Scholarlink

<p>We propose a new density-based topology optimization method for applications where fluid–structure interaction (FSI) plays a significant role. The method utilizes the jump in density between neighboring finite elements to implicitly track the FSI boundary and the FSI load. A pressure penalty term is introduced into the Navier–Stokes equations to mitigate the formation of internal pressurized holes, providing a more accurate representation of the physics and enhancing manufacturability. The method is implemented using a parallelized computational framework that enables efficient optimization of large-scale 3D problems. High-resolution discretization, combined with filtering techniques, minimizes intermediate densities and achieves detailed, binary structures that accurately model the FSI load. This is then exemplified using a classic FSI benchmark (the wall problem) with different objective functions and constraints. A relevant engineering example is then shown, maximizing fluid performance with a mechanical constraint. The approach demonstrates good convergence and provides conceptually robust designs with potential for further refinement.</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

Fluid–structure interaction topology optimization using density jumps for implicit boundary representation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月24日

查看全文
  • Hampus Hederberg,
  • Carl-Johan Thore

摘要

We propose a new density-based topology optimization method for applications where fluid–structure interaction (FSI) plays a significant role. The method utilizes the jump in density between neighboring finite elements to implicitly track the FSI boundary and the FSI load. A pressure penalty term is introduced into the Navier–Stokes equations to mitigate the formation of internal pressurized holes, providing a more accurate representation of the physics and enhancing manufacturability. The method is implemented using a parallelized computational framework that enables efficient optimization of large-scale 3D problems. High-resolution discretization, combined with filtering techniques, minimizes intermediate densities and achieves detailed, binary structures that accurately model the FSI load. This is then exemplified using a classic FSI benchmark (the wall problem) with different objective functions and constraints. A relevant engineering example is then shown, maximizing fluid performance with a mechanical constraint. The approach demonstrates good convergence and provides conceptually robust designs with potential for further refinement.