Thanks a lot Graham!! It goes to show to what extent aerodynamic performance is in the cm-mm level of detail - incredible how much the performance changed! And hopefully this is not the end of it - perhaps we'll go for a second aero package if the community wants it :)
You should do an FSAE aerodynamics project similar to this! The rules are much simpler and will make the project more accessible to people. Coupled with the relevance to actual FSAE teams, I'm sure it would have greater reach than this already successful F1 project.
That's a very cool suggestion Nando. We get request for free sims often from FSAE teams but we simply can't do that, as they require a lot of resolution. But showcasing a public version could help to inspire the teams. Perhaps we should team up with one of the teams, asking them to open source their 3D model!
Surprised there were no modifications to the plank area of the underfloor as that can interact with both flow upstream and downstream. If I were to speculate, I'd say that the strakes/fences at the leading edge of the floor act as an air restriction, and it is that restriction that creates the low pressure, and drives the inwash around the edge wing. The biggest challenge to improve the performance of the strakes, in my dumb opinion, is to generate as much outwash, while limiting secondary flows. The regulations allow a fair bit of fillet radius both at the top and lower sections, up to 50mm which means you can have basically an endplate at the bottom of the floor fences. You can make the strakes look like a snow plow if you exploit the maximum fillet radius. The higher the pressure at the fences, the lower the pressure downstream. The regulations also make you get creative to force outwash at the rear. The turbulence of the tires gets cleaned up dramatically for every degree of airflow outwash you can get away with. Several studies, in both CFD and real world testing show that the turbulence turns into 2 vortices when the relative airflow is outwashing around the wheel. I can't share external links on youtube, but basically if you have a spinning tire that's toed out 5 degrees or more onto oncoming flow, the wake turns into a vortex which is much cleaner than airflow going straight on. Now in F1 you can't have crazy toe angles, because the car would be nearly uncontrollable, especially at the rear. You can however force air outward, so that the air interacts with the tire as if it were toed out.
Hi Rolando, could you perhaps drop some sketches of your ideas on our Reddit channel? www.reddit.com/r/airshaper/ Thanks a lot! PS: Aero Update is going through simulation as we speak - we did work a lot on the fences so hope you find the results interesting!
@@AirShaper I deleted my reddit account for personal reasons and out of principle refuse to go back. I would love to show you some crude drawings, images, research papers, and other things that have led me to my opinions. If you have an E-mail or some way to send external links/images/etc, doesn't have to be personal, a throwaway e-mail is fine, I would love to share my thoughts. I removed myself from discord for similar reasons.
Is this model also accounting for vehicle dynamics (compression of suspension at different speeds, sensitivity to pitch if you hit a curb etc.) ? Really interesting to watch you guys work
@@AirShaper in that case, be careful around modeling/creating aero devices that are very "sharp"/sensitive to minor changes in flow direction/pressure gradients. This was part of the reason blamed for Senas death: the car was designed for active aero, had it stripped out and was therefore far too sensitive for a human to control. It's why current setups aren't "optimal", but rather have a very large range of near optimal roll, pitch and yaw angles, as well as speeds
@@camofelix Very true - a bit lower performance but lower sensitivity to settings can inspire more confidence & create a safer car. Building a full aero map should indicate these sentivities.
I’m doing a project for university and I can’t find anywhere informations about Aerofoil and Angle of attack used on F1 cars, it’s urgent for me to find these info. Any help?
We did apply our "adjoint shape optimization technique to optmize parts of the wing, the strakes, ...: ruclips.net/video/WpFNwfHwFn4/видео.html ruclips.net/video/cZAhPQFINZ8/видео.html
Epic project and respect to everyone who had input. That downforce increase was truly significant.
Thanks a lot Graham!! It goes to show to what extent aerodynamic performance is in the cm-mm level of detail - incredible how much the performance changed!
And hopefully this is not the end of it - perhaps we'll go for a second aero package if the community wants it :)
@@AirShaper❤
You should do an FSAE aerodynamics project similar to this! The rules are much simpler and will make the project more accessible to people. Coupled with the relevance to actual FSAE teams, I'm sure it would have greater reach than this already successful F1 project.
That's a very cool suggestion Nando. We get request for free sims often from FSAE teams but we simply can't do that, as they require a lot of resolution. But showcasing a public version could help to inspire the teams. Perhaps we should team up with one of the teams, asking them to open source their 3D model!
Surprised there were no modifications to the plank area of the underfloor as that can interact with both flow upstream and downstream. If I were to speculate, I'd say that the strakes/fences at the leading edge of the floor act as an air restriction, and it is that restriction that creates the low pressure, and drives the inwash around the edge wing.
The biggest challenge to improve the performance of the strakes, in my dumb opinion, is to generate as much outwash, while limiting secondary flows. The regulations allow a fair bit of fillet radius both at the top and lower sections, up to 50mm which means you can have basically an endplate at the bottom of the floor fences. You can make the strakes look like a snow plow if you exploit the maximum fillet radius. The higher the pressure at the fences, the lower the pressure downstream. The regulations also make you get creative to force outwash at the rear. The turbulence of the tires gets cleaned up dramatically for every degree of airflow outwash you can get away with. Several studies, in both CFD and real world testing show that the turbulence turns into 2 vortices when the relative airflow is outwashing around the wheel. I can't share external links on youtube, but basically if you have a spinning tire that's toed out 5 degrees or more onto oncoming flow, the wake turns into a vortex which is much cleaner than airflow going straight on. Now in F1 you can't have crazy toe angles, because the car would be nearly uncontrollable, especially at the rear. You can however force air outward, so that the air interacts with the tire as if it were toed out.
Hi Rolando,
could you perhaps drop some sketches of your ideas on our Reddit channel?
www.reddit.com/r/airshaper/
Thanks a lot!
PS: Aero Update is going through simulation as we speak - we did work a lot on the fences so hope you find the results interesting!
@@AirShaper I deleted my reddit account for personal reasons and out of principle refuse to go back. I would love to show you some crude drawings, images, research papers, and other things that have led me to my opinions. If you have an E-mail or some way to send external links/images/etc, doesn't have to be personal, a throwaway e-mail is fine, I would love to share my thoughts. I removed myself from discord for similar reasons.
@@rolandotillit2867 You can reach me directly at info@airshaper.com
Great video!
Thanks a lot!! Stay tuned, more updates are coming :)
The underfloor improvements are insane! Just wondering: how do the Cd/Cl coefficients compare to a standard F1 car?
Willem Toet once told me they were at a L/D ratio of around 3, but this was around 4 years ago.
@@AirShaper The W13 2022 Merc is around 4.5
@@TheGamingHungary Seriously! That's insane, I didn't know it was that much... do you have any source for this, so we can dig deeper? Thanks!
very cool, lets go further!!!
I hope you liked the consecutive videos as well!
By the way, we're now running a MotoGP challenge:
airshaper.com/motogp-aerodynamics-challenge
Is this model also accounting for vehicle dynamics (compression of suspension at different speeds, sensitivity to pitch if you hit a curb etc.) ?
Really interesting to watch you guys work
No, we only analyse a static setup in this case. Very valid point!
@@AirShaper in that case, be careful around modeling/creating aero devices that are very "sharp"/sensitive to minor changes in flow direction/pressure gradients.
This was part of the reason blamed for Senas death: the car was designed for active aero, had it stripped out and was therefore far too sensitive for a human to control.
It's why current setups aren't "optimal", but rather have a very large range of near optimal roll, pitch and yaw angles, as well as speeds
@@camofelix Very true - a bit lower performance but lower sensitivity to settings can inspire more confidence & create a safer car. Building a full aero map should indicate these sentivities.
@Williams F1 Team, please take a look at this.
Would be awesome to have Williams involved!
@@AirShaper I think they'd benefit more from having you instead of their current engineers no cap 💀
@@petarpartaloski8889 We did all of this based on community input (see www.reddit.com/r/airshaper/) so credit to all of them!
I’m doing a project for university and I can’t find anywhere informations about Aerofoil and Angle of attack used on F1 cars, it’s urgent for me to find these info. Any help?
random question -- has AI/ML been used to find the optimal design of any car shape of interest?
We did apply our "adjoint shape optimization technique to optmize parts of the wing, the strakes, ...:
ruclips.net/video/WpFNwfHwFn4/видео.html
ruclips.net/video/cZAhPQFINZ8/видео.html
Engineers of F1 teams: 👀✍
We do want. If there's one thing we do is want.
Bro say 'um' every 2 microseconds
lol was your jaw wired shut?