Comprehensive Evaluation of Strength Criteria for Granite, Marble, and Sandstone Based on Polyaxial Experimental Tests: Fid-Bau Portal

Comprehensive Evaluation of Strength Criteria for Granite, Marble, and Sandstone Based on Polyaxial Experimental Tests

Feng, Fan / Li, Xibing / Du, Kun / Li, Diyuan / Rostami, Jamal / Wang, Shaofeng

Cubic specimens of three rocks—granite, marble, and sandstone, representing very strong rock, moderately strong rock, and weak rock, respectively—were tested under polyaxial stress. The objective was to determine the optimal strength criteria for these rocks by using the polyaxial test data. Experimental results indicated that the failure strength considerably depended on $σ_{2}$ and $σ_{3}$ for all the rocks. Seven well-known rock-strength criteria were employed to examine the true-triaxial data with regard to the predictability in practical applications, comparison between the best-fitting functions (experimental data) and theoretical (predicted) solutions, stress trajectories on the deviatoric plane, and stress trajectories on the meridian plane and $τ_{oct} - σ_{oct}$ space. A detailed analysis and comparison of these criteria showed that the Mogi–Coulomb criterion, modified Wiebols–Cook criterion, and modified Lade criterion provided a better prediction of the polyaxial strength for the three rocks in most cases, which is attributed to the high $σ_{2}$ dependence and sensitivity of the selected rocks. The Mogi 1971 and 1967 criteria could not be correlated to the strength parameters (cohesion and frictional angle) and failed to predict the real strength values of an in situ rock mass without complex polyaxial compression tests. Furthermore, the stress trajectories on the deviatoric plane illustrated their disadvantages in theoretical interpretation and numerical implementation. The inferiority of the Mohr–Coulomb criterion and Drucker–Prager (DP) criterion in describing the rock strength under polyaxial stress was verified by the higher deviation between the theoretical (predicted) strength values and experimental strength values. The nonconformity between the stress trajectories and experimental data for the Mohr–Coulomb criterion and DP criterion on the meridian plane and in the $τ_{oct} - σ_{oct}$ space validates the preceding viewpoint.