Determining Number of Specimen Replicates in Support of Pavement Construction Using AASHTO TP 125

The bending beam rheometer (BBR) has long been used to test low-temperature properties of asphalt binders in cold regions. More recently, a newly released TP 125 standard by the American Association of State Highway and Transportation Officials [AASHTO. 2016. Determining the flexural creep stiffness of asphalt mixture beams using the bending beam rheometer (BBR). AASHTO TP 125. Washington, DC: AASHTO.] has been introduced to the asphalt industry to advance the testing method for the low-temperature performance of asphalt mixtures during pavement construction. This paper presents a series of statistical analyses associated with material testing to determine the number of specimen replicates in support of pavement construction using the AASHTO TP 125. Two asphalt mixtures using performance grading (PG) 64-34 and PG 58-28 binder were cut and trimmed to mixture beams in the laboratory. Statistical analyses including confidence interval, coefficient of variance (CV), bootstrapping, and randomization analysis were performed to observe the dispersed distribution of asphalt samples so as to determine the number of mixture beam replicates in order for BBR test results to be valid. Based on statistical analyses, a minimum of 5-10 specimen replicates depending on analysis purposes in the BBR test are recommended. However, highway agencies and contractors can decide the number of specimen replicates to be used as long as the CV of selected specimens is less than 0.2. To demonstrate the applicability of using asphalt mixture beams in the BBR for in situ quality control of asphalt pavements at low temperatures, two asphalt paving projects located in the western region of the United States were selected. Asphalt mixtures from each paving location were collected and tested using the BBR, and the recommended number of specimen replicates to obtain stiffness values and determine their corresponding CV was determined. Based on BBR results, there was an anomaly on a sampling location that indicated an increase in stiffness compared with the rest of the sampling locations. The paper concludes that the statistical analysis results and testing procedures presented in the paper are feasible and capable of measuring material properties (stiffness), settling a dispute, and characterizing an immediate low-temperature characteristic of constructed asphalt pavements. The methodology presented in the paper can be used by highway agencies to perform quality control activates for asphalt paving projects.