Measurement of train aerodynamic phenomena in operational conditions

This is an EPSRC project in which different research methodologies such asfull-scale and model-scale measurements and advanced computer-based simulationsare used to measure train aerodynamic phenomena. The major aim of the projectis "to investigate and measure a range of aerodynamic phenomena observed inreal train operation, both relative to the train and relative to a fixed pointat the trackside, and to compare how such effects match model scalemeasurements and various types of CFD calculation".


The specific objectives of the project are as follows.

  • To carry out a range ofconventional physical model tests and computational fluid dynamics (CFD) trialsto measure the velocity and pressure field around a 200kph Class 43 train,commonly referred to as the HST (shown below in Figure 1 and Figure 2).
  • To instrument the NetworkRail High Speed High Speed Measuring Train (which is based on Class 43vehicles) to measure aerodynamic phenomena during its operation on all mainline routes in the UK. This includes measurement of cross wind forces, trainpassing pressures and pressure transients in a wide variety of tunnels.
  • To carry out a series ofmeasurements at a trackside location on the Western Main Line (a route wherethe Class 43 is widely used), such that the pressure and velocity fields closeto such trains can be measured under operational conditions (example is shownbelow).
  • To compare the results ofthe model scale and computational trials to the on-train measurements todetermine the adequacy of the predictions of established test methods to givethe cross wind forces, moments, displacements and on-train pressure transientsfound in operational conditions.
  • To compare the results ofthe model scale and computational trials to the on-track measurements todetermine the adequacy of predicted slipstream velocities and tracksidepressure fields with those measured in operational conditions.
  • To thus investigate thefundamental hypothesis outlined above and to develop possible methods in whichtesting and codification procedures could be changed to make them moreadequately reflect operational conditions, and thus to make the design processfor new trains less conservative and restrictive.
  • Start of the project: 01 April 2012
  • End of the project: 31 March 2016

CFD model

Figure 1 CFD model

Full-scale measurement train

Figure 2 Full-scale measurement train


There are a variety of aerodynamic effects associated with train design andoperation - the determination of aerodynamic drag, the effect of cross winds ontrain stability, pressure transient loading on trackside structures, thephysiological effect of tunnel pressure transients, the effect of trainslipstreams and wakes on waiting passengers and trackside workers etc. Themagnitude of these effects broadly increases as the square of the vehicle speedand thus with the continued development of high speed train lines aerodynamiceffects will become more significant in terms of design and operation. Now itcan be hypothesised that the techniques that have been used to predictaerodynamic effects in the past (wind tunnel and CFD methods) are likely todetermine magnitudes of pressures, velocities, forces etc. that are higher thanthose observed in practice, where other effects - such as track roughness,variability in meteorological conditions etc. are likely to usually obscureaerodynamic effects to some extent and, because of this, some of the currentdesign methodologies are unnecessarily restrictive and/or conservative. Thusthe aim of the current project is to investigate and measure a range ofaerodynamic phenomena observed in real train operation, both relative to thetrain and relative to a fixed point at the trackside, and to compare how sucheffects match model scale measurements and various types of CFD calculation,and thus to test the validity, or otherwise, of the above hypothesis. This willbe achieved through the instrumentation of the Network Rail High SpeedMeasuring Train to measure aerodynamic effects, as the train carries out itsnormal duty cycle around the UK rail network. Also trackside instrumentationwill be installed at a suitable site that will allow off-train phenomena to bemeasured. Calibration wind tunnel, CFD and moving model tests will be carriedout in the conventional way for comparison with data measured at full scale.The full scale, model scale and computational trials will be carried out byexperienced RFs and will provide data for two doctoral studies, one of whichwill investigate how the train based measurements of cross wind forces,pressure transients etc compare with those predicted by conventionalmethodologies, and one of which will investigate how the track sidemeasurements compare with conventional test results. The investigators willsynthesise the results and make recommendations for future aerodynamic testmethods.


Principle Investigator


Research Fellows

Research Students

  • Justin Morden
  • Martin Gallagher



1- Full Scale measurements

The slipstream calculations will be measured through full-scalemeasurements. The full-scale test campaign will be partially carried out atUffington on the Western Main Line (see Figure 3) and the main series of testson HS1 will take place in the spring of 2015 in conjunction with another EPSRCproject EP/K037676/1 ( However, a preliminary series of tests were carried out in the summer of 2012.At this point there is a double track, 200kph railway which is used by around 6Class 43 trains an hour, in either two power car plus seven coach formation(2+7), or two power car plus eight coach formation (2+8). These tests consistedof slipstream measurements, pressure transient measurements and underbodymeasurements.

map showing Primarily full-scale test location at Uffington

Figure 3 Primarily full-scale test location at Uffington

In addition to the slipstream measurements, two sets of measurements havebeen also carried out on the Network Rail "New Measurement Train" (NMT), aClass 43 train that is used to measure track geometry and condition in the UK(Figure 4). The first measured the pressure around the train nose, which,together with the output from a nose pitot tube, enables the instantaneous yawangle of the flow to be obtained. The second set of measurements was made on aloop around the rear power car of the train (Figure 4).

HST train showing a loop of pressure taping at the rear of theleading car.HST train showing a loop of pressure taping at the rear of theleading car.

Figure 4 HST train showing a loop of pressure taping at the rear of theleading car.

A sample result of these measurements is shown in Figure 5.

graph showing Comparison of pressure transients on train walls through Ampthilltunnel between the NMT and a Class 222

Figure 5 Comparison of pressure transients on train walls through Ampthilltunnel between the NMT and a Class 222.

2- TRAIN Rig measurements

A number of different types of test are being carried out on the TRAIN rig(shown below) using a 1/25th 4 car (2+2 consist) scale model of the Class 43(figure 6). The first set of tests aimed to replicate the full scalemeasurement positions on the Great Western Mainline. The train speeds werearound 40 m/s to meet CEN compliance of a minimum Re number of 250 000.Experiments were also carried out at the same height as in the full scalemeasurements (0.7m above the rail), and ensemble averages of the measuredpressures and velocities were formed from around 20 runs of the rig. The secondset of tests aimed to quantify the effect of ballast shoulder height on theslipstream velocity at (and near to) the TSI position of 2 m from trackcentreline and 0.2 m above TOR. Three ground configurations were selected -flat ground, a typical UK height ballast shoulder (0.3m) and a TSI compliantballast shoulder representative of European conditions (0.7m).

Moving rig modelMoving rig model

Figure 6 Moving rig model.

graph showing physical modelling data (slipstream velocity) from the movingrig

Figure 7 Physical modelling data (slipstream velocity) from the movingrig.

3- CFD calculations

RANS and DES CFD simulations of the flow around the 1/25thscalefour car Class 43 have been carried out using an OpenFOAM generated mesh with atotal of 44.1 million cells. Air speed was set at 40 m/s, which results in aReynolds number of around 300,000 based on the train height. The RANS approachuses the k-? SST turbulence model, The DES approach uses the Spalart-Allmarasturbulence model for the RANS regions. Both RANS and DES calculations givelarge amount of detailed results that have been used to reveal the flow aroundthe model scale high-speed train. Also, results from the DES approach are timeaveraged to enable comparison to the RANS and TRAIN rig results. Figure 8 showsa visualization of the slipstream, using both iso-surface of the secondinvariant of the velocity gradient and a plane coloured by velocity magnitude.A sample of the compareson between the different methodologies is shown inFigure 9.

Visualization of the CFD slipstream flow structureVisualization of the CFD slipstream flow structure

Figure 8 Visualization of the CFD slipstream flow structure.

graph showing Visualization of the CFD slipstream flow structure


Journal publications

  1. Morden, J., Hemida, H., Baker, C. (2014) ‘Comparison of RANS and Detached Eddy Simulation Results to Wind-Tunnel Data for the Surface Pressures Upon a Class 43 High-Speed Train” J. Fluids Eng. 137(4), 041108 Paper No: FE-14-1185; doi: 10.1115/1.4029261

Conference publications

  1. Gallagher, M., Morden, J., Hemida, H., Quinn, A., Baker, C.  Measurement of train aerodynamic phenomena under operational Conditions. International Workshop for Train Aerodynamics, Birmingham, 8-10 April, 2013
  2. C. Baker, A. Quinn, H. Hemida, M. Sterling, M. Gallagher, J. Morden and S. Jordan (2014) The Measurement of Train Aerodynamic Phenomena in Operational Conditions Proceedings of the Second International Conference on Railway Technology: Research, Development and Maintenance, J. Pombo, (Editor), Corsica
  3. A.D. Quinn, C.J. Baker and M. Gallagher (2014) “The measurement of aerodynamic phenomena in operational conditions on a rail vehicle”, First international conference in numerical and experimental aerodynamics of road vehicles and trains (Aerovehicles 1), Bordeaux, France
  4. C. Baker, A. Quinn, H. Hemida, M. Sterling, M Gallagher, J Morden “A comparison of full scale and model scale measurements of train Aerodynamic characteristics”, 21 April 2015 - 23 April 2015 Institution of Mechanical Engineers, London

Contact us

Professor Chris Baker
Phone: 0121 414 5067

Address: Birmingham Centre for Railway Research and Education
Gisbert Kapp Building
University of Birmingham
B15 2TT