CFD - 2022 F1 Car

[JordanMugen]

It's really a halo screen rather than the much hoped-for aeroscreen. And it's even less pleasing to the eye than the halo.

Safety first! If FIA are serious about safety, they need to adopt the additional screen in front of the halo.

[ispano6]

Apologies if this is a repost. From https://aws.amazon.com/f1/
Formula 1 Works With AWS To Develop Next Generation Race Car
AWS re:Invent 2019 – Rob Smedley From Formula 1 Talks About Using AWS to Improve the Fan Experience

[Hoffman900]

With the struggles F1 teams have correlating their wind tunnel to CFD to real world performance, it will be interesting to see how these cars perform in anger.

[jjn9128]

With the struggles F1 teams have correlating their wind tunnel to CFD to real world performance, it will be interesting to see how these cars perform in anger.

A lot of the correlation issues in the wind tunnel stem from wall interactions - i.e. when pushing the front tyre wake outboard the outwash hits the wind tunnel walls and straightens more than it would at the track; and tyre squirt from the bulbous high aspect ratio tyres. With less outwash and lower profile tyres the correlation should be simpler.

[Hoffman900]

With the struggles F1 teams have correlating their wind tunnel to CFD to real world performance, it will be interesting to see how these cars perform in anger.

A lot of the correlation issues in the wind tunnel stem from wall interactions - i.e. when pushing the front tyre wake outboard the outwash hits the wind tunnel walls and straightens more than it would at the track; and tyre squirt from the bulbous high aspect ratio tyres. With less outwash and lower profile tyres the correlation should be simpler.

The bulbous tires and barge boards aren't anything new, and some of the wind tunnels were built after their existence. This is the whole reason for adaptive wall wind tunnels, scale models, etc. I have no doubt this isn't the full issue and part of it is that CFD is hard. I know a Boeing aerodynamics department manager and they struggle with it at times, and with resources that are more than the F1 grid combined, several times over. The amount of computing power needed to approximate the real world is staggering, and even with the best systems, it still can't fully replicate it.

As we all know, models are useless without correlation, be it CFD, FEA, engine pressure dynamics, etc.

A NASCAR builder I was listening to talk mentioned how despite all the R&D they do, and having great engineers, stuff still doesn't work out like what was thought on the computer screen. Those guys have the benefit of long term static rule set so their models are very well correlated. His words are the engineers come back with "but it's dimensionally correct", but sometimes it just doesn't work, and it's back to the drawing board and remeasuring and reworking their models to understand why and find further improvements.

It's fun to talk about, but with the hype surround the rules package and the better racing, I worry we're headed for a NASCAR COT situation when things don't play out exactly how people think or have been hyped.

[Just_a_fan]

All simulations rely on assumptions and approximations. Any "error" in either of those and the output will have an error larger still. The skill is often in getting the correct assumptions and approximations in place and allowing for the fact that the results are, likewise, an approximation.

I remember a lecturer who would ask what the answer to a question was. We'd do the work, run the equations etc. and then the first person to give "the answer" would be berated. The answer was never X, it was approximately X or some similar response. It was important to remember that the equations used contained assumptions and approximations in their background. Work stuff through from first principles and at some point there's usually a "fudge factor" somewhere.

[JordanMugen]

Work stuff through from first principles and at some point there's usually a "fudge factor" somewhere.

Direct Navier-Stokes solution with infinitely large cell count would be exact wouldn't it?

But obviously teams use a limited cell count with Reynolds-averaged Navier-Stokes with large amounts of fudging in the areas of boundary layer and turbulence.

[Just_a_fan]

Work stuff through from first principles and at some point there's usually a "fudge factor" somewhere.

Direct Navier-Stokes solution with infinitely large cell count would be exact wouldn't it?

Yes, but impossible. The only way to be exact is to build the car and run it in real conditions.

[SAEED]

Work stuff through from first principles and at some point there's usually a "fudge factor" somewhere.

Direct Navier-Stokes solution with infinitely large cell count would be exact wouldn't it?

Yes, but impossible. The only way to be exact is to build the car and run it in real conditions.

Assumptions are built into Navier- Stokes equation as well. Although you may argue about their impact.

[PhillipM]

The best way is to build a computer the exact same size and shape as the current universe, and try the car in that.

[mzso]

It's really a halo screen rather than the much hoped-for aeroscreen. And it's even less pleasing to the eye than the halo.

Safety first! If FIA are serious about safety, they need to adopt the additional screen in front of the halo.

Safety is never first. But I think they should engineer out that horrible center column. Using a transparent "glass" shield should help in itself. But maybe they could use something better than an utterly pedestrian polycarbonate. But at least laminate it for heaven's sake.

It's really a halo screen rather than the much hoped-for aeroscreen. And it's even less pleasing to the eye than the halo.

I'd bet it would have a big impact on the downstream aero of the cars - the rear wing might take a big loss, for example. It would be interesting to see some plots of the flow with aeroscreen.

Not isn't, it's the central column that makes the halo look awkward. That bit of glass hardly has any visual impact.
"Ewww it's different" doesn't count, in my opinion.

[Just_a_fan]

It's really a halo screen rather than the much hoped-for aeroscreen. And it's even less pleasing to the eye than the halo.

Safety first! If FIA are serious about safety, they need to adopt the additional screen in front of the halo.

Safety is never first. But I think they should engineer out that horrible center column. Using a transparent "glass" shield should help in itself. But maybe they could use something better than an utterly pedestrian polycarbonate. But at least laminate it for heaven's sake.

It's really a halo screen rather than the much hoped-for aeroscreen. And it's even less pleasing to the eye than the halo.

I'd bet it would have a big impact on the downstream aero of the cars - the rear wing might take a big loss, for example. It would be interesting to see some plots of the flow with aeroscreen.

Not isn't, it's the central column that makes the halo look awkward. That bit of glass hardly has any visual impact.
"Ewww it's different" doesn't count, in my opinion.

The centre column is important because it takes the loads down in to a strong part of the tub. Without the front leg, the loads on the rear attachment points would be "interesting" and ensuring the tub shoulders didn't get ripped out would require a lot of extra, heavy, reinforcement. Making a screen that would transmit those loads is likely to require a screen that is rather thick and that increases the chances of optical effects. It's not impossible but it's possibly more effort/cost than is necessary for a purely aesthetic issue. Just wrapping the existing halo in a screen "because I don't like the halo" is also no justification.

The halo is the best, least-worst solution to guarding the driver's head whilst keeping the car "open cockpit".

[mzso]

It's not impossible but it's possibly more effort/cost than is necessary for a purely aesthetic issue. Just wrapping the existing halo in a screen "because I don't like the halo" is also no justification.

Well engineers should like a challenge. Besides it would only have to be designed once, and only manufactured a few times. I don't expect them hitting damaging objects too often.

But it's not just aesthetic. It blocks a chunk of visibility. Particularly troublesome if the driver needs to look up a bit Spa (and Mugello and Portimao) come in mind.

[Zynerji]

I suggest Analog Computers. Using wave cancellation/augmentation in an acoustic system would accelerate the processing on the NS equations exponentially over conventional computers.

https://hackaday.com/2019/05/03/swiss-c ... -computer/

[Just_a_fan]

It's not impossible but it's possibly more effort/cost than is necessary for a purely aesthetic issue. Just wrapping the existing halo in a screen "because I don't like the halo" is also no justification.

Well engineers should like a challenge. Besides it would only have to be designed once, and only manufactured a few times. I don't expect them hitting damaging objects too often.

But it's not just aesthetic. It blocks a chunk of visibility. Particularly troublesome if the driver needs to look up a bit Spa (and Mugello and Portimao) come in mind.

Any device strong enough to, for example, punch through armco or deflect a flying car, is going to block vision at some point. An aero screen would either fail to perform or it would need to be so thick, or locally reinforced, that it too would block vision around the periphery.