Development of a Multi-scale Surrogate-based Tool for Composite Property Estimation

In this study a numerical-based surrogate model is proposed to estimate the effective elastic properties of unidirectional composite lamina, while accounting for geometric and material property uncertainties at both micro and meso scales. In the multi-scale build-up nature of composites many uncertainties occur, mainly in material properties and geometric characteristics. These uncertainties present a challenge in estimating composite material properties. The currently available property estimation/homogenisation tools are mainly in two categories: analytical equations based on an assumed model configuration and finite element homogenisation methods that are more flexible and accurate, but computationally expensive. Hence, this study develops surrogate models capable of representing various uncertainties based on established numerical homogenisation. This tool significantly decreases analysis duration compared with frequent use of full FEA. Thus, represents many composite uncertainties in an efficient way. This tool is particularly useful for developing reliability-based composite structures design approaches.