Knowledge of the magnitudes of velocities in the slipstream of a train is important for a number of reasons. For example high slipstream velocities can result in dangerous conditions for passengers waiting on platforms and for workers at the trackside, and can cause objects such as push chairs to move. These effects thus need to be taken into account in the development and authorisation of new trains. A consideration of these effects, as well as other aerodynamic issues, has led to the development of a series of standards on train aerodynamics, material from which has been incorporated into the Technical Specifications for Interoperability (TSI), giving limiting values for slipstream velocities. These are being developed to allow trains to run across national boundaries in Europe.  The TSI methodology for the assessment of slipstream velocities is based on a method for assessing the magnitude of the slipstreams of a train and requires that full scale measurements be made at specific points on a platform and at the trackside for 20 train passes within defined vehicle speed ranges, for low wind speed conditions only. The maximum one second moving average velocity for each train pass is then calculated. A value of the mean plus two standard deviations of the ensemble of one second values is then compared with limiting values specified by the TSI. The need for two measurement locations, one at trackside and one on a platform, makes this type of testing somewhat cumbersome, particularly accessing the required platform test site.  A method based on one set of measurements at the trackside that is transferable to any country would be rather more convenient and cost-effective. For this reason, a work package of AeroTRAIN was devoted to investigating the testing procedure for slipstream measurements, with a view to reducing the number of measurement locations.


The objectives of the investigation of WP5 were as follows.

  • To collate existing slipstream data from earlier projects – specifically material from the RAPIDE project and material from UK tests carried out in the 1980s and 1990s.
  • To undertake measurement campaigns on lines in Spain and Germany, to measure the slipstreams for a variety of high speed train and conventional train types at trackside and above platforms.
  • To analyse the experimental data in order to 
    • identify the magnitudes of slipstreams from different vehicles at different heights above the track for both trackside and platform situations;
    • determine a possible revised TSI methodology with a simplified test procedure, ideally at just one location;
    • develop a methodology to assess single vehicles within trains with respect to their relevance / impact on the slipstream effects of a particular train configuration.

Experimental results

The experiments in Spain and Germany were carried out for a range of different train types – high speed single unit trains, high speed double unit trains, conventional passenger units and locomotive / coach combinations. The data that was obtained was supplemented by other data from previous projects – specifically data from the RAPIDE project and earlier measurements made in the UK. Two basic types of calculation were carried out. The first involved a study of the ensemble averages of the slipstream velocities, measured both at trackside and above platforms. The differences between the flows around different train types were elucidated, and the effect of platforms on slipstream behaviour analysed. A brief analysis of the effects of cross winds on slipstream behaviour was also carried out. Through a detailed analysis of slipstream velocity components, the detailed nature of the flow around the nose and in the near wake of the train was investigated, again revealing differences in flow pattern between different trains. Significant similarity in the far wake flows was revealed. The second type of analysis concentrated on the analysis of maximum gusts, in order to make suggestions for modifications to the current TSI methodology. The very large dataset obtained for one particular sort of train (the S-103) enabled the variation of slipstream gusts with vehicle speed and wind speed to be determined. It was also possible to carry out a statistical analysis of the gusts that enabled the standard uncertainty of the TSI gust parameter to be determined. It was shown that for most trains the maximum gusts occurred in the train near wake, but for double unit trains the maximum gusts could occur around the gap between the units and for locomotive / coach combinations the maxima could occur around the nose of the locomotive or at the discontinuity between the train and the locomotive. It was further found that the measurements made at two measuring stations around 30m apart were uncorrelated, which implies that the use of multiple measurement stations could result in a reduction in the number of train passes required. Perhaps the most significant result, which will allow a considerable simplification of the TSI methodology, was that if both trackside and platform measurements for a particular train were plotted against height above the rail, then, with very few exceptions, they fell onto one curve, which implies that a trackside measurement could replace the current required platform measurement.

The proposals

Based on the experimental results a revised TSI testing procedure has been outlined, with the following components.

  • As in the current procedure, the methodology should be based on the determination of the 1second moving average of the velocity for a number of train passes, with the TSI characteristic velocity being the mean plus two standard deviations of the ensemble of runs. 
  • At least 20 valid train passes should be used in the formulation of the ensemble. The uncertainty analysis has shown that this results in a standard uncertainty of about ± 6% (± 12% at 95% confidence) on the characteristic velocity values.
  • For a valid train pass the wind speed for the 15 seconds before the train passes the measurement site should be less than 2m/s, and the train speed should be within 10% of the maximum operating speed. 
  • The measurements of air speed shall be carried out at two heights above at a trackside location – 0.2m and 1.4m above the rail. No platform measurements are required. The acceptable track geometry for such tests shall be as currently specificed. Multiple measurement stations can be used at separation distances greater than 20m, which will result in a reduction of the number of train passes that are required. 
  • Fixed composition trains shall be tested in the configurations that are likely to exist in operational conditions i.e. as single or double units. When testing locomotives, the junction between the locomotive and the carriages is important, and all possible configurations of locos and carriages should be tested. A rake of carriages of 100m in length is required. 
  • In terms of the effect of small design changes, the large confidence limits on the characteristic velocity measurements imply that such will only be able to differentiate between vehicles for relatively large design changes. A fuller definition of what is meant by small and large design changes is still required. 
  • The limit values for characteristic velocity shall be the same as specified at present, with the current “platform” limits being applied to the measurements made at 1.4m above the top of the rail.
  • Perhaps the most important point to emerge from these investigations is a fuller understanding of the stochastic nature of the flow within train slipstreams, that results in large scatter within the ensemble of gust values, and the need to have a full appreciation of the uncertainty and confidence limits of any TSI characteristic velocities that are derived.