The Material Point Method (MPM) is widely utilized for simulating boundary value problems in continuum mechanics, offering unique advantages such as eliminating mesh distortions in large-deformation problems. This study investigates and compares the application of linear shape functions, enhanced by the grid-shift technique, with quadratic B-splines in MPM with respect to the stress oscillations of the solutions. Linear shape functions are computationally efficient but suffer from the cell-crossing error due to discontinuities in gradient fields across element boundaries. The grid-shift technique mitigates this by introducing random shifts to the computational background grid, achieving smoother stress fields without additional computational costs. Quadratic B-splines, on the other hand, naturally provide \(C^1\) -continuity, reducing cell-crossing errors and producing smoother results at the cost of an increased computational overhead.

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

Comparison of Approximations Utilizing the Grid-Shift Technique and B-Splines in the Material Point Method

  • Marvin Koßler,
  • Rainer Niekamp,
  • Jörg Schröder

摘要

The Material Point Method (MPM) is widely utilized for simulating boundary value problems in continuum mechanics, offering unique advantages such as eliminating mesh distortions in large-deformation problems. This study investigates and compares the application of linear shape functions, enhanced by the grid-shift technique, with quadratic B-splines in MPM with respect to the stress oscillations of the solutions. Linear shape functions are computationally efficient but suffer from the cell-crossing error due to discontinuities in gradient fields across element boundaries. The grid-shift technique mitigates this by introducing random shifts to the computational background grid, achieving smoother stress fields without additional computational costs. Quadratic B-splines, on the other hand, naturally provide \(C^1\) -continuity, reducing cell-crossing errors and producing smoother results at the cost of an increased computational overhead.