A data-driven Bayesian optimisation framework for the design and stacking sequence selection of increased notched strength laminates

A novel Bayesian optimisation framework is proposed for the design of stronger stacking sequences in composite laminates. The first to incorporate high-fidelity progressive damage finite element modelling a data-driven methodology. Gaussian process regression used as surrogate model, minimising number computationally expensive objective function evaluations. case open-hole tensile strength investigated and an example problem, considering typical aerospace constraints, such in-plane stiffness, balance plies laminate symmetry about mid-plane. includes methodology that applies constraints without jeopardising model performance, ensuring good feasible solutions are found. Three studies conducted, standard non-standard angle laminates, on-axis misaligned loading, illustrating benefits its application general tool efficiently establish guidelines.