Accurate Estimation of Tensile Failure Parameters via Theoretical–Experimental Correlation and Implications for Mine Fracture Prevention

Abstract

The experimental determination of tensile failure parameters is constrained by sample scarcity, methodological inconsistency in testing, and the inherent complexity of fracture behavior. This study develops statistically robust, lithology‐specific parameter relationships for coal‐bearing strata via integrated theoretical analysis and experimental validation. Statistical regression of over 660 measurements across eight lithologies mitigates data scatter and scarcity. Theoretical derivation and dimensional analysis yield a physically interpretable and theoretically coherent K IC – σ t relationship, and linear regression validates well‐defined quantitative correlations across the majority of rock types. The simplified linear model facilitates engineering parameter estimation and cross‐validated parameter inference. The derived model is integrated into the input calibration framework of FDEM, and simulation responses exhibit strong agreement with experimental observations. Additionally, the correlation model serves as an objective benchmark for evaluating K IC testing protocols, enabling the objective identification of optimal methods and the exclusion of high‐variance methods unsuitable for routine application. This approach bridges a critical gap between the scarce, inconsistent direct σ t measurements in coal strata and the demand for reliable tensile fracture data in engineering applications.