Influence of Test Geometry, Temperature, Stress Level, and Loading Duration on Binder Properties Measured Using DSR

During the last two decades, several test methods and performance-based test specifications have been developed to select asphalt binders for use under different traffic and environmental conditions. The test equipment and methods are continually being improved to enhance the reliability with which materials can be selected for optimal performance. One of the procedures to assess rutting susceptibility of asphalt binders is the multiple-stress creep recovery (MSCR) test. Although the MSCR is a promising test method, test variables such as stress level, test geometry, and the number of cycles (total time) need to be carefully evaluated. This paper presents findings from a laboratory investigation of some of these variables. The dynamic shear rheometer (DSR) was used to evaluate the mechanical response of the asphalt binder by varying the stress level, test geometry, and temperature. The results were analyzed on the basis of recoverable and permanent strain to estimate the effect of changing geometry and increasing stress level. Results show that the test geometry used with the DSR (parallel plate versus cone and plate) has a significant effect on the measured properties, particularly at high stress levels or longer durations of loading. Standard parallel plates with a 1 mm gap may allow the binder to flow because of instability and lead to change in the specimen geometry during a repeated loading test. This tertiary-like behavior, which primarily happens at higher number of cycles (more than 10) might be mistaken for tertiary flow and can be misleading. Although decreasing the gap to 275 micrometers will increase the confining stress and alleviate the problem to some extent, using a cone and plate instead of a parallel plate is highly beneficial to reflect the true material behavior. The findings point out the importance of considering specimen geometry type and size effect while interpreting test results. With cone and plate geometry, the results also demonstrate that tertiary flow is an important part of binder behavior at high stress levels and extensive loading times. These considerations are particularly important when correlations between binder and mixture permanent deformation, as measured in the new flow number test, are under investigation.