My key interest is in the multi-scale progressive damage modelling of composites and the development of variable-kinematics finite element models.

Currently, I work on a data-driven multi-scale modelling framework for shell elements. Large composite aero-components are typically modelled using shell finite elements. Conventional shell modelling attributes the linear elastic response to Classical Laminate Theory (CLT), and the material nonlinearities to failure indices and criteria. This approach falls short in predicting the structural behaviour and often results in overly conservative designs. The reason being that real composite parts contain complex sub-laminate geometry, such as defects and features, that cannot be considered using CLT. Not only that, the progressive failure responses in composite feature the 3D interactions of different damage modes, which are not encapsulated in conventional failure indices.

This project develops a multi-scale modelling framework for shell elements. In this framework, the sub-laminate length-scale is modelled high-fidelity RVE solid models, where the full extent of geometric and material nonlinearites are considered. These RVE models are solved, and their responses are homogenised as nonlinear material models (ABD matrices and resultants) that are used by the macroscopic shell elements, in which the effects of damage on membrane, bending and transverse shear stiffness can be accounted for.