USE OF POLYNOMIAL CHAOS EXPANSIONS IN STOCHASTIC INVERSE PROBLEMS

Abstract eng:
Structural reliability aims at computing the probability of failure of a system which is described by a mechanical model having random input parameters. Accurate predictions require that both the mechanical model and the probabilistic model of its input parameters reflect the real world. In many situations, there is few data available for the probabilistic modelling, if any. However, there are cases where the uncertainty in the input may be grasped through the observation of the scattering of the response of various identical systems (e.g. within an experimental program). In this paper, a method based on the polynomial chaos expansion is proposed that allows the analyst to infer indirectly the probability distribution function of some input parameters. The method is illustrated on the identification of the variability of Young’s moduli on three-point bending tests.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket