CEMENT TESTING 44. The thorough testing of cement, as it is done for the largest public works, should properly be done in a professional testing laboratory. A textbook of several hundred pages has recently been written on this subject. The ultimate analysis and testing of cement, both chemically and physically, is beyond the province of the ordinary engineer. But the ordinary engineer does have frequent occasion to obtain cement in small quantities when testing in professional laboratories is inconvenient or unduly expensive. Fortunately it is possible to make some simple tests without elaborate apparatus which will at least show whether the cement is radically defective and unfit for use. It is unfortunately true that an occasional barrel of even the best brand of cement will prove to be very inferior to the standard output of that brand. This practically means that in any important work, using a large quantity of cement, it is not sufficient to choose a brand, as the result of preliminary favorable tests, and then accept all shipments without further test. Several barrels in every carload should be sampled for testing. It is not too much to prescribe that every barrel should be tested by at least a few of the simpler forms of testing given below. The following methods of testing are condensed from the progress report of the Committee on Uniform Tests of Cement, as selected by the American Society of Civil Engineers. The statements may therefore be considered as having the highest authority obtainable on this subject.
45. Sampling. The number of samples that should be taken depends on the importance of the work but it is chiefly important that the sample should represent a fair average of the contents. The sample should be passed through a sieve having twenty meshes per linear inch, in order to break up lumps and remove any foreign material. If several small amounts are taken from different parts of the package, this also insures that the samples will be mixed so that the result will be a fair average. When it is only desired to
determine the average characteristic of a shipment, the samples taken from different parts of the shipment may be mixed, but it will give a better idea of the uniformity of the product to analyze the dif ferent samples separately. Cement should be taken from a barrel by boring a hole through the center of one of the staves, midway between the heads, or through the head. A portion of the cement can then be withdrawn, even from the center, by means of a sampling iron similar to that used by sugar inspectors.
40. Chemical Analysis. Ordinarily, it is impracticable for an engineer to make a chemical analysis of cement which will furnish reliable information regarding its desirability, but the engineer should understand something regarding the desirable chemical constituents of the cement. It should be realized that the fineness of the grinding and the thoroughness of the burning may have a far greater influence on the value of the cement than slight variations from the recognized standard proportions of the various chemical constituents. Too high a proportion of lime will cause failure in the test for soundness or constancy of volume, although a cement may fail on such a test owing to improper preparation of the raw material or defective burning. On the other hand, if the cement is made from very finely ground material and is thoroughly burned, it may contain a considerable excess of lime and still prove perfectly sound. The permissible amount of magnesia in Portland cement is the subject of considerable controversy. Some authorities say that anything in excess of S per cent is harmful, others declare that the amount should not exceed 4 per cent or 5 per cent. The pro portion of sulphuric-anhydride should not exceed 1.75 per cent. It may be considered that the other tests of cement are a far more reli able indication of its quality than any small variation in the chemical constituents from the proportions usually considered standard.