Aerodynamic drag is the main resistance we have to contend with as road cyclists. The fact that we encounter an almost impenetrable "wall" at 40-50 km/h is due to the effect of air resistance, which increases quadratically with speed. The power required to overcome the resistance even increases to the third power with speed. But even at much lower speeds, air resistance is the measure of all things. Depending on your posture on the road bike, air resistance is the dominant driving resistance from 18 km/h upwards. Cyclists therefore benefit from improved aerodynamics at all common road bike speeds, regardless of whether this is due to a better posture or better material. Whether we can ride a 26 or a 28 km/h average - with the same pedalling performance - is primarily due to aerodynamic reasons.
On steep terrain and with frequent acceleration, weight becomes much more important. Beyond a gradient of 5%, weight is the decisive factor for riding resistance for amateur riders. For professional riders, however, who climb hills much faster, aerodynamics are a factor that should not be neglected up to a gradient of around 7%.
Moving legs for real test data
TOUR has already used a large number of wind tunnels. We have been testing in the GST tunnel on the Airbus site in Immenstaad since 2012. The open channel is a former Dornier aeroplane wind tunnel and offers the best conditions for very high-quality measurements with high resolution and good reproducibility, which are not possible on the road or on the racetrack. Together with the operator, we have developed a measurement protocol that sets the standard for aero tests of bicycles, both in terms of resolution, speed and reproducibility: In order to flow around the racing bike and wheels as realistically as possible, we place a dummy on the bike that pedals along. Compared to a human rider, the dummy has the advantage of infinite endurance and once it is seated, it no longer changes its position, which benefits the measurement quality.
After initial trials with a rigid full-body dummy, the moving rider without upper body is the second evolutionary stage and a TOUR in-house development. We left out the upper body because it has little aerodynamic interaction with the frame, but creates a lot of resistance and therefore increases the measurement error. We always measure the bike and dummy together with the aim of obtaining the most relevant and practical measurement values possible.
We always set the height of the handlebars to the same position and also fit a bottle cage with a 0.75 litre bottle. In order to record wheel effects, we measure the bike with the original wheels (basis for evaluation), but also with a fast reference wheelset (Zipp 404 or Swissside 625). The flow velocity is 45 km/h and we swivel the wheel over an angle range of -20 to +20° relative to the longitudinal axis of the channel. The diagonal flow represents situations where crosswind and airstream add up to a resulting wind that no longer hits the rider head-on but at an angle. We choose a high speed in order to be able to measure well. The results can be transferred to other speeds. It would be wrong to assume that aerodynamics are only relevant at such high speeds.
During the tests, the rear wheel is driven by a roller in the wind tunnel floor. We use a rigid drive with a locked freewheel to drive the cranks, which in turn move the dummy's legs. The front wheel is driven in the same way. The entire structure rests on a six-component balance in the wind tunnel floor, which records the resistance in the direction of travel, lateral forces and torques. Using a separate device, we can also measure the forces that the driver feels in the steering when the wind is in contact with flat rims. The measurement over the entire angle range is carried out continuously in one go, which above all saves measurement time. In less than one and a half minutes, we record the force exerted by the wind on the wheel for 41 measuring points. We select the integration times so that whole leg rotations per measuring point are included in the measurement.
The wind doesn't always come from the front
We use the measurement to determine the cwA values, which have the unit m^2 - the product of the cw value and the flow area. These values can easily be converted to other conditions such as different speeds or air pressures. In order to arrive at a conclusive result, we evaluate the 41 measured values and summarise them into a drag coefficient, expressed in watts, for a typical driving situation. As the upper body is not included in the measurement, the resistance of a real rider is around 50 per cent higher (which we take into account in the simulation calculations).
The conversion of the resistance curve across all angles into a watt value requires some assumptions. This is because the probable angle of attack depends on the typical wind speed and its distribution. We assume a mean wind speed of 10 km/h and a Weibull distribution of wind speeds, as used in wind power analyses. Based on a random wind direction, the probability of the angle of attack can be calculated.
Accordingly, the individual measurement points are weighted differently: we obtain a bell-shaped distribution around zero degrees - this means that small angles are much more likely and therefore have a higher weighting for the formation of the resistance than large ones. If you are travelling in areas with very little wind, the distribution becomes more acute. Small angles are then given an even greater weight. For individual assessment, we therefore also map the drag curves so that you can get an idea of how the drag develops.
The riding speed also plays a role in the angles. Slow riders experience more diagonal flow than fast riders in the same wind. This leads to the paradoxical situation that slow time trialists can gain the most from aerodynamic material, as the advantage of aerodynamic material increases significantly in a diagonal flow.
Measurement accuracy and relevance
The measurement accuracy within a measurement campaign - this is usually the data that is published within a story - is +/- 0.25 W. The repeatability between different campaigns is +/- 1 W.
Many readers ask: Is the data relevant, doesn't the rider make the main drag? It is true that the rider dominates the action. An aerodynamic riding position therefore brings the most speed. But the material advantage is always present and should not be underestimated. Between a very good and a very bad aerodynamic bike, there are around 35 watts at a speed of 45 km/h. Transferred to lower speeds, which are typical for leisure cyclists, the road bike alone can cause an increase of one kilometre per hour with the same riding position.
You can become really fast if you take care of all the components of resistance and optimise your posture, clothing and material. Ideally in this order. To demonstrate the time advantages of aerodynamic optimisation, we use simulation calculations that incorporate the wind tunnel data as well as the weight of the bike. We then simulate different mountainous routes - from 500 metres in altitude per 100 kilometres (flat), over 1000 metres in altitude per 100 kilometres (undulating) to 2000 metres in altitude per 100 kilometres (mountainous). Based on a constant pedalling power - we usually choose a moderate 200 watts here - we can see what the aerodynamics of the material means for which distance in minutes and seconds. To do this, we add the missing upper body to the dummy data, set the weight of our sample rider to 75 kilograms and let the rider race virtually over the various routes.
If the speed achieved is irrelevant, aerodynamics are obsolete, which is why we only assess aerodynamics for competition road bikes. Strictly speaking, however, the weight doesn't matter. Because with the right gear, two or three kilograms don't matter on the mountain.
