Traditional fracture criteria only consider the singular terms in the Williams expansion and neglect higher‐order term coefficients, resulting in significant discrepancies between predicted and measured values. To advance the understanding of rock fracture mechanisms, this study introduces a generalized maximum tangential stress (GMTS) criterion by incorporating crack‐tip stress singularity and T‐stress effects. Laboratory experiments on shale specimens with varied crack inclinations validate the GMTS‐predicted initiation angles and fracture toughness. Results demonstrate that traditional MTS criteria, which ignore T‐stress and rely solely on singular terms in William’s expansion, erroneously predict straight propagation along prefabricated cracks in pure mode I fractures. By contrast, GMTS reveals T‐stress‐induced path deflection, resolving prediction discrepancies exceeding 30% at β < 45° and significant errors in fracture toughness under mode II‐dominated loading. For mixed‐mode I–II fractures, crack deflection angles and toughness exhibit T‐stress dependency: negative T‐stress reduces initiation angles while enhancing toughness, whereas positive T‐stress amplifies angles but degrades toughness. This work establishes T‐stress as a critical regulator of fracture behavior, reconciling theoretical limitations of classical criteria and providing quantitative validation through integrated modeling and experimentation. The GMTS criterion can better predict the initial propagation angle and fracture toughness of mixed‐mode I–II cracks in rocks.
Liu et al. (Thu,) studied this question.