Evaluation of snow accumulation simulation on a train using experimental results

Abstract

Snow accumulation on a train is a severe issue in cold regions. Many numerical analyses have been conducted to identify the cause of snow accumulation and for the design of anti-snow devices. The proper design of such devices depends on the accuracy of numerical methods. However, numerical analysis of snow accumulation has not been validated based on full-scale experiments of a train affected by snow due to lack of experimental data. In this study, the numerical analysis of snow accumulation on a train was validated using recent experimental data. The Eulerian-Eulerian model was employed to evaluate two-phase flow. Four numerical inputs for the snow phase solver, namely, distribution of snow concentration in air, snow particle size, melt ratio of snow particles, and density of accumulated snow, were modeled based on the experimental environment. The experiment was successfully replicated in simulation. The computed thickness of snow accumulated on the front wheels is in good agreement with the experimental values; however, under-estimation is observed on the rear wheels. A comparison of simulation and experimental results for evaluating snow accumulation on a train bogie shows the necessity of modeling four numerical inputs.