During post-season testing in Abu Dhabi, we saw many strange appendages on the cars that we don’t see at race weekends. What were they?
On a race weekend, we can run extra instruments and sensors on the Friday, but they must keep within the regulated bounds of bodywork height, width and overhang. These restrictions don’t apply in testing and so we can use transducers that would otherwise be illegal.
The most visible bits of testing kit you see are the aerodynamic rakes, which are used to understand the airow as it sheds off the bodywork and forms the complex wake structures that surround the car. These devices consist of an array of sensors called Kiel tubes, which are attached in a matrix fashion to a large assembly mounted on the car. These measure the total pressure of the air that they encounter. Measurement of the total pressure is a means of deriving the airspeed, and by understanding the speed and direction of the air around the car, we can understand the contorted vortices that are fundamental to performance.
How precise are these measurements?
The very act of measuring anything actually changes the state of whatever you are trying to measure, something known as ‘The Observer Effect’. These sensors are no exception. Their presence affects the airow because the air has to deect around them. Hence we need to balance the desire for accuracy with the need to have low interference, which is achieved by having a minimum number of sensors.
Surely this is what the windtunnel is for?
Yes it is, and a large part of what we do is to correlate the ow structures between the windtunnel and computational uid dynamics (CFD) to measurements made at the track. In the windtunnel we will mount scale versions of these sensor arrays and see if we get similar results to those measured on the car.
What other unusual sensors are used?
At the end of 2014, we tested the 2015 tyres and, as well as assessing the drivers' subjective view of the effect of the tyres on handling, we need to understand the true operating shape of the tyres. We do this by running a laser scanning device mounted to the upright, which shines a laser line across the tyre sidewall. A camera mounted alongside it lms the laser line over a lap so we can measure the deected shape of the tyre sidewall under true loaded conditions.
Why is the tyre shape so important?
If we were able to seal the gap between the tyre and the diffuser we could nd enormous performance. This is what the pre-2014 exhaust-blown diffusers were about. We were trying to use the high-energy exhaust ow to produce an air curtain that separated the dirty air emanating from the side of the tyre from the clean diffuser ow. They say you can never uninvent something, and we still spend a lot of time trying to understand the ow elds in this region. Fundamental to this understanding is the ever-changing shape of the tyre sidewall. Laser scanning shows us how this shape changes and lets us choose various shapes to simulate in windtunnel and CFD testing.
Does the single ECU compromise your tests?
It can do, in that it has a nite amount of sensor inputs and memory, but during testing we often run with an additional data logging device that lets us record additional inputs and sample them at much higher frequency.
Does running all this extra instrumentation present any danger?
No. While we may not be subjected to FIA scrutineering during testing, we are all responsible engineers and will use our best endeavours to ensure that anything we put on the car, no matter how temporary, is safe.
So is all this instrumentation about correlation? Correlation of on-track results with experimental techniques is vital. Testing gives us a rare chance to compare our results with real life. Do these devices replace the flow-viz fluid and even the wool tufts of days gone by? Largely, but there is a difference in that rakes measure and visualise the ow once it has separated from the surface, whereas ow-viz paint lets us picture the ow while it is attached. Both are necessary to understand the full picture.
Is there anything you can’t measure? The hardest things to measure are downforce and drag. We have no direct means of measuring drag at all, so derive it from measuring engine torque at constant speed while trying to account for the non-aerodynamic resistances, or by letting the vehicle coast from a given speed and then trying to work out the contribution of aerodynamic drag to the resultant deceleration.
It is equally hard to measure downforce. We have load cells in the push and pullrods, but these don’t just measure the downforce, which constantly increases with speed; they also measure the weight transfer associated with cornering, braking and acceleration. We try to separate these but it is not easy. The other problem of measuring downforce via chassismounted sensors is that a large amount of our total load is generated by the multiple winglets on the brake ducts. These produce load directly on the upright and tyre and hence are not seen by a load cell mounted on the pushrod.