Time of Sampling Strategies for Asphalt Pavement Quality Assurance

Brenda McCabe, Simaan AbouRizk, Jim Gavin
2002 Journal of construction engineering and management  
The cost of quality assurance programs in asphalt road construction is high in part because of the need for daily testing. This paper demonstrates the methodology used to investigate the timing of sampling and its effect on the quality assurance program. Fourteen lots over two highway paving projects were tested twice, during construction on a daily basis, and after construction was complete. Three quality measures were compared, namely asphalt content, degree of compaction, and aggregate
more » ... ion. The test data during construction was obtained from two sources: the consultant's quality assurance (DQA) and the contractor's quality control (CQC). One-way ANOVA tests and T-tests were used to compare the means. In addition, the arithmetic difference between the means was reviewed to understand how the time of sampling might affect contractor payment adjustments. BACKGROUND Government agencies at all levels are working to reduce their operating costs by reevaluating their procedures and efficiency. Alberta Transportation & Utilities (AT&U), in the Province of Alberta, Canada, worked with University of Alberta to review some of their policies related to quality assurance testing of newly constructed asphalt pavements. Several issues were considered including the number of samples taken [McCabe et al. 1999], the time at which the samples were taken, and alternative nondestructive technologies for evaluating Quality Assurance (QA) measures. This paper is concerned with the second of these three studies, namely the time of sample taking. Quality Assurance (QA) measures are based upon test results of samples collected on a lot by lot basis. In AT&U specifications, a lot is generally defined as representing a full day of hot mix production and placement. The current practice involves the collection of core samples for each lot as it is completed with further testing in a laboratory setting for each of the QA measures. A study was undertaken to examine the difference between samples extracted at the completion of construction and samples taken during construction. As the cost of the
doi:10.1061/(asce)0733-9364(2002)128:1(85) fatcat:sjifkmanmnbbjh4xeahydomveq