Railway rolling noise prediction under European conditions

Several theoretical models are available for predicting railway rolling noise. However, the TWINS (Track-Wheel Interaction Noise Software) model, based on the mechanics of vehicle and track interactions, agrees well with field measurements. Full-scale validation experiments have shown that the TWINS model gives reliable predictions under European conditions. In this paper, the development of the TWINS calculation procedure in MATLAB as part of RailCRC Project No.1-105 Improved Noise Management is detailed. Details of the prediction modules, Graphical User Interface and validation under European conditions are provided as well as the sensitivity analysis of the MATLAB model (called ‘RRNPS’, meaning Railway Rolling Noise Prediction Software). The predictions of this model are compared with those from the TWINS model and results show the RRNPS model gives reliable predictions under European conditions. Subsequently this model is used to predict how pad and ballast dynamic properties affect the sound powers from the wheel and rail.