Two-level numerical optimization of ride comfort in railway vehicles

Abstract

This article addresses the interaction between and tuning of different mechanical subsystems in railway vehicle design. Application of numerical methods in a two-level optimization process, involving multi-body vehicle dynamics simulation and finite-element structural analysis, is proposed. The focus is on vibration ride comfort and tuning of dynamic properties of the vehicle, running gear, and carbody structure. As a background, a brief overview of multi-disciplinary and structural optimization methods is given. The selected solution based on the loosely coupled collaborative optimization approach, here implemented as a two-level structure, is presented. Numerical examples are provided to illustrate the choice of optimization algorithms suitable for this type of resonant systems and to demonstrate the performance of the numerical procedures on a realistic engineering problem. The leading coach of a three-car train-set is successfully optimized in terms of mass reduction and reduced natural frequency requirements, despite an infeasible initial design.