Comparison of Eulerian and Lagrangian approaches for the numerical study of the concentration of micro-particles generated from a moving train

Recent studies indicate poor air quality inside the subway system due to the successive generation and accumulation of particles. Many of these particles are iron-containing airborne wear particles generated by the abrasion of brake pads. In the present study, the concentration distribution of wear particles emitted from train brake pads during train braking is investigated. Due to the unsteady three-dimensional turbulent flow caused by the train movement inside the large geometry of the subway system, this numerical modeling requires a high computational cost, especially when the Lagrangian method, which tracks every single particle, is used. So, applying the less expensive Eulerian method, which considers the particle phase as a continuum, is preferable. However, the feasibility of using the Eulerian approach instead of the Lagrangian one should be examined, especially for the large particles emitted from a moving body inside a train-induced turbulent flow field. Therefore, in the present work, the predictions of particle concentration obtained through these two methods are compared. The predictions of the Eulerian method are more continuous and uniform, while those of the Lagrangian method, which is based on the accumulation of individual particles, are more scattered. The results show that the Eulerian method can provide reasonable predictions while maintaining computational efficiency.