Railway occupant passive safety improvement by optimal design

This work presents a systematic approach to the railway occupant passive safety improvement of seating layouts using optimisation methods for the identification of the best design options. The numerical model of this layout is developed using multibody models for the dummies and finite element models for the vehicle interiors, such as seats and any structural part considered as potential target surface during impact being the crash scenario is analysed using the MADYMO code. The objective is the minimisation of the injuries as represented by all relevant injury criteria identified for railway passengers, particularly head and neck, chest and leg. The design variables are seat structural characteristics, easily modified during manufacturing or maintenance without influencing cost. The methodology starts by using a meta-modelling method that supports the definition of a surrogate objective function obtained by fitting a n-dimensional surface to the results of a collection of analysis in which the values of the design variables are varied inside specific domains. A standard deterministic optimisation method is then used to solve the surrogate problem implemented in a Matlab environment. The results show that an optimum design of the interior seating layout is obtained with relevant decreases of the most critical injury criteria.