Usage Of Numerical Modelling In Studies Of Trains

Using full-scale experiments to investigate the velocity and pressure generated by high speed trains in confined spaces such as tunnels is very difficult and expensive. It is more practical and economical for researchers to carry out model-scale experiments due to the more controlled environment and the possibility of investigating different parameters in a short time.

The physical modeling, in forms of moving and rotating models, has been used to study a range of aerodynamic investigations in open space problems (Baker 2001, Baker 2008, Baker 2012 , Baker 2013 Gil 2008, Jonson 2009, Sterling 2008). For instance, Baker (2012a) investigated transient aerodynamic loads for a range of different structures in the vicinity of railway tracks using reduced-scale models. The ultimate goal of the study was to use the data to develop relevant European codes and standards, which are not specifically for UK specific loading gauges. Hein and Ehrenfried, (2013), investigated the phenomenon of pressure waves at the exit of a tunnel. The moving-model experiment facility based in Gottingen, Germany was used to investigate the passing of an ICE3 through a 2 m tunnel at 1/25th scale. The operating speed of the train was between 38 m/s and 44 m/s. Hwang (2000) studied the effects of a number of parameters including train speed, blockage ratio, nose shape and air shafts on the compression wave generated by a high speed train in a tunnel using model-scale experiments. Gilbert (2013) studied the pressure and velocity variations in the tunnels of different lengths.

Numerical modelling

Numerical simulations have been used extensively to study the flow around trains subjected to crosswinds (Baker 2011, Cheli 2010, Hemida and Baker 2010, Hemida and Krajnovic 2009) slipstream investigations (Hemida 2010, Hemida and Baker 2012, Huang 2014, Flynn 2014) and drag reduction tests (Vladimir 2012 and Hong-qi 2009). However, due to the complexities associated with the moving train simulations in CFD, few researches have been carried out to study the flow around trains in tunnels. Baron (2007) studied the nature of pressure waves induced by a high-speed train travelling through a tunnel. The study used two different methods: model–scale physical experiments and numerical simulations. The data obtained were compared in order to validate the results from both methods. Khayrullina (2014) studied the flow around Dutch passenger and freight trains running through underground tunnels using transient CFD simulations. The research considered factors influencing the strength of pressure waves inside a tunnel such as blockage ratio, train speed, portal shape, tunnel length and roughness. The study found that the passenger train caused higher velocities than the freight train because of a higher operating speed and lower blockage ratio. Hwang (2000) attempted to visualize the propagation of compression waves through a tunnel using CFD.

Baker (2006) and Baker (2010) discussed a number of experimental and numerical studies on the assessment of the slipstream gusts caused by passing trains in open field, with and without cross winds. They also described the potential effects of wind gusts on exposed people. They mentioned that there is a large variability in experimental data due to different boundary conditions, train types and complex flow structures induced by a moving train. Three regions around a train moving in open field were distinguished: the nose region, the boundary-layer region, and the wake region. Also, these authors highlighted the development of turbulent gust flows in the near wake for high-speed trains and in the growing boundary layer of freight trains. These unsteady flows can cause discomfort or even destabilize people standing alongside the moving train, by gusts with speed above 15-20 m/s (Baker 2006). Sterling (2008) analyzed experimental data for high-speed passenger trains and freight trains in open field. They examined the different flow regimes within the three regions around a train and, in line with the previously discussed studies, highlighted the intermittent behavior of the near wake flows. The velocities were found to be higher in the near wake and the boundary layer regions than in the nose region of the train. They also mentioned that the boundary layer development was slightly different between full-scale and reduced-scale measurements and that this could influence the near wake flow.

Baker (2006) and Baker (2010) discussed a number of experimental and numerical studies on the assessment of the slipstream gusts caused by passing trains in open field, with and without cross winds. They also described the potential effects of wind gusts on exposed people. They mentioned that there is a large variability in experimental data due to different boundary conditions, train types and complex flow structures induced by a moving train. Three regions around a train moving in open field were distinguished: the nose region, the boundary-layer region, and the wake region. Also, these authors highlighted the development of turbulent gust flows in the near wake for high-speed trains and in the growing boundary layer of freight trains. These unsteady flows can cause discomfort or even destabilize people standing alongside the moving train, by gusts with speed above 15-20 m/s (Baker 2006). Sterling (2008) analyzed experimental data for high-speed passenger trains and freight trains in open field. They examined the different flow regimes within the three regions around a train and, in line with the previously discussed studies, highlighted the intermittent behavior of the near wake flows. The velocities were found to be higher in the near wake and the boundary layer regions than in the nose region of the train. They also mentioned that the boundary layer development was slightly different between full-scale and reduced-scale measurements and that this could influence the near wake flow.

Gil (2008) mentioned considerable run-to-run variability in the measured data for a 1/25th scale train with 3 carriages moving on a circular track with speeds of about 5 m/s to 15 m/s. They experimentally showed that higher train speeds cause higher ratios of slipstream velocity to train speed. However, Hemida (2010) studied a 1/25 scale model of an ICE train running on a circular track in an open space using validated LES simulations and showed that the Reynolds number effect on normalized slipstream velocities is negligible for trains moving with speeds varying within 20%. Finally, Hemida (2014) in their LES study investigated the effect of the platform height on the slipstream velocity. The slipstream velocities that occurred with a higher platform were increased due to the blocking of the developing slipstream flow. They also monitored the instantaneous flow in the wake of the train and confirmed the presence of highly turbulent vorticity. The maximum velocities and the largest turbulence intensities were observed in the near wake of the passenger train.