Scholarlink

Phase-field models (PFMs) have been demonstrated to accurately predict complex crack patterns such as crack branching, and merging. However, these models come at a significant computational expense. We propose an adaptive mesh refinement (AMR) algorithm in the present study to efficiently solve the thermomechanical fracture problem. This investigation focuses on three widely recognized PFMs—AT1, AT2, and PF-CZM, to implement them in a single codebase, and utilize them to investigate the effectiveness of the AMR algorithm in solving the thermomechanical fracture problem. We analyze the five different stages of mesh adaptivity, including Solve, Estimate, Mark, Refine, and Check. The adoption of the mesh adaptive algorithm results in a substantial reduction in the total simulation time, ranging from 5 to 90 times faster, contingent upon the type of the problem, in comparison to simulations employing non-adaptive mesh refinement a priori.

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

A Computationally Efficient Method for Modeling Thermo-Mechanical Fracture Using Phase Field Method

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • U. Meenu Krishnan,
  • Abhinav Gupta,
  • Rajib Chowdhury

摘要

Phase-field models (PFMs) have been demonstrated to accurately predict complex crack patterns such as crack branching, and merging. However, these models come at a significant computational expense. We propose an adaptive mesh refinement (AMR) algorithm in the present study to efficiently solve the thermomechanical fracture problem. This investigation focuses on three widely recognized PFMs—AT1, AT2, and PF-CZM, to implement them in a single codebase, and utilize them to investigate the effectiveness of the AMR algorithm in solving the thermomechanical fracture problem. We analyze the five different stages of mesh adaptivity, including Solve, Estimate, Mark, Refine, and Check. The adoption of the mesh adaptive algorithm results in a substantial reduction in the total simulation time, ranging from 5 to 90 times faster, contingent upon the type of the problem, in comparison to simulations employing non-adaptive mesh refinement a priori.