A Thermo–Hydro–Mechanical Coupled Shear Damage Constitutive Model for Rock‐Concrete Interfaces

ABSTRACT The shear mechanical behavior of rock‐concrete interfaces (RCIs) critically governs the stability of support structures in tunnel and underground engineering. However, these interfaces represent the weakest link in composite systems, especially under high geothermal conditions. In such environments, post‐construction concrete is subjected to high‐temperature curing in direct contact with hot surrounding rock, leading to thermal damage. During tunnel operation, geothermal water infiltration induces further interface degradation through thermo‐hydro‐mechanical (T–H–M) coupling. To address these challenges, this study develops a novel shear damage constitutive model for RCIs under T–H–M coupling by integrating statistical damage theory with the JRC‐JCS joint strength criterion. A systematic parameter calibration methodology is established to ensure model accuracy and reliability. Comprehensive shear tests on concrete‐granite interfaces demonstrate the model's strong capability to replicate the full‐range shear stress–strain behavior under T–H–M coupling. Furthermore, the model was further modified to overcome two critical limitations of compressive‐type statistical damage approaches: (1) inadequate representation of nonlinear concave curvature during pre‐peak yielding, and (2) poor characterization of residual strength evolution post‐peak. Based on the developed shear damage constitutive model, the evolution of interface damage during high‐temperature curing in the construction stage and the subsequent T–H–M‐induced degradation operation are systematically analyzed. These findings provide theoretical foundations for optimizing tunnel support design in geothermal environments, particularly by identifying early microcrack initiation thresholds and guiding damage‐informed reinforcement strategies. Summary Shear tests identify four‐phase T–H–M behavior: compaction, elasticity, hardening, and softening. Novel T–H–M‐coupled shear damage model integrating JRC‐JCS criterion with statistical damage theory. Modified statistical formulation resolves pre‐peak nonlinearity and post‐softening limitations. Experimental validation shows <2% error in peak strength and residual stress prediction. Critical damage thresholds identified (D = 0.32 for microcracks, D = 0.6 for macro‐fractures).