Preliminary design optimization of stiffened panels using approximate analysis models
Autor: | Theodore F. Johnson, Luciano Lamberti, Raphael T. Haftka, Satchi Venkataraman |
---|---|
Rok vydání: | 2003 |
Předmět: |
Numerical Analysis
Mathematical optimization Engineering stiffened shell structures business.industry Applied Mathematics Probabilistic-based design optimization Multidisciplinary design optimization General Engineering Preliminary design optimization Finite element method Sizing Engineering optimization Conceptual design Aerospace business Global optimization |
Zdroj: | International Journal for Numerical Methods in Engineering. 57:1351-1380 |
ISSN: | 1097-0207 0029-5981 |
Popis: | Preliminary sizing of structural components in the conceptual design phase of new aerospace vehicles requires a large number of structural optimizations. Use of detailed finite element models for global optimization of stiffened shell structures is not affordable. Designers have to compromise between the complexity in the structural analysis model (e.g. small models or linear analysis) and that of the optimization method (local versus global optimization). When designing new concepts where good initial designs are not available, global design optimization needs to be performed, and this is not currently possible with complex models. In this paper, we discuss the use of simple analysis methods for performing global optimization of stiffened shell structures. We demonstrate the efficiency of using approximate analysis models by using PANDA2 for a practical design trade-study. Often approximate models are based on assumptions that are not satisfied throughout the design space. Hence, such models require constraints that will keep the designs in regions where the analysis assumptions are valid. We illustrate with examples situations where such design constraints result in sub-optimal designs. We conclude that performing global optimization using approximate methods can permit a greater exploration of the design space than the use of local optimization with complex models. Copyright © 2003 John Wiley & Sons, Ltd. |
Databáze: | OpenAIRE |
Externí odkaz: |