Integration of crosswind forces into train dynamic modelling

C. Baker, H. Hemida, S. Iwnicki, G. Xie, D. Ongaro

Research output: Contribution to journalArticle

28 Citations (Scopus)

Abstract

In this paper a new method is used to calculate unsteady wind loadings acting on a railway vehicle. The method takes input data from wind tunnel testing or from computational fluid dynamics simulations (one example of each is presented in this article), for aerodynamic force and moment coefficients and combines these with fluctuating wind velocity time histories and train speed to produce wind force time histories on the train. This method is fast and efficient and this has allowed the wind forces to be applied to a vehicle dynamics simulation for a long length of track. Two typical vehicles (one passenger, one freight) have been modelled using the vehicle dynamics simulation package 'VAMPIRE®', which allows detailed modelling of the vehicle suspension and wheel-rail contact. The aerodynamic coefficients of the passenger train have been obtained from wind tunnel tests while those of the freight train have been obtained through fluid dynamic computations using large-eddy simulation. Wind loadings were calculated for the same vehicles for a range of average wind speeds and applied to the vehicle models using a user routine within the VAMPIRE package. Track irregularities measured by a track recording coach for a 40 km section of the main line route from London to King's Lynn were used as input to the vehicle simulations. The simulated vehicle behaviour was assessed against two key indicators for derailment; the Y/Q ratio, which is an indicator of wheel climb derailment, and the Δ Q/Q value, which indicates wheel unloading and therefore potential roll over. The results show that vehicle derailment by either indicator is not predicted for either vehicle for any mean wind speed up to 20 m/s (with consequent gusts up to around 30 m/s). At a higher mean wind speed of 25 m/s derailment is predicted for the passenger vehicle and the unladen freight vehicle (but not for the laden freight vehicle).
Original languageEnglish
Pages (from-to)154-164
Number of pages11
JournalProceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit
Volume225
Issue number2
Early online date4 Feb 2011
DOIs
Publication statusPublished - 1 Mar 2011
Externally publishedYes

Fingerprint

Derailments
Wind tunnels
Wheels
Aerodynamics
Computer simulation
Vehicle wheels
Vehicle suspensions
Large eddy simulation
Fluid dynamics
Unloading
Rails
Computational fluid dynamics
Testing

Cite this

@article{737066366bf24c998c351a748fde0664,
title = "Integration of crosswind forces into train dynamic modelling",
abstract = "In this paper a new method is used to calculate unsteady wind loadings acting on a railway vehicle. The method takes input data from wind tunnel testing or from computational fluid dynamics simulations (one example of each is presented in this article), for aerodynamic force and moment coefficients and combines these with fluctuating wind velocity time histories and train speed to produce wind force time histories on the train. This method is fast and efficient and this has allowed the wind forces to be applied to a vehicle dynamics simulation for a long length of track. Two typical vehicles (one passenger, one freight) have been modelled using the vehicle dynamics simulation package 'VAMPIRE{\circledR}', which allows detailed modelling of the vehicle suspension and wheel-rail contact. The aerodynamic coefficients of the passenger train have been obtained from wind tunnel tests while those of the freight train have been obtained through fluid dynamic computations using large-eddy simulation. Wind loadings were calculated for the same vehicles for a range of average wind speeds and applied to the vehicle models using a user routine within the VAMPIRE package. Track irregularities measured by a track recording coach for a 40 km section of the main line route from London to King's Lynn were used as input to the vehicle simulations. The simulated vehicle behaviour was assessed against two key indicators for derailment; the Y/Q ratio, which is an indicator of wheel climb derailment, and the Δ Q/Q value, which indicates wheel unloading and therefore potential roll over. The results show that vehicle derailment by either indicator is not predicted for either vehicle for any mean wind speed up to 20 m/s (with consequent gusts up to around 30 m/s). At a higher mean wind speed of 25 m/s derailment is predicted for the passenger vehicle and the unladen freight vehicle (but not for the laden freight vehicle).",
keywords = "large-eddy simulation, side winds, train aerodynamics, train dynamics, VAMPIRE{\circledR}",
author = "C. Baker and H. Hemida and S. Iwnicki and G. Xie and D. Ongaro",
year = "2011",
month = "3",
day = "1",
doi = "10.1177/2041301710392476",
language = "English",
volume = "225",
pages = "154--164",
journal = "Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit",
issn = "0954-4097",
publisher = "SAGE Publications Ltd",
number = "2",

}

Integration of crosswind forces into train dynamic modelling. / Baker, C.; Hemida, H.; Iwnicki, S.; Xie, G.; Ongaro, D.

In: Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, Vol. 225, No. 2, 01.03.2011, p. 154-164.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Integration of crosswind forces into train dynamic modelling

AU - Baker, C.

AU - Hemida, H.

AU - Iwnicki, S.

AU - Xie, G.

AU - Ongaro, D.

PY - 2011/3/1

Y1 - 2011/3/1

N2 - In this paper a new method is used to calculate unsteady wind loadings acting on a railway vehicle. The method takes input data from wind tunnel testing or from computational fluid dynamics simulations (one example of each is presented in this article), for aerodynamic force and moment coefficients and combines these with fluctuating wind velocity time histories and train speed to produce wind force time histories on the train. This method is fast and efficient and this has allowed the wind forces to be applied to a vehicle dynamics simulation for a long length of track. Two typical vehicles (one passenger, one freight) have been modelled using the vehicle dynamics simulation package 'VAMPIRE®', which allows detailed modelling of the vehicle suspension and wheel-rail contact. The aerodynamic coefficients of the passenger train have been obtained from wind tunnel tests while those of the freight train have been obtained through fluid dynamic computations using large-eddy simulation. Wind loadings were calculated for the same vehicles for a range of average wind speeds and applied to the vehicle models using a user routine within the VAMPIRE package. Track irregularities measured by a track recording coach for a 40 km section of the main line route from London to King's Lynn were used as input to the vehicle simulations. The simulated vehicle behaviour was assessed against two key indicators for derailment; the Y/Q ratio, which is an indicator of wheel climb derailment, and the Δ Q/Q value, which indicates wheel unloading and therefore potential roll over. The results show that vehicle derailment by either indicator is not predicted for either vehicle for any mean wind speed up to 20 m/s (with consequent gusts up to around 30 m/s). At a higher mean wind speed of 25 m/s derailment is predicted for the passenger vehicle and the unladen freight vehicle (but not for the laden freight vehicle).

AB - In this paper a new method is used to calculate unsteady wind loadings acting on a railway vehicle. The method takes input data from wind tunnel testing or from computational fluid dynamics simulations (one example of each is presented in this article), for aerodynamic force and moment coefficients and combines these with fluctuating wind velocity time histories and train speed to produce wind force time histories on the train. This method is fast and efficient and this has allowed the wind forces to be applied to a vehicle dynamics simulation for a long length of track. Two typical vehicles (one passenger, one freight) have been modelled using the vehicle dynamics simulation package 'VAMPIRE®', which allows detailed modelling of the vehicle suspension and wheel-rail contact. The aerodynamic coefficients of the passenger train have been obtained from wind tunnel tests while those of the freight train have been obtained through fluid dynamic computations using large-eddy simulation. Wind loadings were calculated for the same vehicles for a range of average wind speeds and applied to the vehicle models using a user routine within the VAMPIRE package. Track irregularities measured by a track recording coach for a 40 km section of the main line route from London to King's Lynn were used as input to the vehicle simulations. The simulated vehicle behaviour was assessed against two key indicators for derailment; the Y/Q ratio, which is an indicator of wheel climb derailment, and the Δ Q/Q value, which indicates wheel unloading and therefore potential roll over. The results show that vehicle derailment by either indicator is not predicted for either vehicle for any mean wind speed up to 20 m/s (with consequent gusts up to around 30 m/s). At a higher mean wind speed of 25 m/s derailment is predicted for the passenger vehicle and the unladen freight vehicle (but not for the laden freight vehicle).

KW - large-eddy simulation

KW - side winds

KW - train aerodynamics

KW - train dynamics

KW - VAMPIRE®

UR - http://www.scopus.com/inward/record.url?scp=79951859258&partnerID=8YFLogxK

U2 - 10.1177/2041301710392476

DO - 10.1177/2041301710392476

M3 - Article

VL - 225

SP - 154

EP - 164

JO - Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit

JF - Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit

SN - 0954-4097

IS - 2

ER -