Wind Tunnel Test

Klik hier om het artikel in het Nederlands te lezen!

Our aerodynamic engineers, Bart and Charlotte, have made the design of the VeloX 9 as aerodynamic as possible and produced it over the last months (read more about the production of the shells here). The next step is to measure the aerodynamics of the shells, among other things to validate the used models but also to investigate the influence of crosswind for example. For this purpose we moved to the Open Jet Facility (OJF), one of the wind tunnels on the campus of the Delft University of Technology.

Aerodynamic Engineers Bart and Charlotte monitor the data during testing. Picture: Bas de Meijer

The OJF is a low speed wind tunnel with a test diameter of 285 x 285 cm and a maximum wind speed of 35 m/s. Thanks to the open jet and the large outlet opening, large test objects can be measured in the OJF, including the VeloX 9. Another big advantage is the OJF’s range: the maximum wind speed is 35 m/s, equivalent to 122.5 km/h, making the range almost equal to the speeds to be achieved during the World Human Powered Speed Challenge in Battle Mountain.

A schematic drawing of the working of the OJF.
Source: TU Delft

Various test setups were used in which different values could be measured. The three main methods of testing were: measuring the air resistance by means of force measurement, listening with a microphone to the streamlines, and following the streamlines with the help of UV-sensitive paint.

The most important measurement is to determine the CdA value of the shells. The VeloX 9 is fixed to the stand and the wind turbine is turned on. The stand is attached to an accurate measuring instrument that measures the forces in the X-Y-Z direction and also registers the force moments. The measured forces are converted into a concrete CdA value for the various wind speeds used for testing using a Matlab script. This makes it possible to determine which race shells are the most aerodynamic, and also to validate the model used.

By measuring the vibrations we can separate laminar and turbulent flow areas. Picture: Bas de Meijer

A similar measurement was carried out with last year’s VeloX 8. By measuring two of last year’s race shells (one that has been repaired and one that hasn’t been damaged) and comparing these results with last year’s measurements, it is possible to compare the difference in aerodynamic resistance between a repaired shell and an undamaged shell. For example, is it worth repairing a shell if, after repair, it experiences significantly more friction than the other race shells?

It is important to identify where around the VeloX laminar and turbulent flow occurs. Turbulent flow increases resistance and should therefore be minimized. There are various instruments for this, including measuring with noise. With laminar flow, the microphone does not absorb vibrations in the air (i.e. sound), whereas with turbulent flow it does.

Furthermore, wind plays a major role during the Battle Mountain race. For example, next year we want to map the wind profile over the entire 10 kilometre long race during the various runs. Also during the wind tunnel tests the influence of the wind is examined by angling the set-up with the VeloX 9 and mapping the influence on the change in CdA value.


One way to map the flowlines on the surface is by using UV-sensitive light. First, an even layer of paint is applied to the VeloX 9. After this, the normal lights are switched off and the UV lamp is switched on. When the turbines are switched on, the paint ‘runs’ over the surface and the streamlines become visible. It is also possible to map the extent to which separation occurs, a phenomenon in which the boundary layer of air that runs over the surface is released, resulting in a high amount of resistance. The less separation, the better!

These measurements are especially useful for the current team to be able to make tactical choices during the race. What are the weaknesses of the different shells, and which ones are the best to choose under favourable conditions? In addition, the results are processed in new team’s hand-over, for example to validate the models and possibly adjust the parameters.


Share on facebook
Share on twitter
Share on linkedin