We have analyzed in our previous entries some of the most important aerodynamic parameters of a racecar, namely downforce, drag, balance, efficiency and the aerodynamic coefficients. Today we will deal with how these data are measured, collected and organized.
It is important to keep in mind that downforce, drag and aerodynamic balance (or centre of pressure position), as well as the aerodynamic efficiency, change very dynamically as the car travels around the track. We have seen already how drag and downforce depends on the square of the speed, which means that anytime that speed changes, the car will experience even a bigger change for downforce and drag (and this is one of the reasons why of downforce and drag coefficients).
This is only a part of the whole picture. Even when considering the aerodynamic coefficients multiplied by the frontal area (CzA and CxA), thus not considering the direct influence of speed on downforce and drag, a racecar’s aerodynamic properties are extremely sensitive to ride heights and rake.
This means that even considering that a car travels at constant speed, it will experience a different downforce, drag, balance and efficiency depending on car’s front and rear ride heights. How drag and downforce depend on ride heights and rake is linked very much on car’s design (and, hence, on the technical rules) and on which portion of the aerodynamic forces are generated by components seating close to the ground, as we will see. Normally, downforce is more sensible than drag to ride heights variation.
In general, high downforce vehicles, like LMP1 and LMP2 cars, are aerodynamically very ride height sensitive and this is why race engineers play so much attention in controlling what is often called car’s “platform”, referring to car’s ride heights and rake.
Ride height and rake sensitivity are the reasons why Formula 1 teams in the past invested so much energy and resources in developing active suspensions (until they were banned, at the beginning of the nineties). Their goal was to have always the “ideal” ride heights in each driving condition the car could encounter on the track.
To “monitor” how drag, downforce and balance change with respect to car’s attitude, manufacturers normally compiles tables filled with values of CzA, CxA and Balance corresponding to many combinations of front and rear ride heights. These tables are called aeromaps. An example of an aero map (actually a portion of it) is shown in the following picture.
It contains data taken from Perrin “MyP1” project, some years ago. Nicolas Perrin designed an LMP1 vehicle with the intention to bring it to Le Mans, through a very innovative and open project, involving also the fans. Normally, a proper aeromap has much more values in, but since manufacturers and teams data are kept very confidentially, they cannot be shared. On the other hand, Perrin data were made available to the public as the project was based on an open source platform and we can use them to show a couple of important points.
The map above refers to a low downforce configuration, meant for Le Mans. Although the downforce is pretty low, we can still see how much of an influence ride heights have on downforce and balance.
If we look at the line corresponding to a front ride height of 15 mm, we have a CzA of 3.36 for a rear ride height of 25 mm and a CzA of 3.43 for a rear ride height of 31 mm. This means, a rear ride height change of 6 mm produces a CzA relative change of about 2.1%, which can seem small but is actually a substantial difference.
It is also interesting to see the difference in balance between the two above configuration: we have 43.4% of the downforce acting on the front axle for the first one, 44.9% for the second one. The difference between the two is a 1.5%. Professional drivers are normally able to perceive clearly setup changes leading to balance differences below 1%.
As we anticipated, the difference in drag for the same ride heights values is lower than the downforce one, with CxA of 0.761 for a rear ride height of 25 mm and of 0.771 for a ride height of 31 mm. This means a relative change of about 1.3%.
We will not deal with the effects of yaw and roll angles on the aerodynamic forces, but the reader should keep in mind that these effects exist and can be very important. Yaw angle effects on downforce, drag and balance can be relatively big and cars are often optimized also setting targets relative to the yaw angles expected on track.
Aerodynamic engineers always design cars that create a stabilizing aerodynamic Mz (yaw) moment in presence of a yaw angle (hence, in corner), whose magnitude grows with speed as any other aerodynamic action. This helps to increase vehicle’s controllability, but these topics are out of the scope of these articles.
The data contained in an aeromap can be collected in different ways. We will briefly introduce three of them: CFD, wind tunnel testing and track testing.
During the design phase, when the car only exist in a virtual environment and the engineers are still evaluating high-level aspects, most of the manufacturers normally employs CFD (computational fluid dynamics) simulations to understand, at least qualitatively, which could be the best route to follow. Sometimes, CFD is also used in other phases of a project or by teams running a car they didn’t design themselves, as an alternative to wind tunnel testing, that would be very expensive if done with a full-scale car or with a purpose-built scale model.
CFD simulations are based on a numeric approach to solving fundamental fluid dynamic equations. The “control volume” (term identifying the volume that will be simulated, which is normally much bigger than the car itself) is divided in finite elements (smaller volumes of defined shape and dimensions) and fluid dynamic equations are solved for each volume and each time step.
An iterative process is used and the calculation runs until “convergence” is reached, which means until the error between two following calculations is small enough to be ignored.
The car itself is represented as a surface that the fluid cannot penetrate, depending on the assumptions (elements like radiators and engine intakes could be made permeable).
Specific mathematic models are used to simulate what happens when the fluid gets closer to car’s surface, because of fluid viscosity and local turbulence.
Important elements for racecars aerodynamics, like rotating wheels and moving floors can also be reproduced in CFD environment and this surely helps to improve results accuracy. In the real world, the car moves with respect to the ground, but aerodynamic tests are normally performed keeping the car in a certain position and letting the air and the floor move at the same speed with respect to the car.
From a pure computational perspective, using CFD efficiently requires extremely big and powerful computers and these simulations are normally run using clusters, which are none but many computers working in parallel at the same time. This is one of the reasons why, if the car under analysis is symmetric with respect to its longitudinal middle plane, CFD engineers normally only study half of it, as shown in the previous picture.
The most interesting aspect about CFD, beside numeric results defining cars performance, is that everything can be analyzed in the very details, in a way that would be impossible with physical or lab testing (see wind tunnel). The engineers can potentially look at what happens in every point of the control volume, no matter how small or how difficult to reach in the real world.
Beside this, of course designers can extract the most important aerodynamic metrics out of a CFD run, like pressure distribution, air velocity in the whole control volume, streamlines, downforce, drag and balance.
The biggest question mark with CFD remains its accuracy. We have to keep in mind that every test, no matter how much effort the engineers profuse to control testing conditions and make them repeatable, will never be able to reproduce exactly what the car experience on track. This is actually not really the goal of testing, because a race track is, most of the times, the best example of uncontrollable boundary conditions (teams have no power in deciding weather conditions, temperature, track conditions, etc.) and uncontrollable boundary conditions is exactly the opposite of what you desire to do proper testing.
Anyway, some aero engineers trust other methods more, in order to get reliable quantitative results, although successful car manufacturers, like Oreca, make extensive use of CFD to develop their cars.