Repeatability Blog

Repeatability is key to understanding and validating wind tunnel test data. Unless variables are meticulously monitored and controlled, the variability of results could be greater than the marginal differences attempting to be measured. How a wind tunnel session is managed can often influence the quality of results obtained.

The wind tunnel hall is temperature controlled and the wind tunnel control software compensates for changes in ambient conditions to prevent drift in results through maintaining either a constant Dynamic Pressure or constant Reynolds Number, normalising drag for STP (Standard Temperature and Pressure) conditions. This also helps to achieve comparable results if comparing data to that recorded on another day, where the ambient conditions may differ.

What the wind tunnel cannot account for is how a human body reacts to ambient conditions and physical exertion. A change in body temperature or perspiration could well affect air flow characteristics and therefore drag.

The test variable is the thing that is to be changed in order to measure aerodynamic differences (whether a different component, equipment or change in rider position). To accurately measure and quantify the differences that a change has produced, all other variables must remain constant. Examples of variables that could influence results and therefore should be considered are:

  • •Bike installation, vertical alignment.
  • •Bike accessories: computer head unit, drinks bottle, bottle cages etc.
  • •Rider or mannequin position: hands, wrists, arms, head, shoulders, back, position on saddle.
  • •Helmet fitment and orientation.
  • •Clothing fitment: zip position, pocket positions, sleeve lengths, short lengths, sock heights, position and orientation of seams and fabric detail.
  • •Rider effort, power, cadence.
  • •Gear selection, wheel speed.
  • •Ability of rider to maintain a position.
  • •Rider fatigue, temperature, perspiration.
  • •Crank position (if bike only testing).
  • •Wheel orientation and tyre valve position (if wheels are static).
  • •Wheel fixture position (if testing a wheel in isolation).
  • •Tyre pressure.
  • •Yaw sweep direction.
  • •Possible aerodynamic instabilities of static objects (such as a mannequin).- •Movement in cables
    •Loose stickers or tape on the frame and wheels.
    •Handlebar alignment.

A rider simply getting off the bike and straight back on again can induce positional, helmet and clothing differences as well as a change in weight distribution on the force balance, all of which could have an effect on the drag measurements.

Utilising software tools such as the photos, live athlete edges and taking a tare before a run will help to minimise unwanted variables. It is therefore important to conduct baseline repeats to understand the variability of a test specimen to validate changes and results.

Testing protocol

To demonstrate the repeatability of a pedalling rider, we recently went into the Sports Performance Wind Tunnel to carry out the following test procedure:

A Giant Trinity TT bike was used without change to the set up between runs.

The speeds selected for the test were 45 kph and 50kph, measured at 0,3,8, and 12 degrees of yaw, providing a representative range of conditions faced in real-world riding/ racing.

The riders’ edges were captured during the baseline test, which were then projected in front of them for each subsequent run as a reference point to maintain the same position on the bike and ensure repeatability. A 10 second pedalling tare was also completed prior to each run to account for any changes in weight distribution.

The only changes made between runs was the helmet, changing from a Kask Mistral to a MET Drone, before then repeating for a second time.


Comparing the results between the two runs, the average difference in CdA between the Kask Mistral and repeat was 0.5% across the two speeds and yaw angles.

The averaged difference in CdA between MET Drone and repeat was 0.1%.

At SSEH, we would typically advise customers’ to aim for repeatability to be within the range of 0.5 – 1%, this test was therefore within an acceptable range to have confidence in the deltas between runs.


Based on the results shown, the importance of a rider maintaining position is critical to accurate data recording and confidence in tunnel data output.

It can often be tempting to make equipment changes before establishing an accurate rider baseline, however, without knowing the variability of the rider, it is difficult to determine if the deltas between runs are due to a positional change by the rider or the piece of equipment being tested.

By ensuring a base line repeat is included in your test programme (as a minimum), you can then start to determine measurable differences with greater confidence in the data provided.

If you would be interested in arranging a session in the Sports Performance Wind Tunnel to optimise your own equipment choices, please refer to the Performance Consultants section of our website for further information –