Author(s)

M. Cid Montoya, J. D´ıaz & S. Hern´andez

Abstract

Structural designs are progressively more conditioned by uncertainty in a wide range of fields, and new designs have to meet the requirements of safety and efficiency. Probabilistic optimization is a powerful tool able to improve and optimize initial designs into others which are more efficient. In spite of the continuous growth of computational power, this kind of optimization usually turns out to be unaffordable, due to the computational cost of the methods involved. However, parallel computing and surrogate models can overcome this drawback by reducing drastically the total computational cost of the process. In this paper, a survey of the performance of several surrogate models combined with reliabilitybased design optimization is presented. Two examples are selected to show the behavior of the methods when applied to different models. Keywords: uncertainty quantification, reliability based design optimization, surrogate models.

Keywords

uncertainty quantification, reliability based design optimization, surrogate models.