Comparison of Simple Performance Test |E^*| of Accelerated Loading Facility Mixtures and Prediction |E^*|

Abstract

The dynamic modulus, |E*|, of asphalt mixture has recently drawn much attention in the asphalt community. The |E*| test has been suggested by NCHRP Projects 9-19 and 9-29 as the simple performance test (SPT) to verify the performance characteristics of Superpave® mixture design. In addition, it has been recommended as the potential quality control-quality assurance parameter in the field. The dynamic modulus, |E*|, is an input to the Mechanistic-Empirical Pavement Design Guide (M-E PDG) and also supports the predictive performance models developed as part of NCHRP Project 1-37A. The importance of this parameter on one hand and its time-consuming determination on the other hand necessitated reliable prediction of dynamic moduli of asphalt mixtures at different temperatures and loading frequencies. In this respect, a number of predictive equations have been developed. The most widely used model is the NCHRP 1-37A, which is currently used for calculation of |E*| in Levels 2 and 3 of M-E PDG. The model has a reasonably good prediction capability of high modulus values; however, it tends to over-predict the moduli at the lower spectrum. This study shows that a recent modification to the NCHRP 1-37A model has improved the overall prediction of |E*| but has not significantly improved the overprediction of the lower moduli. While the replacement of the binder viscosity, η, with the binder shear modulus, |G*|, has increased the accuracy of the model, the inclusion of the binder phase angle, δ, in the revised equation was not found to have a significant effect.

Accuracy of Predictive Models for Dynamic Modulus of Hot-Mix Asphalt

“…There are several observations in the literature that the various models lose accuracy at the low and/or high temperature extremes (Pellinen, 2001;Schwartz 2005;Dongre et al 2005;Bari and Witczak 2006;Al-Khateeb et al 2006;Azari et al 2007). Table 1 summarizes the average prediction errors for all of the models stratified by temperature.…”

“…Several studies have indicated that the Witczak |E*| models show significant scatter especially at the low and/or high |E*| modulus extremes (Pellinen, 2001;Schwartz 2005;Dongre et al 2005;Bari and Witczak 2006;Al-Khateeb et al 2006;Azari et al 2007). There are also suggestions that the Witczak |E*| predictive models are dominated by the influence of temperature and understate the influence of other mixture parameters (Schwartz 2005).…”

Comparison of Witczak NCHRP 1-40D & Hirsh dynamic modulus models based on different binder characterization methods: a case study

“…The studies conducted by Al-Khateeb et al [14] and Yousefdoost et al [15] showed that the NCHRP 1-37A model produced more accurate E* predictions compared to the NCHRP 1-40D model, particularly at the higher temperature range [. Whereas a recent [19] concluded that Hirsh and NCHRP 1-37A models generated more accurate E* values compared to the NCHRP 1-40D model based on 1032 E* measurements from 25 mixes from the state of Idaho.…”

“…Several studies focused on using laboratory measured E* data to compare and evaluate the accuracy of the two well-known Witczak models which are the NCHRP 1-37A and NCHRP 1-40D [14][15][16]. The studies conducted by Al-Khateeb et al [14] and Yousefdoost et al [15] showed that the NCHRP 1-37A model produced more accurate E* predictions compared to the NCHRP 1-40D model, particularly at the higher temperature range [.…”

Developing Master Curves, Binder Viscosity and Predicting Dynamic Modulus of Polymer-Modified Asphalt Mixtures

“…For example, at Level 1, the asphalt binder and the HMA are tested in the laboratory to measure dynamic modulus. However, the measurement of dynamic modulus in the laboratory is not always feasible because of the tedious experiments and it may also take several days to develop a single master curve [6]- [8].…”