Strength theory is the basic theory for calculating and designing the strength of engineering materials in civil, hydraulic, mechanical, aerospace, military, and other engineering disciplines. Therefore, the comprehensive study of the generalized nonlinear strength theory (GNST) of geomaterials has significance for the construction of engineering rock strength. This paper reviews the GNST of geomaterials to demonstrate the research status of nonlinear strength characteristics of geomaterials under complex stress paths. First, it systematically summarizes the research progress of GNST (classical and empirical criteria). Then, the latest research the authors conducted over the past five years on the GNST is introduced, and a generalized three-dimensional (3D) nonlinear Hoek-Brown (HB) criterion (NGHB criterion) is proposed for practical applications. This criterion can be degenerated into the existing three modified HB criteria and has a better prediction performance. The strength prediction errors for six rocks and two in-situ rock masses are 2.0724%-3.5091% and 1.0144%-3.2321%, respectively. Finally, the development and outlook of the GNST are expounded, and a new topic about the building strength index of rock mass and determining the strength of in-situ engineering rock mass is proposed. The summarization of the GNST provides theoretical traceability and optimization for constructing in-situ engineering rock mass strength.

To verify the novel method of achieving a true-triaxial stress path with the pseudo-triaxial apparatus, a series of drained and undrained tests were carried out for the identical scheme with pseudo-triaxial apparatus and true-triaxial apparatus respectively. The differences between the two types of tests were quantified. The results show that the novel method effectively achieved the true-triaxial stress path by controlling the loading ratio of the pseudo-triaxial apparatus. The relationships of q - epsilon 1 and eta - epsilon s measured by the two apparatuses had a higher similarity which decreases slightly with the b increase. When 0 <= b < 0.5, the slope of the critical state line measured by both apparatuses was almost identical. When 0.5 <= b <= 1, the slope of the critical state line measured by the novel method was slightly lower, but the biggest change was within 10% compared with the two Mohr-Coulomb criteria, the peak strength measured by the two apparatuses was distributed near the criteria, indicating the feasibility and rationality of the novel method. The tests show that the novel method greatly enriches the test range of pseudo-triaxial apparatus, which not only simplifies the process of soil 3D testing but also reduces the test cost.