Non-destructive testing of concrete strength: statistical control

Some problems of statistical control of concrete strength in non-destructive testing are considered. The procedures for rejection of gross errors and accounting for the actual law of concrete strength distribution are given. In the non-destructive control, the reliability and accuracy of a test result, fi, as well as the probability of a specified concrete strength, f′ c , should be the same as in testing with the use of a standard method (specimen testing). Proceeding from this concept, known acceptance criteria are analysed and those recommended by the author are described. The use of the latter permits a considerable reduction of the consumer's risk. Procedures for specifying a required average strength, f cr , are also analysed. In the use of the proposed acceptance criteria, the value of f cr may be reduced in many cases, and hence the cement content may be decreased.