this report describes work carried out in a site laboratory to estimate the repeatability and reproducibility of two methods for analysing fresh concrete, namely, the constant-volume method (ram - rapid analysis machine) developed by the cement and concrete association and the pressure-filter method devised by bavelja of messrs sandberg. the following estimates of the precision data were obtained. for the pressure-filter a repeatability value of 28 kg/m3 with two estimates for the reproducibility of 26 kg/m3 and 34 kg/m3, and for two ram's, repeatability values of 33 kg/m3 and 41 kg/m3 with a reproducibility value of 61 kg/m3. results from the pressure-filter experiment on repeatability show that contributions to the overall variability were approximately one-quarter associated with batching, a similar amount with sampling and sub-sampling, and the remaining half with testing. a similar comparison of the variabilities obtained during the ram experiment showed that the batching was greatly improved, the sampling remained much the same, but the testing variability was larger than that obtained using the pressure-filter method. these results are based on a limited amount of data, and relate only to the materials used and to the conditions on the particular site. the work has shown the need for practical site-sampling techniques; it has highlighted the problems that arise in organising multiple tests on samples from the same batch of concrete, and it has shown the necessity to provide adequate drainage facilities to deal with the large volume of cement in suspension issuing from the rapid analysis machines.(a)

Want to know more about this project?