Keywords: ballast, CFD, underbody flow, rough ground.

The current paper presents a procedure to consider the effect of the geometry of the railway track on the flow under a high-speed train at a low computational cost, by using the combination of a flat wall and an equivalent roughness. For the reproduction of a realistic ballasted track, the algorithm of Voronoi´s tessellations was used to generate a statistical distribution of 3D particles with the characteristics of real ballast stones. This geometry was used in simplified simulations of an air flow over a track section to determine the value of the equivalent roughness. The simulations including only the ballast layer led to an equivalent roughness of 0.072 m, whereas when including both ballast and stones the equivalent roughness increased to 0.177 m. The adequacy of these values was satisfactorily compared with experimental results. Next, the geometry of the underbelly of a high-speed train was included in the model, and different simulations were performed to compare the underflow generated with either a flat ground or a flat ground and the equivalent roughness. The results showed significant differences in the velocity profile, with higher values in the case of flat wall with no roughness (+13.5% in the middle of the section between ground and train), which affect to the calculation of the aerodynamic resistance. Differences were also revealed in the contours of wall shear stress on the track. The rough ground led to values up to 5 times higher, which were especially remarkable in the wake region. This behaviour would be especially significant regarding the ballast flight phenomena, and would affect to the estimation and prediction based on friction parameters. In addition, the simulations were repeated at different velocities, showing no influence on the Reynolds number. As a conclusion, the determination and use of the equivalent roughness as a representation of the effect of the ballast and sleepers was proven as an interesting alternative in the CFD simulations of high-speed trains, showing significant differences with respect to a flat ground and avoiding the increase of the computational costs.

download the full-text of this paper (PDF, 8 pages, 901 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description