Optimisation of railhead profiles for curved tracks using improved non-uniform rational B-splines and measured profiles

Abstract

Optimisation of railhead profiles using improved non-uniform rational B-Splines (NURBS) and measured profiles is presented for curved tracks to minimise wear. Measured geometry of rail profiles have been parameterised for the NURBS model. These parameters have then been decomposed as constants and variables to represent any other generalised rail profile. The variable parameters have been defined as design variables for the optimisation modeling method that minimises wear of the railhead profile suitable for curved rail track. A genetic algorithm optimisation method was formulated incorporating the parameters of the measured rail profiles with improved NURBS curves described as genes for high and low rails in curved tracks and combined them as chromosomes. The method can handle optimisation of any unique rail profile, although examples given in this paper refer to only the 60 kg/m Chinese Rail CH60. Optimal rail profiles are shown to mitigate wear of wheel-rail contact under the operation of a vehicle passing through a curved track. The optimal shape was shown to minimise contact stresses, increase rolling radius difference and conicity for improved passenger comfort and vehicle safety.