Lightweight variable angle stiffened prisms with cutouts: An ANN-GA approach for optimized buckling resistance

Composite structures with variable angle designs offer greater flexibility compared to conventional straight designs, particularly in stiffened laminate cylinders with cutouts. However, these structures introduce significant design complexity due to intricate geometries and laminate sequences. This paper proposes a novel lightweight and better buckling resistant composite decagonal prism structure with cutouts, incorporating Variable Angle Stiffeners (VAS) with Double-Double (DD) laminates to achieve lighter, homogenized, easier-to-produce, modular manufacturing and the best mechanical properties, while reducing design variables. To enhance computational efficiency in optimizing a large number of design variables, this paper proposes an Artificial Neural Network-Genetic Algorithm (ANN-GA) optimization framework. This framework efficiently identifies optimal geometric and stacking sequence parameters for improved buckling performance. It also demonstrates faster optimization and superior solutions compared to conventional finite element analysis (FEA) methods. The structure and framework were validated through three case studies with varying boundary and loading conditions, using a conventional stiffened laminate prism as a benchmark. Results indicate that the proposed variable angle stiffened laminate prism offers enhanced buckling resistance, and the ANN-GA framework outperforms traditional optimization approaches in terms of efficiency and solution quality.