In rock engineering in cold regions, the damage evolution of rocks under freeze–thaw cycles (FTCs) exhibits significant nonlinearity and discreteness, yet existing statistical damage models based on Weibull distribution struggle to accurately characterize their full-stage mechanical responses. To address this issue, this study refines and optimizes the improved Harris function, establishes a theoretical framework describing the mechanical properties of rock micro-units, and constructs a statistical damage model based on this framework. Comparative validation with Weibull distribution and power function distribution models shows that the improved Harris function performs significantly better in characterizing the damage evolution features of freeze-thawed rocks.
Wang et al. (2026) studied this question.