Summary
Tests on laboratory-made concrete of various mix proportions, types of cement, and ages at test indicate that, for a constant degree of control, the standard deviation within a batch is directly proportional to the mean strength of the concrete. A similar relation applies to the standard deviation between batches. These relations do not hold good for very low strengths—below about 1,500 lb/in2.
An examination of field data, carefully separated according to the degree of control exercised, suggests that for site-made concrete also the standard deviation increases with the increase in the mean strength. Thus the present practice of designing mixes on the basis of a constant coefficient of variation, as recommended by Road Note No. 4, is believed to be satisfactory.
