Non-parametric shape optimization method for thin-walled structures under strength criterion

• Published in: Computer aided optimum design in engineering X Vol. 91; pp. 179 - 188
• Main Authors: SHIMODA, M, TSUJI, J, AZEGAMI, H
• Format: Conference Proceeding Journal Article
• Language: English
• Published: Southampton WIT 01.01.2007; W I T Press
• Subjects: Applied sciences; Criteria; Design; Design optimization; Design parameters; Exact sciences and technology; Hilbert space; Lagrange multiplier; Mechanical engineering. Machine design; Parameterization; Shape optimization; Thin wall structures; Geometrical shape; Gradient; Optimization method; Parametric programming; Non parametric method; Minimization; Adjoint method; Parameterization; Traction; Geometrical model; Lagrange multiplier; Optimal solution; Hilbert space; Curvature; Thin wall; Mathematical programming
• ISBN: 1845640705; 9781845640705
• ISSN: 1746-4498; 1743-3509
• DOI: 10.2495/op070171

This paper presents a numerical optimization method for shape design to improve the strength of thin-walled structures. A solution to maximum stress minimization problems subject to a volume constraint is proposed. With this solution, the optimal shape is obtained without any parameterization of the design variables for shape definition. It is assumed that the design domain is varied in the in-plane direction to maintain the curvatures of the initial shape. The problem is formulated as a non-parametric shape optimization problem. The shape gradient function is theoretically derived using the Lagrange multiplier method and the adjoint variable method. The traction method, which was proposed as a gradient method in Hilbert space, is applied to determine the smooth domain variation that minimizes the objective functional. The calculated results show the effectiveness and practical utility of the proposed solution in solving minmax shape optimization problems for the design of thin-walled structures under a strength criterion.