Scholarlink

<p>Investigating rock behavior at the mineral scale provides critical insights into macroscopic failure mechanisms. This study developed a modular micromechanical modeling framework integrating mineral parameter characterization, three-dimensional (3D) digital core generation, and GPU-accelerated material point method (MPM) simulations to analyze granite under uniaxial compression. The framework uses nanoindentation tests, MPM simulations, and deep neural networks to determine mineral elastoplastic parameters. A convolutional neural network generates high-fidelity 3D digital cores representing mineral spatial distribution and geometry. The elastic modulus and uniaxial compressive strength derived from the numerical simulations and experimental results are in good agreement. The mineral’s spatial distribution and morphology significantly influence rock strength. The texture model, which preserves intergranular interlocking and heterogeneity, yields 13.4% higher strength than the Voronoi model. Crack evolution analysis identifies biotite as the primary site for microcrack initiation and mineral interfaces as preferential crack propagation paths. Mineral interface properties affect macroscopic behavior. Cohesion and friction angle influence compressive strength, whereas the interfacial elastic modulus is positively correlated with macroscopic stiffness. This work elucidates multiphase, coupled damage evolution in granite and provides an effective method for predicting macroscopic responses from microstructure.</p><p><b>Highlights</b><UnorderedList Mark="Bullet"> <ItemContent> <p>High-fidelity three-dimensional (3D) digital granite cores with realistic mineral structure and mechanical properties were developed and validated.</p> </ItemContent> <ItemContent> <p>The 3D digital granite cores were generated through the self-similarity of mineral structure.</p> </ItemContent> <ItemContent> <p>The texture model outperformed the Voronoi model, yielding 13.4% higher load capacity and demonstrating the effect of microstructural interlocking on macroscopic strength.</p> </ItemContent> <ItemContent> <p>Interface cohesion and friction angle govern granite’s uniaxial compressive strength, whereas its modulus influences stiffness.</p> </ItemContent> </UnorderedList></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mineral Information-Driven 3D Simulation of Meso-Mechanical Behavior and Fracture Process in Granite

  • 期刊论文

  • 发表日期：

    2026年3月23日

查看全文
  • Yanmin Zhou,
  • Binwei Xia,
  • Sisong Zhang,
  • Lei Zhou,
  • Chengpeng Zhang,
  • Xingguo Zhang

摘要

Investigating rock behavior at the mineral scale provides critical insights into macroscopic failure mechanisms. This study developed a modular micromechanical modeling framework integrating mineral parameter characterization, three-dimensional (3D) digital core generation, and GPU-accelerated material point method (MPM) simulations to analyze granite under uniaxial compression. The framework uses nanoindentation tests, MPM simulations, and deep neural networks to determine mineral elastoplastic parameters. A convolutional neural network generates high-fidelity 3D digital cores representing mineral spatial distribution and geometry. The elastic modulus and uniaxial compressive strength derived from the numerical simulations and experimental results are in good agreement. The mineral’s spatial distribution and morphology significantly influence rock strength. The texture model, which preserves intergranular interlocking and heterogeneity, yields 13.4% higher strength than the Voronoi model. Crack evolution analysis identifies biotite as the primary site for microcrack initiation and mineral interfaces as preferential crack propagation paths. Mineral interface properties affect macroscopic behavior. Cohesion and friction angle influence compressive strength, whereas the interfacial elastic modulus is positively correlated with macroscopic stiffness. This work elucidates multiphase, coupled damage evolution in granite and provides an effective method for predicting macroscopic responses from microstructure.

Highlights

High-fidelity three-dimensional (3D) digital granite cores with realistic mineral structure and mechanical properties were developed and validated.

The 3D digital granite cores were generated through the self-similarity of mineral structure.

The texture model outperformed the Voronoi model, yielding 13.4% higher load capacity and demonstrating the effect of microstructural interlocking on macroscopic strength.

Interface cohesion and friction angle govern granite’s uniaxial compressive strength, whereas its modulus influences stiffness.