Multi-laboratory precision of Marshall design related tests

Abstract

The Marshall method is still the method of choice for the design of Hot-Mix Asphalt (HMA) in South Africa. During the validation of a HMA mix design, considerable variability was encountered in Marshall test results for the same mix supplied by different laboratories. The variability was of such a degree that it would influence design decisions or could lead to conflict over product acceptance. A study was undertaken to investigate the extent and consequences of the inter-laboratory variability. The results, presented in this paper, show that the tests included in the Marshall design method are subject to considerable interlaboratory variability. The variability should be taken into account during the design phase as well as during quality control and product verification. One of the deficiencies in the current standard test method is that it is lacking an effective method of calibrating the compaction effort for different hammers. The results indicate a need to calibrate Marshall hammers of all parties taking part in a project and for every HMA mix that is used. Also, a need was identified for a national proficiency testing scheme to aptly identify subsequent correct, sources of variability in the testing processes