Effects of rail models on aerodynamic characteristics of trains in crosswinds at a large yaw angle

Abstract

Rail models may be simplified at different levels in wind tunnel tests and numerical simulations. This work used delayed detached eddy simulations to analyze and compare the aerodynamic loads (drag, side force, lift, and rolling moment), surface pressure, and flow characteristics around a train model placed on subgrades with three different rail models: no-rail, simplified rail, and realistic rail. The simulations were conducted at a large yaw angle of 60°. The mean aerodynamic characteristics of the car bodies and bogies were similar with the simplified rail model and realistic rail model, while those were completely different without the rail model. The difference in aerodynamic loads between the no-rail model and the realistic rail model was as high as 53.1%, whereas the difference between the simplified and realistic rail models was only 7.7%. And for fluctuating aerodynamic characteristics of the train, there was still differences between the cases with the simplified and realistic models. Further, the rail model significantly impacted the flow (such as the velocity, turbulent kinetic energy, and vorticity) around the train, especially under the train, resulting in significant differences in the aerodynamic loads in time and frequency domains. For accurate and precise measurements of mean aerodynamic loads, incorporating a rail model (with the simplified model being sufficient) can improve accuracy by 48.0% and precision by 66.9%. When focusing on fluctuating aerodynamic loads, using a realistic rail model is essential to accurately capture amplitude-frequency characteristics. The rail model setup recommendations outlined in this work aim to enhance the accuracy and reliability of various wind tunnel tests and numerical simulations related to train aerodynamics.