Design of composite materials and structures across the scales: physical and data-driven models

The ambitious goal of reaching net-zero emissions in aviation by 2050 can only be met by science-based disruptive innovations, including those that have the potential to reduce aircraft structural weight. Based on the hypothesis that composite materials and structures are currently not used to their full potential, a vision of a systems approach for the inverse design of composite materials and structures across different spatial scales will be presented. The main building blocks of the proposed approach, linking the macrostructure of a composite structure to the microstructure of a composite material, will be discussed. At the micromechanical level, these include the generation of representative volume elements, appropriate material models for the constituents, reduced order models and associated surrogates. A recently developed mesomechanical modelling approach that simulates the propagation of the different failure mechanisms observed in composite laminates under finite strains using a new homogenization-based smeared crack formulation will be discussed. At the structural level, building on data generated analytically and on reduced representations of composite lay-ups, four machine learning algorithms are used to predict the strength of composite laminates with notches of several geometries and the corresponding statistical distribution, associated to material and geometrical variability. It will be shown that the proposed approach may lead to significant reductions of the time required to virtually certify composite aircraft structures.