Three-Dimensional Nonlinear GZZ Strength Criterion of Sandy Dolomite: Verification by True Triaxis Test and Analysis of Intermediate Principal Stress Effect
摘要
In response to the complex mechanical characteristics of sandy dolomite in the Central Yunnan Water Diversion (CYWD) Project, this study constructs a multi-parameter strength prediction model (the power function relationship between UCS and SH, Is(50), Vp, ρ, n, with an R2 value ranging from 0.593 to 0.973) through in-situ tests (Schmidt rebound, point load, wave velocity test) and laboratory tests (uniaxial compressive strength and triaxial compression). It also proposes a three-dimensional constitutive model parameter calibration method based on the Generalized Zhang-Zhu (GZZ) criterion. True triaxial tests demonstrate that the intermediate principal stress (σ2) exerts a significant influence on the brittle-ductile transition and strength properties of sandy dolomite, as evidenced by a 133% increase in peak strength of σ1 at σ2 = 0.8 σ1. By calibrating parameters (GSI = 84.57, a = 0.5, s = 0.4285), the theoretical yield surface predicted by the GZZ criterion shows a high degree of agreement with the experimental data (average error < 5%), thereby confirming its applicability in a three-dimensional stress field. The parameter sensitivity analysis indicates that a has a significant effect on the yield surface morphology (the yield surface is expanded twofold when a = 0.8), while s only causes the overall translation. Engineering case studies further indicate that optimized parameters (a = 0.7, s = 0.45) effectively reduce the extent of the plastic region in surrounding rock by 23%, providing a robust tool for engineering stability analysis of deep sandified rock mass engineering.
Highlights Intermediate stress σ2 increases peak strength by up to 133% and mitigates brittle failure. The multi-parameter model establishes a strong correlation between UCS and parameters such as SH, Is(50), and Vp, with an R2 value reaching as high as 0.973. Calibrated GZZ criterion demonstrates < 5% error and reduces the plastic region by 23% with optimized parameters. Validated in deep tunnel excavation significantly enhances stability analysis for sandy dolomite engineering.