Structural similitude design for a scaled composite wing box based on optimised stacking sequence

Testing appropriately-designed scaled-down structures instead of the full-size prototype structure is beneficial for quickly understanding the prototype structural behaviour in a cost-effective way. In this study, a novel approach is proposed to design scaled composite structures that can be used to predict the structural responses (e.g. static bending, modal behaviour, and compressive buckling) of the prototype. The objective is to overcome the main drawback of the conventional design method, which tends to result in low accuracy of the prototype prediction when certain variables cannot be appropriately scaled due to manufacturing constraints (e.g. ply thickness). In the present work, a set of scaling laws being independent of boundary conditions were firstly derived for plates and beams respectively based on their governing equations. The genetic algorithm (GA) was then applied to help design the stacking sequence of the scaled models, accommodating the mismatch in similarity conditions resulting from the manufacturing constraint in ply thickness. This GA-based design method was demonstrated to be effective in designing scaled plate, I-beam, and stiffened plate models, with improved accuracy in predicting the prototype structural behaviour compared with the conventional method. The application of this new design method was also extended to an A320 size wing box structure, validating its robustness for complex structures.