Metamodel-Based Uncertainty Quantification for the Mechanical Behavior of Braided Composites

Abstract

This chapter presents an uncertainty quantification framework for triaxially braided composites simulation, dealing with the stochastic stiffness prediction via numerical multiscale analysis. Efficiency is achieved by using various metamodeling techniques, such as neural networks, polynomial chaos expansion and Kriging modeling. Uncertainties accounting for material and geometric randomness are propagating through the scales to the final scatter of the mechanical properties of the macroscale. Information about the stochastic input and the dominating uncertain parameters is offered via application of a variance-based global sensitivity analysis. All methods employed in this work are non-intrusive, hence the framework can be used for all sorts of composite materials and numerical models. The need for realistic uncertainty quantification is highlighted.