Comparison and statistical evaluation of Marshall stability and stiffness modulus for asphalt mixtures

Abstract

The Marshall test has been used for decades as a basic approach to optimize asphalt mix design. It has gradually been replaced by new approaches over the last two decades, yet it can serve as a quick indication for predicting the deformation behavior of an asphalt mix. The stiffness of the asphalt mixture expressed in terms of its modulus, which is determined by different test methods usually at a temperature of 10-20 °C, represents one of the key factors for the mechanistic design of the asphalt pavement structure. In contrast to the Marshall test, the determination of the stiffness modulus is always based on a non-destructive principle, which also uses measurements in the linear viscoelastic region of the bituminous binder. The negative aspect of the non-destructive test for determining the stiffness of the asphalt mix is the higher demand on laboratory equipment and at the same time the use of appropriate software for the actual stiffness calculation. This is a limiting for many commercial laboratories and limits the tests to universities or research centres. For this reason, in recent years, attention has been given to verifying the possibility of an indicative prediction of the sufficient stiffness of an asphalt mix based on the Marshall test characteristics, which can be performed by essentially any road laboratory. Measurements have been made on more than 250 different types of asphalt mixtures in order to find correlations between the stiffness determined by repeated indirect tensile loading and the Marshall stability, which is the defining characteristic of the Marshall test. The results are presented in the paper.