返回
A study on partial pivot ACA boundary element method for elasticity problems
DOI:10.1016/j.enganabound.2024.105803.png)
摘要
En 中文
The boundary element method (BEM) employing the partial pivot adaptive cross approximation (PACA) algorithm has been observed to experience convergence failures and reduced solution accuracy when solving elasticity problems, especially at large scales. To address this issue, this paper proposes an improved algorithm for both 2D and 3D elasticity problems. Our investigations revealed that when the normal of an element is parallel to the coordinate axes and the element is in the same plane as the source point, zero entries with regular distributions will appear in the coefficient matrix. This results in the premature satisfaction of the convergence criterion of the PACA algorithm, leading to the omission of some rows and columns of the coefficient matrix. To address this issue, this paper proposes improvements to the PACA algorithm through the Coordinate Rotation Method (CRM) and the Component Traversal Method (CTM), and validates the effectiveness of these two improved methods. Furthermore, we investigate reasons behind the decrease in accuracy when employing the CTM to solve largescale problems. We attribute this to the oscillatory nature of the estimated relative error and propose the modified PACA (M-PACA) algorithm. M-PACA not only maintains the advantages of the PACA algorithm in reducing storage space and accelerating coefficient matrix generation but also addresses its limitations, ensuring more accurate and reliable solutions for elasticity problems.
Keyword:
Partial pivot ACA
BEM
Elasticity
Convergence improvement
M-PACA algorithm
期刊
IF:
4.1
论文数:
5.8K
被引数:
9.4K
机构
引用论文
Desulfovibrio capillatus sp. nov., a novel sulfate-reducing bacterium isolated from an oil field separator located in the Gulf of Mexico
Anaerobe
IF0
A fast multipole boundary element method for solving the thin plate bending problem求解薄板弯曲问题的快速多极边界元法

