An experimental study and theoretical evaluation on the effect of specimen geometry and loading configuration on recorded fracture toughness of brittle construction materials

Geometry and loading condition of cracked specimens may affect the fracture toughness results and hence the evaluation of differences in measured fracture toughness values is necessary for many construction materials. In the absence of standard methods for measuring the general mode fracture resistance of asphaltic concrete materials, in this study the fracture toughness values for such materials were measured and compared by means of three different test specimens and methods. All three fundamental fracture modes (i.e. pure I, pure II, and pure III) and related mixed-mode loading cases (i.e. I/II and I/III) were evaluated. The test specimens were edge-notched disc bend (ENDB), edge-notched diametrically compressed disc (ENDC), and semi-circular bend (SCB). During the experimental program the specimens made of asphaltic concrete were tested at two temperatures of −5 and −25 °C and analyzed using the linear elastic fracture mechanics (LEFM) principles. Also, several theoretical frameworks were compared and examined to evaluate their applicability and accuracies for predicting the trends of experimentally obtained test data. The results showed that the geometry and loading configuration can significantly affect the measured fracture toughness values under pure mode II and pure mode III. The value of mode I fracture toughness measured by three specimens was almost the same (with small differences of 4%) and equal to 0.68 MPa√m. At pure mode II, the fracture toughness evaluated by the ENDB specimen was about 20% higher than that of the SCB specimen (with averages of 0.83 and 0.68 MPa√m for the ENDB and SCB samples, respectively). For pure mode III case, the measured fracture toughness by the ENDC specimen was approximately 60% higher than that of the ENDB specimen (with averages KIIIc = 0.62 MPa√m for the ENDC and 0.42 MPa√m for the ENDB specimen). Among the investigated mixed mode fracture criteria, the GMTS criterion provided accurate predictions for KIIc/KIc ratios. Also, 3D-MTSN, and 3D-MTSED criteria were more precise than others for mode I/III predictions. The evaluations indicated that the discrepancy between the measurements is mainly due to sign and magnitude of T-stress parameter.