The Impact of Wake Turbulence on Downforce.

[Pierce89]

One thing that gets mentioned a lot is that cars with ground effects should lose less downforce when following others. Is that true? For example if we take an extreme type of car which has no front wing, makes ~80% of it overall downforce from the ground effects and uses the rear wing to mainly to balance the car (let's just imagine the aero balance would be good for the sake of argument).

Would that kind of car actually create better racing? Would the car behind keep most of its downforce? Would the car behind get drag benefit from running in lower pressure air? How would the turbulence affect the car behind?

For the 2017 season I was hoping the regulations would move lots of the downforce to the the bottom of the car but for racing and overtaking what would the data say about that kind of aero arrangement?

To be honest, your question has no straight forward answer. If all the DF is made by underbody venturis with the low pressure peak under the CoG, then theoretically, they would still lose some downforce but balance wouldn't change as much as now. But in real life most floors have 2 suction peaks, one at leading edge of floor and one at diffuser kink line. This makes them susceptible to balance change in traffic even though it's "ground effect" downforce

[livinglikethathuh]

It the relationship between downforce in wake turbulence and distance exponential?

it could probably be simplified to that, but like so many other dynamic problems, it's not really something that can be written as an equation, only approximated.

Turbulence, due to its chaotic nature, cannot be fully expressed in equations, but the dissipation of downforce is by air molecules slowing each other down. This friction is loosely proportional to their speed, think of air molecules like objects flying through the air and having their own coefficient of drag. Therefore, the force that slows them down is proportional to the square of their speed, which makes their speed/time relation a decaying third order equation. In other words, the reduction of turbulence per unit time is inversely proportional to the time spent since the creation of the turbulence.

Another point; the correlation between aero performance of a car and the turbulence of the air it goes through must be analysed, and IMO it wouldn't be as simple as a third order equation. Furthermore, this turbulence/downforce relation would vary from car to car.

Though I'm pretty sure teams know how much performance they lose in different strengths of downforce, and are making design choices based on it. For example, Mercedes are expecting to have the race in clean air, so they look for extra clean air performance in exchange for huge downforce loss in dirty air, whereas Toro Rosso with its hotshot drivers opt for aero devices that play nice with turbulent air in exchange for all-out clean air/qualifying aero performance.

[Just_a_fan]

The Navier-Stokes equations do account for turbulence but they're basically impossible to use for anything the size of a car.

[Tommy Cookers]

iirc the leading book on wind tunnel work says that the Navier-Stokes equations (and so CFD) are rather .... wrong

[Just_a_fan]

CFD approximates N-S. That's one of the issues with it and why wind tunnels are still essential. To compute N-S directly is computationally prohibitive even with today's supercomputers.

[livinglikethathuh]

iirc the leading book on wind tunnel work says that the Navier-Stokes equations (and so CFD) are rather .... wrong

Care to elaborate?
Many, MANY principles in aerodynamics are obtained through simplifications of N-S equations.
If they are wrong, I may have to burn my diploma (that I will hopefully acquire in June)

[trinidefender]

iirc the leading book on wind tunnel work says that the Navier-Stokes equations (and so CFD) are rather .... wrong

Care to elaborate?
Many, MANY principles in aerodynamics are obtained through simplifications of N-S equations.
If they are wrong, I may have to burn my diploma (that I will hopefully acquire in June)

Something I was reading before is that some CFD calculations use more than just N-S equations to arrive at their turbulence and other modelling.

[OO7]

Regarding the 55% downforce loss at half a car length and general loss of downforce in wake turbulence, the problem compounds because as downforce bleeds off ride height increases which will lead to further downforce loss.

I wonder how much of this 55% total downforce loss is simply down to the increase in ride height?

[Just_a_fan]

iirc the leading book on wind tunnel work says that the Navier-Stokes equations (and so CFD) are rather .... wrong

Care to elaborate?
Many, MANY principles in aerodynamics are obtained through simplifications of N-S equations.
If they are wrong, I may have to burn my diploma (that I will hopefully acquire in June)

Something I was reading before is that some CFD calculations use more than just N-S equations to arrive at their turbulence and other modelling.

All CFD (except DNS) uses turbulence approximations of one sort of another. Most CFD will be using RANS (Reynolds-averaged Navier-Stokes equations) with one of a number of turbulence "model" such as "k-epsilon" or "shear stress transport". This gives a good compromise between accuracy and computer requirements for most uses. You can also use RANS for most of the flow and embed the more accurate (but more computer intensive) LES (Large Eddy Simulation) for specific areas of interest. This is called hybrid RANS-LES. In this case you would use "cheap" RANS for areas of simple flow (attached and steady) e.g. the top of the nose/tub or the leading edges and upper surface of a F1 wing. You would use "expensive" LES for areas of unsteady (turbulent) flow such as the lower surfaces of F1 wings (in areas where you feel separation of the flow is occuring), around suspension arms and wheels etc.

I would bet that the main difference between the teams with good area and those with ok aero is down to how well they deal with turbulence modelling. I'd bet a team like RedBull or Mercedes are using some cutting edge hybrid RANS-LES on some very expensive hardware. They may even be writing some of their own systems. A team at the back of the grid might be using something off the shelf that is at least a couple of years behind.

[Cold Fussion]

Don't forget teams are limited to 30 terraflops (presumably continuously) of super-computing power.

[Just_a_fan]

So having a very good way of dealing with turbulence in finely defined critical areas is even more important.

[riff_raff]

That's a great point. How a car reacts to wake turbulence is probably a greater concern. A car that has a more benign response to wake turbulence going into a corner is much easier for the trailing driver to operate at the limits of braking and steering.

[toraabe]

If the rule was that groundeffect with venturis and a scupltured unerfloor to generate the downforce ( like Indy ) and simple front and rear, and strictly defined angles on the elements, this problem would not been the case. Just look at this video how close they are able to follow.. https://www.youtube.com/watch?v=EWF_VEtDnKo

[OO7]

If the rule was that groundeffect with venturis and a scupltured unerfloor to generate the downforce ( like Indy ) and simple front and rear, and strictly defined angles on the elements, this problem would not been the case. Just look at this video how close they are able to follow.. https://www.youtube.com/watch?v=EWF_VEtDnKo

Indy cars also suffer with this problem as can be seen on the ovals, however it is not as severe as with as F1 car.

[toraabe]

If the rule was that groundeffect with venturis and a scupltured unerfloor to generate the downforce ( like Indy ) and simple front and rear, and strictly defined angles on the elements, this problem would not been the case. Just look at this video how close they are able to follow.. https://www.youtube.com/watch?v=EWF_VEtDnKo

Indy cars also suffer with this problem as can be seen on the ovals, however it is not as severe as with as F1 car.

Specially after the new rules, but not in the 90"
If you look at the picture http://superhachi.com/theory/downforce/indydiff.JPG
The underfloor tunnels had to stop at the diff. And the angle and the total volume is clearly defined in the rules.
Now they are able to continue towards the end http://static2.businessinsider.com/imag ... actice.jpg And therefore creating a bigger Wake behind the car. Still most of the downforce in Indy is created through this floor or what Colin Chapman said. " Groundeffect" The biggest mistake was to ban groundeffect in total in F1. Indy has implemented this very well.