American specifications for portland cement require the crushing of mortar cubes for compressive strength of cement and the breaking of cylinder specimens for compressive strength of concrete. It has been assumed that there is a reasonable correlation of mortar and concrete strengths. This has been questioned. A preliminary study of sixty samples showed a significant correlation but indicated a very low order of confidence. A more extensive research was undertaken, which is reported in this paper.
This research involved 42 job cement samples and twelve repetitions of a monitor cement. Each of the cements in mortar was subjected to the standard 3, 7, and 28-day strength tests, a repeat of the 28-day test, and two accelerated strength tests. The corresponding cements were tested in cylinders of ready-mixed concrete, sampled at both the batching plant and at the job site. Test specimens of these concretes were broken at seven days and at 28 days. Concrete from the job site was also tested after autogenous curing for two days. The same cements were used in two additional concrete batches at the research laboratory, the first using job mix proportions and the second using a standard formulation. Both batches for each cement used job sand and gravel. These latter concretes were tested after standard curing for 7 and 28 days, and other cylinders after accelerated curing under two different sets of conditions.
Much information was obtained concerning the cements, the five mortar tests, the five concrete tests, and their many relationships—mortar/mortar, concrete/concrete, and mortar/concrete. Results were subjected to recognized statistical analysis. The paper includes 25 tables of data, 13 tables of processed information, and 18 figures.
An evaluation of the testing laboratories indicated an acceptable level of competency; an analysis of the cement and concrete variations showed normal good levels of operation. These two factors provided limitations but also gave validity to the conclusions that have been drawn from the research.