Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 105
PROCEEDINGS OF THE NINTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY Edited by:
Paper 7
Comparison of Turbulent Models applied to the Aerodynamics of a High Speed Train C. Paz, C. Gil, E. Suárez and J. Porteiro
Industrial Engineering School, University of Vigo, Spain , "Comparison of Turbulent Models applied to the Aerodynamics of a High Speed Train", in , (Editors), "Proceedings of the Ninth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 7, 2014. doi:10.4203/ccp.105.7
Keywords: high speed train, Fluent, turbulent models, Strouhal number.
Summary
In this paper the aerodynamics of a high-speed train has been studied with
computational dynamics simulations. Three turbulent models have been compared:
k-epsilon (realizable), k-omega (SST) and detached eddy simulation (DES with kepsilon
realizable). The train used is the TVG duplex at 300 km/h, and it has been
discretized with a structured grid with a y+ between 30 and 300. Different time steps
have been employed. The software used for this purpose has been Ansys Fluent on a
cluster of Intel ® Xeon ® Quad -Core E5530 2.4 GHz processors. The aim of this
research is to quantify the influence of the unsteady phenomena and to analyse the
effect of the turbulent model used on the size of the generated vortexes, the final
drag coefficient and other characteristic aerodynamic parameters. Moreover, the
stability, convergence level, and accuracy of each model have been compared. The
results show differences in the behaviour of the turbulent models used, and the
influence of the unsteady. The results obtained agree with previous general
conclusions about the versatility and range of application of the k-epsilon and komega
models and the higher capacity of DES model to capture details associated
with detachment processes and the wake.
purchase the full-text of this paper (price £20)
go to the previous paper |
|