I-II mixed fracture characterization of hot mix asphalt under tensile test using notched semi-circular specimens

I-II mixed mode fracture is a common fracture pattern of asphalt pavement under the coupled tension and shearing loads. To address the roles of reclaimed asphalt pavement (RAP) and loading effect, this study proposed a novel test method to characterize the I-II mixed fracture toughness of two types of asphalt mixture. Within the novel tensile test, direct tension load could be applied on the bottom of the notched semi-circular specimens at -10 degrees C. I-II mixed fracture could be generated by varying the notch angle. Fracture parameters, including the stress intensity factors of mode I and mode II fracture (Krc and Krrc) and the T stress were obtained for specimens with five notch angles, including 0 degrees, 15 degrees, 30 degrees, 45 degrees and 60 degrees, respectively. Results indicated that from 0 degrees to 30 degrees, pure mode I to pure mode II fracture could be generated. Under the same loading rate, compared to mixture A, mixture B presented superior fracture toughness in terms of mode I and mode II stress intensity factors (Kr and Krr), which was attributed to the fact that the inclusion of RAP in mixture A deteriorated the fracture resistance. Loading rate played positive effect on the fracture toughness. From 1 mm/min to 5 mm/min, the improvements of Krc were 52% and 33% for mixture A and B, respectively. For Krrc, the increase of loading rate enhanced the pure mode II fracture resistance by 30% and 20% for mixture A and B, respectively. Several fracture criteria were adopted for the fracture prediction. Comparison was performed between the test data and the predictions. It was found that compared to MTS criterion, the inclusion of T stress in GMTS criterion made the prediction closer to the test data. Compared to GMTS and GEMTS criteria, GEMTS criterion could provide better estimates for I-II mixed mode, indicating considering the effect of fracture mode on the critical distance (r,) could make the prediction with a higher reliability.