Aerodynamics (and CFD) of Cornering Vehicles

[Greg Locock]

You will struggle to get representative CFD FSAE results in a straight line let alone complicating it with a yaw condition.

True, but you may see the trend in the differences. Of course if you start with one of the defined bluff bodies you can calibrate your results.

Incidentally we do plot CD and Cl and the other 4 vs yaw angle, so that we can simulate crosswind performance for passenger cars. I've held off describing the curves because they are on my other computer. In general i remember that Cd drops at large yaw angles, but I'm guessing that for cornering we are only looking at 3 degrees or so.

[bhall II]

My question is, given a constant speed, do all F1 cars lose some down force while in yaw/side-slip when compared to the same speed in a straight line?

Yes, but the better cars will tend to be those that lose the least.

For instance, back in the days of too-much-is-not-enough, "viking horns" and similar devices redirected air flow in yaw toward downforce-producing elements downstream in an effort to minimize the loss.

[OO7]

My question is, given a constant speed, do all F1 cars lose some down force while in yaw/side-slip when compared to the same speed in a straight line?

Yes, but the better cars will tend to be those that lose the least.

For instance, back in the days of too-much-is-not-enough, "viking horns" and similar devices redirected air flow in yaw toward downforce-producing elements downstream in an effort to minimize the loss.

http://i.imgur.com/vhK72UW.jpg

Cheers bhall, that is what I suspected. I'm composing an idea I have that I'll post later, in which this loss could be beneficial to the sport.

BTW I didn't realise the primary purpose of those appendages was flow control in yaw.

[gixxer_drew]

Consider that a modern car, say an LMP1 goes from having downforce measured in tonnes to lift sufficient to pull it off the ground in ~90 degrees of yaw with an array of anti blow-over devices... yes a few degrees change everything.

[Greg Locock]

cite please, in my experience the first few degrees make little difference.

[Pierce89]

Consider that a modern car, say an LMP1 goes from having downforce measured in tonnes to lift sufficient to pull it off the ground in ~90 degrees of yaw with an array of anti blow-over devices... yes a few degrees change everything.

I've also seen wind tunnel studies of modern lmp cars in RCE that show little loss or even DF gains at small yaw angles

[Cold Fussion]

90 degrees is the opposite of a small yaw angle.

[gixxer_drew]

cite please, in my experience the first few degrees make little difference.

Only my experience in my every day workload, Yaw is significant. What sort of cars were you testing on?

Consider that a modern car, say an LMP1 goes from having downforce measured in tonnes to lift sufficient to pull it off the ground in ~90 degrees of yaw with an array of anti blow-over devices... yes a few degrees change everything.

I've also seen wind tunnel studies of modern lmp cars in RCE that show little loss or even DF gains at small yaw angles

Be careful with RCE stuff, isnt that done in a non moving belt scenario with relatively low downforce cars? I am talking about more than a "few" degrees of yaw, but not a lot more either.

Lets say you get a change of 3-5% front downforce, thats a very significant change for something at high level professional motorsports. Someone who worked to get those losses back will be significantly faster than someone who didn't.

Also how they are simulating yaw, I look at it with the car in attitude in every way, rotating and steered tires, roll + yaw, tire distortion. Just steering the tires has a pretty large effect so I would check first they are doing all that and not just having the car on a rotating table top. Thats pretty commonly used for low end wind tunnels and meant for crosswind NVH prod car stuff. Anyone saying the effect is small catches my suspension right away that they dont have a correct perspective on what is "small" or they arent testing properly. I can tell you right now in pro level, anyone not developing in yaw will not be competitive.

[gixxer_drew]

I had pondered on this since my last post and thought it would be good to put some perspective on it through numbers. I tried to come up with a good test, but it was difficult since the straight line tests I do now days are on geometries optimized in yaw. I pulled up an old file which was done completely in straight line only testing with a symmetry plane (CFD) alongside a workhorse wind tunnel that only does straight ahead in 20% scale. CFD and WT had good correlation but this was some years ago and compute power was limited as was WT budget.

I thought this would be a good test of an optimized geometry for straight ahead, to see how it will perform in yaw. This is a relatively high downforce car based on a production saloon. I updated the file just enough to complete a yaw run in CFD for a typical "average corner" scenario, similar to the newer processes.

Front downforce change was around -20% and rear was +5%. On this sort of car 20% front downforce change assuming balanced rear is well over a second a lap on a mid speed ~90 second lap.

This is my opinion, but I think SuperGT is a good example to look at because of how recently yaw work took off. One manufacturer began doing full motion WT work around 2008 (wasn't F1 nearly a decade earlier?). After the first car came out on the back of F1 based technologies (then a recently cancelled program), it was obvious they were sandbagging all season and won easily. Everyone looked at the redevelopment costs of existing cars, but chose not to. Then simply began the move to DTM regs which meant full time internal teams and all new wind tunnels.

[gixxer_drew]

90 degrees is the opposite of a small yaw angle.

I would agree, but the point I was trying to make is that if 90 degrees is a change in downforce/lift of a few TONNES on an LMP car, what might a "few" degrees be worth?

[Greg Locock]

Thanks for the info. For obvious reasons you left the yaw angle out. I has assumed that the yaw angle in still air would be the slip angle of the tires at maximum lateral force, ie 3-5 degrees.

[gixxer_drew]

Thanks for the info. For obvious reasons you left the yaw angle out. I has assumed that the yaw angle in still air would be the slip angle of the tires at maximum lateral force, ie 3-5 degrees.

I would say thats a good assumption between yaw and slip. But I think the issue you get into with rule of thumb numbers is what the steered angle is at any point on track fitting into what sort of car and tires you are on. You can imagine how any one number becomes a "wrong" number in a hurry. This is a big puzzle in terms of how to test the car aerodynamically.

[Greg Locock]

Oh and I forgot to add that of course in passenger car work the effect of the front wheel steer angle is ignored, I can believe it has a big influence on an open wheeler.

[gixxer_drew]

Oh and I forgot to add that of course in passenger car work the effect of the front wheel steer angle is ignored, I can believe it has a big influence on an open wheeler.

My limited experience with prod car work confirms yours. The steer angles in a crosswind lift or whatever are a fraction of a racing car in a low-mid speed corner.

[Tim.Wright]

The "slip angle" of the body airflow at the front axle is going to be the delta between the steer angle contribution and the tyre/suspension slip angle contribution (which are both in opposing directions).

When you steer in a certain direction, this imposes a kinematic slip angle of the body in the direction of the steering motion. The suspension slip angle (made up of the tyre slip and suspension K&C) then works in the opposite direction, cancelling out the kinematic component to varying degrees.

The resulting body (airflow) slip angle can be in the same direction as the steer angle, the opposite direction and even zero depending on how much steer angle and slip angle you have.

The first thing to get in order here is the coordinate system - otherwise nothing else will make sense. I use the SAE system which has steering, slip angle and lateral acceleration positive to the right.

To simplify the problem for CFD analysis, you can assume the tyres are always at their "optimum" slip of around 4-6deg (maybe 1-2deg less on the rear) then airflow slip angle is basically a function of corner radius.

Large radius corners will have low steer angles and the body slip angle will be more or less equal to the suspension slip angle of 4-6deg. So for a positive steer angle you get a negative body slip angle at the front axle.

Tight radius corners will have large steer angles but the same suspension slip angle so the body slip will be in the direction of the steered wheels but will be slightly reduced by the suspension slip angle. So for a positive steer angle you get a positive body slip angle at the front axle.

There will also be a condition (say medium radius) where the steer angle and slip angle are the same (i.e. you steer the front wheels 5deg to the right and the tyre/suspension slips 5deg to the left) in this case there is zero body slip on the front axle.

The rear axle can be assumed to have always the same airflow slip angle.