95. Tests for Compressive Strength.—The compressive strength of concrete has commonly been tested on 6-inch or 12-inch cubes, and much of the available data is based upon such tests. The desir ability of eliminating the corners has led to the use of cylindrical specimens, which are found much easier to make and handle satis factorily. The use of a test piece whose height is greater than its lateral diameter is also found advantageous. Blocks of concrete under crushing loads usually yield through shearing on surfaces making angles of about. 60° with the horizontal, and by using blocks whose heights are twice the diameter sufficient freedom is allowed for such action. Tests have seemed to indicate that the strength of blocks varies somewhat with the ratio of height to diameter, cubes showing 25 to 35 per cent more strength per square inch than cylin ders whose heights are twice their diameters. Blocks of greater relative height show a further loss of strength, but to much less degree, those having a height five times the diameter giving an average strength about 90 per cent of those with a ratio of 2 to 1.
A Committee of the American Concrete Institute recommended I that a cylindrical test piece be used, whose height is twice the diameter, and diameter at least four times the maximum size of the aggregate. Cylinders S inches in diameter and 16 inches high are used as standard by the Joint Committee on Concrete, though cylinders 6 inches by 12 inches are also sometimes used.
Forms made of cast iron, with a metal base, machined smooth and true, are made by manufacturers of testing apparatus; they are somewhat expensive, but are far more satisfactory in use than lighter forms. Forms made of sheet metal with flanges to bolt or clamp together may be used without the metal base by placing them on plate glass.
Sampliruj the materials should he carefully done to insure obtain ing fair samples for the tests. To secure an average sample, the method of quartering is sometimes which consists in taking shovelfuls of material from different parts of the pile, mixing them together and spreading out on the mixing surface. The resulting layer of material is then divided into quarters, two opposite quarters are shoveled away, the other two quarters are mixed again and the operation repeated until a sample of the size desired is obtained. When the materials are not uniform and vary in different parts of the supply, it may be desirable to take separate samples from each part and make comparative tests.
Measurement of materials should be by weight in making tests. When the proportions used in work are by volume the volume weight of the materials should first he ascertained, and the weight proportions for the test pieces determined accordingly.
_Mixing should he done on an impervious surface. The cement and sand should first be rnixecl thoroughly, to a uniform color, and spread evenly on the mixing surface. The coarse aggregate should
then be spread over the clry mixture of sand and cement and the whole turned several times dry. Water may then be added in a crater and the mass turned and wet until it is thoroughly worked to uniform consistency.
On important work, it may often be desirable to test the con crete as it is being used, by making test pieces from the concrete as delivered for placing in the work.
Forming the Block.—The concrete should be tamped into the forms in layers so as to bring the mortar to the surfaces and leave no open spaces around the edges. After the concrete has set, the top of the block may be smoothed by leveling with cement paste or mortar, or with plaster of Paris, and a piece of glass pressed down on top and left until the mortar has set.
Storage.—In making tests for purposes of comparison, the test pieces should be kept moist while they are hardening. Blocks left in dry air do not gain strength normally.
Testing.—In making the tests, spherical bearing blocks should be used and care taken to permit the adjustment of the test pieces to uniform bearing, properly centered.
96. Tensile and Transverse Strength.—There are very few data available concerning the tensile strength of concrete, as it is not used it is subject to direct tension, and this strength is of com paratively little interest. The tensile strength is called into play in unreinforced beams, but the action is quite different from that of a direct pull. The results of tests indicate that the tensile strength of concrete commonly varies from about one-fifteenth to one-twelfth of the compressive strength.
The transverse strength is dependent upon the tensile resistance of the material, and plain concrete is therefore a weak material for use in beams. On account of this weakness, concrete is seldom used for beams without reinforcement, and in the computation of reinforced beans, the resistance of the concrete on the tension side of the beam is neglected.
The few data available indicate that the modulus of rupture for plain concrete beams varies from about one-eighth to one-fifth of the compressive unit strength, or that it is approximately twice the strength in direct tension. These values are based upon the appli cation of the common theory of flexure, and the usual formulas for homogeneous materials. The difference between the modulus of rupture and tensile strength may be partly accounted for by the fact that the modulus of elasticity is not constant and the neutral axis does not remain at the gravity axis, but changes in position, approach ing the compression side of the beam as the load increases, so that the actual tension does not reach the computed modulus of rupture.