scart.dlr.de
Friday, 14. December 2018
 

Large Eddy Simulation of train aerodynamics in cross winds

Cross wind stability is not only important for the certification of high-speed trains but it is also a field of interest in applied research. Due to the effort of engineers in transportation industry and research institutions just a few accidents were caused by cross wind in the past. With increasing speed and decreasing weight of high-speed trains the investigation of cross wind stability is becoming a pressing issue. Other crucial factors impacting the danger of a train derailment are the wind conditions and the topology of the surrounding area. The magnitude of the wind component normal to the train traveling direction in relation to the weight of the train determines the cross wind stability of the train. From wing aerodynamics, it is well known that dynamic changes of the oncoming flow can lead to overshooting aerodynamic loads on the train. In this respect, critical places with dynamically changing cross winds and therefore high overturning risks are tunnel exits, bridges and embankments.
Today, the cross wind stability of a new high speed train is already taken into account in the early design phase by computational fluid dynamics (CFD). The used methods commonly solve the Reynolds-averaged Navier-Stokes (RANS) equations. It is well known, that the mainly unsteady flow physics and the cross wind induced aerodynamic loads are not reliably predicted in RANS simulations. A more promising, but also computationally more expensive technique for predicting the complex unsteady flow fields is the Large Eddy Simulation (LES) which solves the filtered Navier-Stokes equations together with a subgrid-scale turbulence model.
Currently, we are conducting LES of the flow around a concept high-speed train referred to as the next-generation train (NGT) at a low Reynolds number of Re=2.1x105 based on the model's width. The final objective is to perform a comprehensive LES study of side wind effects on realistic train geometries using various subgrid-scale turbulence models and different grid resolutions in a large Reynolds number range. So far, simulations with zero side wind and with the standard and the dynamic Smagorinsky model were performed.

 

(a) Angle of attack: 0°
(b) Angle of attack: 30°

Time averaged streamlines projected onto a plane for different angles of attack of the flow around the simplified train model.

 

Contact:

Dr. Keith Weinman
German Aerospace Center (DLR)
Institute of Aerodynamics and Flow Technology, Department Ground Vehicles
Göttingen
Phone: +49 551 709-2339

 
German Aerospace Center (DLR), Institute of Aerodynamics and Flow Technology, SCART
Bunsenstraße 10, 37075 Göttingen, Germany