Correlation between the evolution of double-K fracture toughness and microstructural parameters in thermally treated granite

This study integrates Digital Image Correlation (DIC), Acoustic Emission (AE) in-situ monitoring, and quantitative microstructural analysis techniques to reveal the micromechanical control mechanisms of Double-K fracture toughness in heat-treated granite (25–800 °C). The observed evolution of the Fracture Process Zone (FPZ) by DIC, the broadband characteristics of AE time–frequency signals, and microstructural analysis collectively demonstrate that above 400 °C, grain boundary microcracks evolve from independent nucleation to networked interconnection, directly inducing a transition in granite fracture behavior from linear-elastic to nonlinear. By quantitatively characterizing microstructural parameters such as grain equivalent diameter, aspect ratio, and fractal dimension, it was found that the synergistic control of decreasing grain diameter, increasing aspect ratio, and rising fractal dimension governs the attenuation of Double-K toughness. Based on microstructural parameters including grain morphology and microcrack networks, an Artificial Neural Network (ANN) prediction model was constructed, achieving high-precision prediction of Double-K fracture toughness.