Combining digital image correlation and probabilistic approaches for the reliability analysis of composite pressure vessels

Abstract

The accuracy of reliability analysis of composite solutions depends on the robust estimation of the uncertainties associated with the mechanical properties of this material. On this basis, we propose a methodology able to exploit the full field strain data provided by the digital image correlation approach in order to extract the probabilistic density functions of the mechanical properties. These probabilistic density functions are complemented by a global sensitivity analysis based on the polynomial chaos expansion and a random variable approach, based on the latin hypercube sampling method, with the aim of obtaining a stochastic evaluation of composite pressure vessels.