Isogeometric topological shape optimization using dual evolution with boundary integral equation and level sets. (January 2017)

• Record Type: Journal Article
• Authors: Lee, Seung-Wook; Yoon, Minho; Cho, Seonho
• Is Part Of: Computer aided design. Volume 82(2017)
• Page Start: 88
• Page End: 99
• Publication Date: 2017-01
• Subjects: Topological Shape optimization -- Level set method -- Isogeometric analysis -- NURBS -- Boundary integral equation -- Dual evolution; Computer-aided design -- Periodicals; Engineering design -- Data processing -- Periodicals; Computer graphics -- Periodicals; Conception technique -- Informatique -- Périodiques; Infographie -- Périodiques; Computer graphics; Engineering design -- Data processing; Periodicals; Electronic journals; 620.00420285
• DOI: 10.1016/j.cad.2016.08.004
• Languages: English
• ISSNs: 0010-4485

Abstract

Abstract: An isogeometric topological shape optimization method is developed, using a dual evolution of NURBS curves and level sets; the NURBS curves feature the exact representation of geometry and the level sets help to detect and guide the topological variation of NURBS curves. The implicit geometry by the level sets is transformed into the parametric NURBS curves by minimizing the difference of velocity fields in both representations. A gradient-based optimization problem is formulated, based on the evolution of the NURBS curves. The control points of NURBS curves are taken as design variables. The necessary response and design sensitivity are computed by an isogeometric boundary integral equation method (BIEM) using the NURBS curves. The design sensitivity is obtained on fixed grids and utilized as the velocity to update the Hamilton–Jacobi equation for the level sets. To obtain the whole velocity field on the fixed grids, an interpolation and velocity extension scheme are employed. The developed method provides accurate response and enhanced sensitivity using isogeometric BIEM. Also, additional post-processing is not required to communicate with CAD systems since the optimal design is represented as NURBS curves. Numerical examples demonstrate the accuracy of design sensitivity on fixed grids and the feasibility of shape and topological optimization. Highlights: A topological shape optimization method is developed using a dual evolution scheme. Design sensitivity at fixed grids is derived using isogeometric boundary integral equation. Transformation of implicit geometry from level sets into parametric NURBS curves. NURBS geometry is used for shape variations and the level sets for topological ones. Minimization of energy functional to resolve mismatch between level sets and NURBS.