Investigation of the use of dynamic modulus as an indicator of hot-mix asphalt peformance

True material characterization of asphalt mixtures is covered by the theory of visco-elasticity and visco-plasticity, due to the time and temperature dependent nature of the mix behavior. Thus, a mix that may perform well in a cold climatic region may exhibit permanent deformation at regions with a warmer climate depending upon the raw material types and mix composition used. This implies that the performance criteria of asphalt mixtures must be tied to climatic as well as structural aspects of the pavement system.; The major objective of this study was to demonstrate that dynamic modulus (stiffness) could be used as a performance indicator for selecting the most promising parameter for the Simple Performance Test (SPT) to complement the Superpave volumetric mix design system. The stiffness was evaluated using a variety of techniques including predicted modulus, triaxial dynamic modulus testing and ultrasonic wave propagation testing. A total of 26 mixtures were tested for three major field sites using unconfined testing in the linear visco-elastic region, and unconfined and confined testing in the nonlinear region. Research indicated that the dynamic (complex) modulus (|E*|) correlated very well with the in-situ permanent deformation of the field test sections of the three major sites. Correlation was sensitive to binder stiffness, as expected, but it was also test temperature dependent. The dynamic modulus was also found to correlate well with fatigue cracking and moderately with thermal cracking. The stiffness obtained from ultrasonic wave propagation testing was found to have a fair correlation with in-situ distresses. The unconfined testing in the linear visco-elastic region gave better correlation to rutting than confined testing using high stress levels. The measured parameters were dynamic modulus (|E*|) and phase lag (&phis;). Also, stiffness factors |E*|/sin&phis; for permanent deformation and |E*|sin&phis; for cracking were studied as analogues to the Superpave binder specification. The correlation of stiffness factor |E*|/sin&phis; to rutting varied based on test temperature/frequency used; it was very good at intermediate temperatures but was found to develop poorer correlation at high temperatures. The modulus (|E*|) was found to be a more stable parameter and was selected as the Simple Performance Test parameter for rutting, as well as for fatigue cracking. The stiffness of asphalt mixtures did not turn out to be good enough as a performance indicator for thermal cracking.; A provisional methodology of constructing stress dependent non-linear master curves for asphalt mix was developed based on bulk stress and octahedral shear stress. In addition, an initial predictive equation, similar to the predictive techniques available for the linear-visco-elastic dynamic modulus, was developed.; Finally, a conceptual methodology of developing stiffness criteria for the parameter (|E*|) is introduced using mechanistic pavement design program, which incorporates several rational mechanistic models for designing flexible pavements. Criteria included climatic and structural aspects. This study clearly shows the future possibility of such an approach to develop the final guideline for a Simple Performance Test. Final criteria will be developed using the fully validated 2002 Guide for the Design of Pavement Structures currently under development in the National Cooperative Highway Research Program (NCHRP) project NCHRP 1-37A.