|
| IGA-SIMP method based stress-constrained topology optimization of continuous structures |
| Received:September 19, 2017 Revised:November 30, 2017 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx20170919002 |
| KeyWord:Topology optimization of structures isogeometric analysis IGA-SIMP method stress constraints P-norm function |
| Author | Institution |
| 刘宏亮 |
大连理工大学 工程力学系 工业装备结构分析国家重点实验室, 大连 |
| 杨迪雄 |
大连理工大学 工程力学系 工业装备结构分析国家重点实验室, 大连 |
|
| Hits: 1860 |
| Download times: 1013 |
| Abstract: |
| This paper establishes an isogeometric analysis (IGA)-SIMP method for the stress-constrained topology optimization of continuous structures.Based on the popular SIMP model,the unified NURBS (non-uniform rational B-spline) function is utilized for geometry modeling,structural analysis and design parameterization,which can well integrate structural analysis with optimal design.By virtue of high-order continuous NURBS basis functions,the isogeometric analysis improves the computational accuracy of stress and its sensitivities,and thus enhances the credibility of optimization results.For handling the numerous local stress constraints,a STM (stability transformation method) correction-based P-norm stress constraint strategy is proposed to overcome the iteration oscillation and convergence difficulty of topology optimization problem.Several representative topology optimization examples of 2D plane stress problems demonstrate the effectiveness and accuracy of the present design method.Numerical examples,including the stress-constrained volume minimization design and the mean compliance minimization design with the constraints of both volume and stress,indicate that the STM-based stress correction strategy is able to suppress the iteration oscillation in volume minimization design problems and realize stable convergence.By comparison,the optimization iterations for the mean compliance minimization design with the constraints of both volume and stress are more stable,and thus the exact stress correction strategy is suitable for addressing this issue. |
|
|
|