally mixed together in a dry state, the proportion of each being determined by custom, rule of thumb, or experience.

In former days, when lime concrete was more used, a common mixture was,

1 quicklime. Or 1 quicklime.

2 sand. 7 Thames ballast (which contains 5 or 6 gravel, broken stone, or brick. sand and shingle).

The same proportions were for some time blindly adhered to, irrespectively of the nature of the materials used.

The best proportions for the ingredients of a cubic yard of concrete to be made with any given materials may, however, always be arrived at by ascertaining the contents of the voids in a cubic yard of the aggregate (without sand), and adding to the latter such materials as will make mortar of the best quality and in sufficient quantity to perfectly fill those voids. Where the concrete is not required to be of the best quality, as for example in the backing of heavy walls, the mortar may be made poorer accordingly.

If the aggregate contain sand (as in the case of gravel or ballast), the sand should be screened out of the sample before the voids are measured, and the amount of sand thus screened out will be deducted from that required for the mortar which is to form the matrix of the concrete.

In practice a little more mortar than is actually required to fill the voids should be provided, in order to compensate for imperfect mixing and waste.

Thus, supposing the aggregate available for making concrete to be clean shingle containing 9 cubic feet of voids per cubic yard, a first-rate concrete can be made by adding to each cubic yard of aggregate 4 cubic feet of Portland cement and 8 cubic feet of sand, which will make 102/3 cubic feet of 2 to 1 Portland cement mortar (see p. 205), or a little more than sufficient to fill the 9 cubic feet of voids in the shingle.

Again, if the aggregate were ballast, itself containing 41/2 cubic feet of sand in each cubic yard, and 41/2 cubic feet of voids besides, it would be necessary to add to each cubic yard 4 cubic feet of Portland cement as before, but only 31/2 cubic feet of sand, because there are already 41/2 cubic feet of sand in the aggregate, making 8 cubic feet of sand altogether, which, with the 4 cubic feet of Portland cement, will make 102/3 cubic feet of 2 to 1 Portland cement mortar, or more than sufficient to fill the 9 cubic feet of voids that there are in the ballast without the sand.

If the concrete is not required to be of the first quality, as for example in the backing of heavy walls, the mortar may be made poorer accordingly.

Thus, to make a poorer concrete, with clean shingle for the aggregate, to each cubic yard may be added 12 cubic feet of sand and only 2 cubic feet of Portland cement, making 6 to 1 mortar (or mortar of 6 sand to 1 cement) in more than sufficient quantity to fill the voids.

The chief point to be considered is the quality of the mortar in the concrete. This should be arranged as above described so as to be good enough for the work in which it is to be used, and sufficient in quantity to thoroughly fill the voids of the aggregate, with a little to spare in case of imperfect mixing.

It is a curious thing that engineers have not agreed upon any short way of describing concrete so as to indicate at once its proportions and quality.

As recently as in November 1886 Mr. Hayter said at the Institute of Civil Engineers : - 1

1 M.P.I.C.E., vol. lxxxvii. p. 161.

"In describing concretes it was customary to say that they were mixtures consisting of so many parts of gravel or shingle and sand to 1 part of cement. But in Mr. Hayter's experience two concretes so described might mean admixtures of two different strengths. Thus, assuming a concrete that might be called a 6 to 1 mixture. In specifiying such one engineer might say the concrete was to consist of 1 part Portland cement and 6 parts of gravel or sand of approved quality. Another engineer might say that the concrete was to consist of 1 part Portland cement, 4 parts of gravel or shingle without any sand, and 2 parts of sand."

He pointed out that these two concretes, though both called 6 to 1 concretes, were very different, for in the first there is 1 part of Portland cement to 6 of gravel containing sand, whereas in the second, after the sand is mixed with the shingle, it merely fills the interstices, and the concrete is composed of 1 part Portland cement to 4 of shingle containing sand in its interstices (see p. 213).

The present practice as to briefly describing concrete differs and is often very misleading. It is necessary, therefore, to be very careful in specifications to state exactly how much of each ingredient, shingle or stone, sand and cement is required.

It has been stated that concrete can be made equally good without sand, but sand is a necessary ingredient in all cases where the concrete is required to be waterproof, and it is also desirable on account of strength. Recent experiments have shown that with different aggregates - the proportion of cement, etc., being the same - the concrete made with sand was far stronger both as regards transverse and tensile stress and crushing than that without sand.

Concrete is much used for paving, being made with the very best Portland cement into slabs, and then laid like ordinary stone flags.

For this purpose it is preferable to use an aggregate, such as shingle or granite, much harder than the matrix, and to use very little sand in the latter.

As the matrix becomes worn away, the pebbles of the aggregate project slightly, making the surface slightly rough, and therefore less slippery, and at the same time the matrix is protected from further wear.