Evolution of Aerodynamic Testing in F1

[jjn9128]

I think it's more likely a scale and compressibility effect - like the Williams issue. We have to remember always that Wind Tunnels and CFD are just simulations - they approximate reality and there are compromises.

I understand on a basic, literal level what scale and compressibility are in a wind tunnel, but I don't know what you really mean by scale and compressibility as the problem. What is it that wind tunnels can't simulate (accurately?) due to scale and compressibility effects? Or, how do some teams handle that limitation better?

Wind tunnel testing is limited by the FIA to 60% models with 60m/s wind speed (216km.hr = 135mi/hr), i.e. they can only simulate the air flow around the car at the equivalent to 36m/s = 129.6km/hr = 81mi/hr for the full scale car. So they could be having issues with the car at speed - there's a separation which isn't occurring at the low Reynolds numbers of the wind tunnel, or a boundary layer is transitioning earlier or something. They are also limited to using air at atmospheric pressure - so they can't increase the Reynolds number in the tunnel by compressing the air or using a different gas.

The speed of sound is 343m/s = 1234.8km/hr = 771.8mi/hr - in dry air at sea level - normally for subsonic aero you consider the air to behave as an incompressible fluid, but that's only true up to ~Mach 0.3. Under the floor and front wing in particular air will be travelling between 2 and 2.5x the speed of the car (Cp>-4). At 60m/s in the wind tunnel that air is travelling at 120-150m/s or Mach 0.35-0.43 so the air is going to be compressible. But at the dynamically similar speed (36m/s) on the track the air under the car is only at Mach 0.2-0.26... so the air in the tunnel is compressible but with dynamic similarity on track (at the same Reynolds number) it isn't. Including flow compressibility in CFD simulations has quite a big effect - especially under the floor, which is where most of the downforce comes from so it's an important area.

Now no car will travel at 81mi/hr for long - and they do their constant speed aero testing at ~240km/hr = 150mi/hr to more closely match to the average speed for the car where maximum downforce is required - which also more closely matches the Mach number of the wind tunnel, but not the Reynolds number. So you can get some things right but others wrong or the other way around - best to test on the track at an appropriate speed.

In CFD is easier to match everything (scale/Reynolds#/Mach#) than wind tunnels but there are limitations there too.

This is a long answer and I'm not sure if it answered the question....

[Vyssion]

The only thing I would add to jjn's reply is the example case of tyres...

I will preface this with stating that my background is much more heavily dominated by CFD methods, rather than experimental methods, but I have had some dealings with them.

Tyres are the bane of all wind tunnel, CFD and even track testing aerodynamicist trying to work out what the hell is going on. You have 6 vortices which are shed from a rotating tyres, and two counter rotating at the top, two and the bottom (squirt region) and then two in the middle which have the lovely habit of flip flopping around which direction they decide to randomly shed in. Wind tunnel tyres include special pneumatics within them, and all sorts of other software controlled wizardry to try and best match what is going on with regards to things like rotation balancing, toe, camber, caster, squirt, vortex shedding, deformation of sidewall, tyre slippage in straight line and cornering, warp of the tyre, frequency of vortex shedding (strouhals number), any oscillitory behaviour specific to the natural harmonics of the rubber, loading in 6 degrees of freedom..... blah blah blah... the list goes on for ages...

And as far as I am aware at least, Pirelli supply a set of tyres to each team to copy each of their compounds - but to say that they exactly model all of these things in addition to all of the scaling factors of Reynolds Number (which is a measure of turbulence based on a characteristic length and veloccity) within what is an extremely separation dominated flow regime, as well as Mach number of not only the forward airspeed, but also the rotating tyre speed relative tot he full scale.... Its nigh on impossible to fully resolve EVERYTHING at each speed or setup you would like to run.

It becomes important then to work out at which conditions you can check certain things at and make extremely careful and well-educated assumptions from those single datum points as to how the flow behaviour is LIKELY to be at speed.. Now, granted now a days this process is quite robust and teams are amazing in what they can do with this information etc, but there are some things that are simply just too complicated with how the regulations are currently set up to deal with maximum fidelity.

This is just for tyres, but it isnt hard to see how other bits and pieces of the car which are heavily dependant on specific flow regimes or types causing problems when trying to see how they work in a tunnel vs. on the track.

[godlameroso]

Are the vortices shed off the tire itself or do they start at the boundary layer(which grows as the tire spins faster).

[Vyssion]

Tyres are the bane of all wind tunnel, CFD and even track testing aerodynamicist trying to work out what the hell is going on. You have 6 vortices which are shed from a rotating tyres, and two counter rotating at the top, two and the bottom (squirt region) and then two in the middle which have the lovely habit of flip flopping around which direction they decide to randomly shed in. Wind tunnel tyres include special pneumatics within them, and all sorts of other software controlled wizardry to try and best match what is going on with regards to things like rotation balancing, toe, camber, caster, squirt, vortex shedding, deformation of sidewall, tyre slippage in straight line and cornering, warp of the tyre, frequency of vortex shedding (strouhals number), any oscillitory behaviour specific to the natural harmonics of the rubber, loading in 6 degrees of freedom..... blah blah blah... the list goes on for ages...

Are the vortecies shed off the tire itself or do they start at the boundary layer(which grows as the tire spins faster).

Not entirely sure what you mean here, but I'll try and explain the process with a couple pictures and you can tell me if I didn't understand the question or not. If all else fails, google "isolated rotating wheel CFD" or something along those lines, you will be able to come up with a whole host of papers and studies looking into this sort of thing.

What I think you're asking is whether the vortices would be the same in structure and number whether the tyre was stationary or not, or whether their structure and size is dependent on how fast they are going; i.e. is there some "critical rotation speed" where 4 vortices become 6, or a flip-flop happens to the centre ones...? Is that about right?

The short answer is that the rotation of the wheel heavily defines how the aerodynamic flow will develop, and that there is probably some relative Reynolds Number below which these vortical structures do not form, however, I would imagine this to be impractically slow in terms of wheel rotational speeds. How much they develop, on the other hand, is linked to their rotational speed - to a point!

With wheels in any sort of flow, in general terms, the separation and wake characteristics are strongly influenced by the rotation of the wheel; i.e. the separation point may change by as much as 90% when compared to that of a stationary wheel, whilst drag can be close to 20% less than the rotating wheel. Downforce can also be up to 40% greater, though there are quite a few instances where the wheels actually generate lift, instead of downforce.

Here are a few pictures that hopefully will give you a better understanding of whats going on:












These two videos here (I personally would read the "Formula 1" part in the title quite loosely , and the domain looks way too small - but its crude LES and so should give you an idea of whats going on regardless) should show how the 3D vortical structures may look like, given how it looks on a single plane through space.

[jjn9128]

I think that LES is the work of Axerio which IIRC was fairly low Reynolds number but F1 model scale (grooved tyre at 50 or 60%). Saddington and Knowles is another good reference, but again grooved tyres - there's a lot of research out there with old F1 grooved tyres from 10-15 years ago.

[hardingfv32]

Just watching a VW Pikes Peak 2018 car build video. In it they show a lot of activities in the wind tunnel. At one point a technician pulls a tire warmer off the model's tire. Why would tire warmers be needed?

Brian

[J.A.W.]

Just watching a VW Pikes Peak 2018 car build video. In it they show a lot of activities in the wind tunnel. At one point a technician pulls a tire warmer off the model's tire. Why would tire warmers be needed?

Brian

To warm the tyres & reduce variables to best replicate* actual use, as far as practicable.

*Hot air flows measurably differently, as it happens..

[jjn9128]

Just watching a VW Pikes Peak 2018 car build video. In it they show a lot of activities in the wind tunnel. At one point a technician pulls a tire warmer off the model's tire. Why would tire warmers be needed?

Brian

I can't find the video in question - could you link it?

They did model and full scale testing - model at Dallara, full scale at one of the VAG wind tunnels. I imagine this would be part of a full scale test... I wonder if they're warmers or just a protective wrap??

[godlameroso]

Just watching a VW Pikes Peak 2018 car build video. In it they show a lot of activities in the wind tunnel. At one point a technician pulls a tire warmer off the model's tire. Why would tire warmers be needed?

Brian

To warm the tyres & reduce variables to best replicate* actual use, as far as practicable.

*Hot air flows measurably differently, as it happens..

Yep, little known fact about air, it's kinematic viscosity increases with temperature, the effect is small but there none the less.

[jjn9128]

Just watching a VW Pikes Peak 2018 car build video. In it they show a lot of activities in the wind tunnel. At one point a technician pulls a tire warmer off the model's tire. Why would tire warmers be needed?

Brian

I can't find the video in question - could you link it?

They did model and full scale testing - model at Dallara, full scale at one of the VAG wind tunnels. I imagine this would be part of a full scale test... I wonder if they're warmers or just a protective wrap??

Aah, found it! From the model scale testing.


It's not something I've come across myself.... as others have mentioned there's a temperature effect on the air properties but at model scale this is all wrong anyway...

I imagine it's for the tyre pressures, as the tyres roll on the treadmill they heat up (as does the belt because it's being sucked onto the plenum - to prevent it lifting and stalling the floor - which creates a huge amount of friction & heat) so the tyre profile slightly changes as a run progresses - which is obviously not ideal for absolute accuracy. So I'd guess to keep tyre pressure constant. That's just a guess though!!

[PhillipM]

Few things, main thing as you said would be the tyre profile and the squirt around the tyre contact patch - you could do it with higher pressures but that doesn't account for the softer rubber and carcass at running temperatures.
The next one is the heat in the upright, brakes and wheel alters the way the brake ducts and tins flow quite a lot, and as already stated, the air over the surface of the tyre itself is impacted marginally by temp.

[Frank06]

Thanks for posting links to your work on aero testing techniques, very interesting.

[hollus]

....The speed of sound is 343m/s = 1234.8km/hr = 771.8mi/hr - in dry air at sea level - normally for subsonic aero you consider the air to behave as an incompressible fluid, but that's only true up to ~Mach 0.3. Under the floor and front wing in particular air will be travelling between 2 and 2.5x the speed of the car (Cp>-4). At 60m/s in the wind tunnel that air is travelling at 120-150m/s or Mach 0.35-0.43 so the air is going to be compressible. But at the dynamically similar speed (36m/s) on the track the air under the car is only at Mach 0.2-0.26... so the air in the tunnel is compressible but with dynamic similarity on track (at the same Reynolds number) it isn't. Including flow compressibility in CFD simulations has quite a big effect - especially under the floor, which is where most of the downforce comes from so it's an important area....

Is this compressibility also noticeable over the tires? The top of each tire is moving at twice the car's speed. Isn't this equivalent to the situation under the floor that you are describing?

P.S. Is the air below the car really moving at twice the speed of the car relative to the car? Wouldn't that mean that, relative to the track, it is moving backwards at 300km/h???

[jjn9128]

Is this compressibility also noticeable over the tires? The top of each tire is moving at twice the car's speed. Isn't this equivalent to the situation under the floor that you are describing?

P.S. Is the air below the car really moving at twice the speed of the car relative to the car? Wouldn't that mean that, relative to the track, it is moving backwards at 300km/h???

The top of the tyre is travelling at the speed of the car but in the opposite direction!? The bottom of the tyre is moving at the speed of the car... assuming no slip.

Cp under the car peaks at between -3 to -4.

where is local velocity near the surface and is the freestream/road speed. So the ratio of local to freestream velocity is:

Which for Cp = -3 is the square root of 4 which is 2x the freestream velocity.

[hollus]

The frame of reference of the moving road in the wind tunnel might muddy the waters, but on the road, the point of the tyre in contact with the tarmac is, for an brief instant, not moving at all (the road does not move and tyre slip is negligible). With the center of the tyre moving constantly at the same speed as the car, and the wheels averaging the same speed as the car, the top of the tire must move twice as fast.

Maybe this illustrates it better.