CFD - 2022 F1 Car

[godlameroso]

I am simulating a car (not a F1, it is nearer to a F3) with a F1 2021 floor layout: I am pretty sure (as you wrote on the article) that the optimum rake will be around 0°, no matter if the front wing will be (a bit) more distant from the ground. You will have a significant drag reduction and you don't have a diffuser peak to balance.

Perhaps a little bit of rake will be needed just for balance, the shift in rake at speed may also have very interesting effects on that balance. Particularly roll, and pitch.

What if more, not less body roll and pitch is favorable to the 2021 generation of aero rules. It would certainly force drivers to adapt a new style of driving, pitching and rolling the cars to set up for corners.

Very light brake application on corner entry to get the front end to turn in, and having to semi inertial drift the turns to get the most of the aero platform.

[Jolle]

With the skirts on the rear axle, isn’t a high rake possible if teams (read: Newey) finds ways to seal the floor again with vortexes up to this skirt? Could be a trail and error how far you can go with rake per circuit before you stall the floor.

[Just_a_fan]

Where do they generate the vortices? The floor has three strakes either side and that's about it. F1 aero is going to be close to spec. Where does Newey find the bits to play with.

Obviously, once Newey has figured it out, the other teams will copy him. Just like that last few years. Oh, wait...

[hollus]

So, if the front wing is overpowered... would the rules allow to sacrifice its center section to replicate the current y-250 vortex?

[godlameroso]

So, if the front wing is overpowered... would the rules allow to sacrifice its center section to replicate the current y-250 vortex?

Maybe, shape the wing slots to alleviate the separation at the tunnel entrance. We have maximum surface volume but they can be smaller, there is some sculpting possible, but the general shape will remain the same.

[jjn9128]

So, if the front wing is overpowered... would the rules allow to sacrifice its center section to replicate the current y-250 vortex?

Unfortunately not. I pondered the same thing in the other '21 aero thread.

Where do they generate the vortices? The floor has three strakes either side and that's about it. F1 aero is going to be close to spec. Where does Newey find the bits to play with.

Obviously, once Newey has figured it out, the other teams will copy him. Just like that last few years. Oh, wait...

The more I look at these rules the more I realise they are funneling teams into a particular spec, all the teams can then really do is faff about with the detail. The shape of the underbody tunnels and fences could be slightly different between teams, but it'd be difficult to tell even under close scrutiny.

[timbo]

Great stuff!

Guys, if that would help I have a semi decent PC, an older intel 6-core with 32Gb RAM, and if software installation is not too complicated, I could totally run the calculations for several hours a day.

[CMSMJ1]

Great thread and news post on first page.

Just like old times

Keep on chaps

[Andres125sx]

Amazing job, I have no idea about aero but that looks like really hard work to do. Congratulations and thanks for sharing!

I see once you´ve done the hard work about modelling, representing results looks like an even bigger headache. How do you analyse results? With those gifs or with numerical data? Be kind with me, I said I have no idea Gifs are not numerical data, but I guess numerical data about this would be hideous and completely unmanageable. Any option I can think about cause me a big headache when I imagine how hard will be that work, so my most sincere congratulations and admiration

[jjn9128]

Amazing job, I have no idea about aero but that looks like really hard work to do. Congratulations and thanks for sharing!

I see once you´ve done the hard work about modelling, representing results looks like an even bigger headache. How do you analyse results? With those gifs or with numerical data? Be kind with me, I said I have no idea Gifs are not numerical data, but I guess numerical data about this would be hideous and completely unmanageable. Any option I can think about cause me a big headache when I imagine how hard will be that work, so my most sincere congratulations and admiration

Thank you everyone we try. Vyssion may disagree that the modelling is the hard bit as he runs the CFD while I do the CAD

We tend to look at things graphically (that's just imposing the numerical data onto the mesh) as we're only looking at the generalities of these things rather than fine detail. Vyssion has a script set up to auto-output about 150 4k images for each simulation, so we have direct comparison from case-to-case, we then try to select a few which describe the thesis of an article without overloading the reader.

Gifs are a good way to show the flow around the car as you can put 1 image box into an article which would otherwise require 20/30 pictures. It's also a good way to see how things flow from 1 slice to the next (so along and around the car) rather than looking at individual static slices.

Just looking at data graphically can be a bit misleading as you're superficially comparing one coloured blob to another maybe slightly different blob. In our cases so far the differences are so obvious it's hardly worth going any deeper. My PhD is full of 1d/2d as well as 3d plots of wakes so it's certainly possible to represent data in other ways and give value to differences, but you start requiring a higher level of very specific education, or huge swathes of explanatory text to fully understand the plots though. We're trying to keep these articles accessible.

So in short, yes it could be a bit of a nightmare, but we have standard plots and took the pain in the first case to set it all up so everything is self consistent.

[maunde]

Amazing job, I have no idea about aero but that looks like really hard work to do. Congratulations and thanks for sharing!

I see once you´ve done the hard work about modelling, representing results looks like an even bigger headache. How do you analyse results? With those gifs or with numerical data? Be kind with me, I said I have no idea Gifs are not numerical data, but I guess numerical data about this would be hideous and completely unmanageable. Any option I can think about cause me a big headache when I imagine how hard will be that work, so my most sincere congratulations and admiration

Thank you everyone we try. Vyssion may disagree that the modelling is the hard bit as he runs the CFD while I do the CAD

We tend to look at things graphically (that's just imposing the numerical data onto the mesh) as we're only looking at the generalities of these things rather than fine detail. Vyssion has a script set up to auto-output about 150 4k images for each simulation, so we have direct comparison from case-to-case, we then try to select a few which describe the thesis of an article without overloading the reader.

Gifs are a good way to show the flow around the car as you can put 1 image box into an article which would otherwise require 20/30 pictures. It's also a good way to see how things flow from 1 slice to the next (so along and around the car) rather than looking at individual static slices.

Just looking at data graphically can be a bit misleading as you're superficially comparing one coloured blob to another maybe slightly different blob. In our cases so far the differences are so obvious it's hardly worth going any deeper. My PhD is full of 1d/2d as well as 3d plots of wakes so it's certainly possible to represent data in other ways and give value to differences, but you start requiring a higher level of very specific education, or huge swathes of explanatory text to fully understand the plots though. We're trying to keep these articles accessible.

So in short, yes it could be a bit of a nightmare, but we have standard plots and took the pain in the first case to set it all up so everything is self consistent.

First of all cheers to Vyssion and yourself for your interpretation of the 2021 regs, it's always nice to see the pretty colours

I have a few questions:

1 - How did you go about modelling this car, what software did you use and what was your basic process: model the legal volumes than put something that fits within that? I assume you did individual parts seperately and blended them together (as mentioned in the regs)?

2 - This is more of a question for Vyssion but how did you plot your Lambda 2 for vorticity? I assume you used a vorticity core feature in Ansys, but how did you choose your Lambda value for the iso surface?

Thanks in advance, and keep up the great content!

Maunde

[jjn9128]

First of all cheers to Vyssion and yourself for your interpretation of the 2021 regs, it's always nice to see the pretty colours

I have a few questions:

1 - How did you go about modelling this car, what software did you use and what was your basic process: model the legal volumes than put something that fits within that? I assume you did individual parts seperately and blended them together (as mentioned in the regs)?

2 - This is more of a question for Vyssion but how did you plot your Lambda 2 for vorticity? I assume you used a vorticity core feature in Ansys, but how did you choose your Lambda value for the iso surface?

Thanks in advance, and keep up the great content!

Maunde

The car is modeled with OnShape which is a free online CAD package - IIRC the guys responsible were previously at Dassult and a lot of functionality is familiar from Solidworks even if the UI is considerably simplified. It's not bad, it's also not great, high quality surfacing can be a massive pain in the backside. There are certainly things I'd like to do better but can't, but it's free and beggars can't be choosers!!

I start as any team would, position the big masses, set a wheelbase, and produce the rules boxes. I have a basic engine model and create a simple gearbox case and radiators so I have internal things to wrap with the bodywork, which hopefully makes things a bit more realistic than some of the F1 CAD models I've seen. The bodywork then is a case of reading the rules and making sure it's legal. Each part is made in it's particular volume and then the car is assembled.

There are a few volumes missing in the current draft of the rules, some I think have yet to even be defined by the FIA; e.g. the part ahead of the drivers bum/crotch at the front of the floor, the plank, the side impact structures (a new single structure is going to replace the double impact spars), the more futuristic style halo (if it actually appears, the rules mention the current one we used), if there's a different cockpit entry template - so where those are we had to make some assumptions and tried to err on the side of caution.

[CAEdevice]

Hi, would you share the max/min floor distance from the ground (I mean the flat region where the plank is bolted). Thanks!

[jjn9128]

Hi, would you share the max/min floor distance from the ground (I mean the flat region where the plank is bolted). Thanks!

Min at the front edge of the plank is ~15mm ahead of that is slightly curved upwards

[CAEdevice]

Hi, would you share the max/min floor distance from the ground (I mean the flat region where the plank is bolted). Thanks!

Min at the front edge of the plank is ~15mm ahead of that is slightly curved upwards

Thanks!