Study on development characteristics of mining-induced fracture networks in hard rock strata of coal mines based on multifractals

● A physical similarity model for coal mining under thick-hard roof strata was established. ● Multifractal theory was utilized to investigate the development patterns of mining-induced fractures. ● Multifractal spectra dynamically quantify the complexity and stochasticity of mining-induced fractures. ● Multifractal parameters provide indicators for assessing the evolution of fractures. The evolution of mining-induced fracture networks is one of the causes triggering structural instability of overlying strata and potential safety hazards. In this paper, similarity simulation experiments and multifractal theory are combined to investigate the evolutionary characteristics of fracture networks induced by coal seam mining under thick-hard roof conditions. With the advancement of the working face, the multifractal spectrum width Δα first increases and then fluctuates downward, reaching a peak of 1.09 at 45 cm and decreasing to 0.72 at 140 cm, reflecting the staged evolution of mining-induced fractures from local inhomogeneity to global homogeneity. The spectral end-point dimension difference Δ f shows a staged evolution (generally negative), and the multifractal spectrum exhibits a left-hook shape. In the initial mining stage, the absolute value of Δ f is large, indicating strong asymmetry in fracture development. As the working face advances, Δ f approaches 0, suggesting a gradual reduction in the complexity difference between densely and sparsely fractured zones. At 135 cm, Δ f is −0.09, indicating that fracture development reaches a staged equilibrium. The spectral parameters can quantitatively characterize the evolution of mining-induced fracture networks under thick-hard roof conditions. Multifractal theory provides an effective quantitative method for revealing the evolution law of mining-induced fractures, and can serve as a reference for overburden stability evaluation and mine disaster prevention and control.