Post-buckling optimisation strategy of imperfection sensitive composite shells using Koiter method and Monte Carlo simulation

A numerical stochastic strategy for optimising composite elastic shells undergoing buckling is presented. Its scope is to search for the best stacking sequence that maximises the collapse load considering the post-buckling behaviour. Its feasibility is due to a reduced order model built for each material setup starting from a hybrid solid-shell finite element model exploiting a multimodal Koiter method. The approach has no limitations concerning geometry, boundary conditions and material properties distribution. The collapse load is evaluated using a Monte Carlo simulation able to detect the worst imperfection shape, including a posteriori the imperfections in the reduced order model. For a limited number of parameters the proposal allows to analyse all the possible layups. In the general case, it uses a Monte Carlo scanning of the design parameters with different levels of adaptability. The optimisation of curved panels, also with stiffeners, confirms the feasibility and reliability of the proposed strategy.