By Jason Ansell

This article is the result of three years of research into aerodynamics of Superstox. It would seem that that win designs being used at present are not all they could be.
When aerofoils became the trend on top of Superstox, Cyril Ansell of A B C Racing, said "We'd better study up on these aerofoils son, to see if they really do any good." A friend at university put us in touch with Mike Gastor, a professor in aerodynamic engineering and the project was born. Professor Gastor has worked with aerodynamics for 30 years and did work with the McLaren F1 team in the 80's, so definitely knows his stuff when it comes to airflow.
The very first time we took the Superstox to the university, he had a good look and said that the aerofoil looked ok, but he was a bit worried about the roof plate directly underneath. The aerofoil at that time was based on the NACA 63-418 profile with one trailing edge flap. His initial opinion was that although the aerofoil was a good profile, the roof plate underneath would destroy any flow needed to generate good downforce. He speculated that is would probably produce about 70lbs of downforce, but a venetian blind type system would work much more efficiently with this disturbance, probably producing in the region of 200lbs of downforce. He agreed to do a project on the car with two of his students, James and Richard, for their degree, using a 1/5 scale model of the car.

Aerodynamic theories.
Ask anyone how an aerofoil works and nine out of ten people will say the air hitting the top of the aerofoil pushes the car down - wrong. The fact is only 15-20% of an aerofoil's effectiveness is generated this way. The other 80% is produced by low pressure underneath sucking the aerofoil down. Low pressure underneath is created by the increased speed of airflow underneath and decreased speed above the wing. This is why a smooth flow of air is so extremely important. If the flow is interrupted or is not clean, it separates from the surface of the wing and most of the downforce is lost. The majority of aerofoils, certainly the ones on top of the race cars at the moment, will not work at any angle of greater than 17 degrees. This is known as the angle of attack. If this angle is too great the wing 'stalls' and again the air separates from the surface and downforce is lost.
The question is bound to be asked why do the aerofoils work so well on American sprint cars? There are several reasons for this. The biggest factor is they have no roof plate, just the roll cage. It really does make that much difference also, they are travelling at much faster speeds, up to 130mph. Downforce increases by the square of the speed, e.g. twice the speed = 4 times the downforce. 3 times the speed = 9 times the downforce. 4 times the speed = 16 times the downforce. For instance, if a given wing produces 50lbs of downforce at 40mph, at 80mph it will be producing 200lbs of downforce and at 120mph it will produce 450lbs of downforce. So the American Sprint Cars travelling at over 100mph without the roof plate have a very definite advantage.
The standard single wing section of type aerofoils, with end fences, generally available at the moment do work, partially, but as explained before, only the top surface of the aerofoil is working. The roof plate did disrupt the flow as the professor feared and several options were tried with the same result, turbulent airflow beneath the wing, the effect of this is shown in the results.
Initial tests in the wind tunnels showed very poor results. Firstly, the basic car without aerodynamics shows an uplift of about 20lbs. A basic aerofoil showed a downforce of about 40lbs, but with tremendous amounts of drag caused by the turbulent airflow. The standard aerofoil with a trailing edge flap set at different angles showed a slight improvement, up to 70lbs of downforce. Remember what the professor said when he first saw the car with the wing. Not a bad guess professor. This set up was still producing far too much drag however.

Pitching Moment
Unless the wing is moved up away from the roof about 2ft it isn't going to work as an aerofoil. These type of wings on cars that are racing at present can only be described as air deflector plates. Don't be fooled, they are not working as aerofoils. A few other things should be considered here. A force known as 'pitching moment' comes into effect because your wing is above your wheels, (hopefully), and is inclined into the airflow it has to produce drag. This drag creates a rotation lifting the front wheels and pushing down the back, (grip!) and the more drag you get the more pressure this force has. Unfortunately, you can't forget about the drag which is slowing you down. You have to find a compromise between this drag and the grip you are getting. This will show on the stop watch when you alter the angle of your wing, but wouldn't it be nice if you could get more downforce and less drag - you can. If you can get the air flowing properly.
In the wind tunnel we started to move away from aerofoil sections on to simple curved plates. Surprisingly a simple curved plate produced better results than an aerofoil. Probably because it catches more air, but it still has a poor airflow. Most of the work on any aerofoil is done about one quarter of the length of the wing back from the nose underneath and its the tiny bit of air about one quarter of an inch from the surface all over the wing which is most important, known as laminar flow. The theory is that a wider but narrower wing is more effective than a long thin one, (known as the aspect ratio), so if you have three or four wings 12" deep instead of one big wing 4" long it will be more efficient, rather like a venetian blind.
Interestingly whilst watching the Grand Prix recently, I noticed that a definite change has occurred from single and double wings to now up to four or five fins and the last flap is inclined at something like 45 percent. These F1 cars are literally like an engine powered wing with a driver. I watched the McLaren series on the television when the cameras went behind the scenes with great interest. Apparently, these wings have so much downforce you could drive one on the ceiling at 70mph and it wouldn't fall off! Ten years ago the F1 cars were using single wing sections.
Although it is probably far more technical and much more thoroughly tested, even the world's most expensive racing cars are using venetian blind type wings. Have a close look in the next Grand Prix. The basic theory behind this is the clean air hits the first wing and flows around it and is slightly turned upwards because of the angle of attack. The second wing then can be inclined at a greater angle
because the air is already flowing upwards. So for instance, if your first wing is inclined at 8 degrees then when the air exits it will be flowing at 8 degrees or possibly slightly more. The second wing can then be set at 20 degrees to horizontal, but the effective angle of attack to the airflow is 12 degrees, (20 degrees ~ 8 degrees), so it doesn't stall and so on. Your last flap can therefore end up at 45, 50 or even 60 degrees and the wing doesn't stall. The more flaps you have the better results. You are limited though from a strength point of view, a 4" deep wing 48" wide is not very strong and there's a lot of pressure bearing down when its working correctly. So why doesn't the roof plate destroy the airflow on this wing? Now this is the really clever bit. The gaps between the wings allow the fast moving high pressure air on top to flow through to the low speed low pressure air beneath, significantly after the work has been done, so keeping the air moving. The interaction between the plates is very important, every time you reach a gap you are almost starting again with negative pressure fast flowing clean air that has already been turned upwards by the preceding wing. Just the job. It's not clever really, its an age old method called boundary layer control used on aircraft to keep air moving and it works perfectly where any obstructions in the air stream might interfere, such as a roof plate for instance.
The size of the gaps is quite critical to achieve the best airflow. This can only really be optimised in a wind tunnel. Interestingly the gaps are not the same along the wing although the cross section of each wing is the same, only the angle changes.

Wind Tunnel Test
As we progressed with the wind tunnel testing things became more and more exciting, we began to wonder just how much downforce we would end up with. It seemed every time the students altered something, the dial went up a little further. Unfortunately, the whole thing was Mongolian to me as it was measured in kilonewtons and the speed in metres per second, so we had to wait for the translation before we really knew what was happening.
In the meantime, every different wing was tested in the smoke tunnel. This is another wind tunnel in a dark room with a spotlight on coloured smoke. The smoke flows over the car and you can see the flows and how it changes and reacts to the wing. The final wing we ended up with actually turns the smoke from horizontal to almost vertical flow on exit and the flow is perfectly clean. This was all of course captured on video. I have made out a table of results on some of the different wings tested.
Air speed is a constant 56mph, wings are all 4" x 48", and the actual downforce figure is how much more the car weighs at this speed in total. Remembering the car with no wing had an uplift of 20lbs, the downforce from the wing itself is 20llbs more than shown, 20lbs of course countering this uplift. A slight re-design of car shape would eliminate this uplift - probably a front spoiler of some description. This downforce figure takes into account the force from pitching moment also.
Don't ignore the drag figures, these are very important. This force is trying to push you exactly the opposite to the way you want to go. Bolting one of these wings on your car isn't going to suddenly transform it into a race winner. If the car is set up correctly and you achieve more grip then you will go faster, although the difference in grip may mean you have to alter your set up to compensate. All the other variables have to be right, i.e. torque, loading, weight distribution, spring rates, roll bar rates, tyre pressures etc.