Optimal And Robust Design Of NonlinearStructures Under Uncertain Loads UsingStatistical Optimization Method
Qiang Yu, Takayoshi Kashiwamura, M. Shiratori, Toshiki Yamada · 1999
The Statistical Design Support System**™ is based upon the design of experiments (orthogonal array design) and FEM analysis. The orthogonal array is used to define a set of finite element analyses (FEA), and by the analysis of variance of the FEM results, the correlation between the objective functions and the design factors can be expressed as a set of chebyshev's orthogonal functions. And then, the other design processes are treated using these approximate functions (estimation expressions). In this study, the authors proposed a basic concept to evaluate the dispersion due to an uncertain variable (uncontrollable variable), and a method of optimal and robust design of nonlinear structures by SDSS, and gave an example of the optimal and robust design for an artificial ankle joint. It was concluded that the SDSS can be used as an optimum design tool taking into consideration its robustness.