A simplified model for structural stiffness and crashworthiness optimisation of composite fuselages

In this paper, a multi-objective optimisation for the initial design of crashworthy composite structures is proposed. The focus is on the optimisation model with its minimisation of inhomogeneous deformation capability. It promotes homogeneous contribution of all crash elements, as a representative of structural crashworthiness. For the first time, homogeneous contribution has been identified as a metric for crashworthiness of composite structures and been transferred into a mathematical expression. The structural model uses discrete elements for very efficient computation in combination with the genetic algorithm NSGA-II. Clustering as a machine learning technique is applied to the Pareto set of solutions in order to identify representative structural solutions. The approach uses positioning of elements and the shape of the spring characteristic of these elements as variables. The method enables the simultaneous consideration of static and crash loads, which is demonstrated by a case study featuring a composite aircraft fuselage substructure. So far, static and crash loads have been only considered separately, but never at the same time. The novelty of this approach is in the combination of an appropriate simplified modelling technique and a new formulation of the optimisation model. The proposed optimisation is beneficial in improving the crashworthiness of composites, as optimisation of the geometry and material behaviour enables a non-linear response to be obtained in an otherwise brittle material.