Observations on current performance tests and prediction models for asphalt concrete mixes

This paper presents a brief summary and review of 15 years of progress on pavement related research. The pavement research was conducted in the United States as part of the Strategic Highway Research Program (SHRP) and National Cooperative Highway Research Program (NCHRP) as well as other related research projects. In particular this paper reviews Hot Mix Asphalt (HMA) pavement tests either developed or refined at Arizona State University as part of NCHRP Project 9-19 "Simple Performance Tests" and NCHRP 1-37a "2002 Pavement Design Guide", now commonly referred to as the Mechanistic Empirical Pavement Design Guide (MEPDG). These tests were originally used to characterize hot mix using unmodified asphalt; however additional testing was conducted in Arizona and Portugal using crumb rubber modified asphalt. The properties obtained from these tests for all types of hot mixes are used as inputs into mechanistic models to predict pavement performance. SHRP recommended a suite of tests to characterize mix properties to predict permanent deformation (rutting) and cracking distresses caused by either traffic loading or climatic conditions or the combination of the two. Recommended tests included the repetitive simple shear test at constant height to predict rutting, four point flexural bending fatigue test to predict fatigue cracking, Indirect Tension Test (IDT) and Temperature Stress Restrained Specimen Test (TSRST) test to predict low temperature cracking. Most of this research was originally based on laboratory specimens prepared by the rolling wheel compactor. Later NCHRP Project 9-19 was tasked to develop simple performance test(s) using specimens prepared with the SHRP gyratory compactor, which subsequently lead to the recommendation of uniaxial / triaxial simple performance tests. At about the same time the NCHRP Project 1-37a involved taking existing mechanistic models as well as the SHRP and simple performance to predict the cracking, rutting and roughness over a ten to twenty year design period for varying traffic loading and climatic conditions. The purpose of this paper is to highlight some test developments and identify some correlation's made using the above tests and mechanistic models in Arizona and Portugal. Inferences from the test results and models are drawn which potentially may lead to improvements.