Flap development in F1

Here are our CFD links and discussions about aerodynamics, suspension, driver safety and tyres. Please stick to F1 on this forum.
Post Reply
timbo
111
Joined: 22 Oct 2007, 10:14

Flap development in F1

Post

Today's generation of F1 cars has ridiculous flap detail. Deflectors in front of the sidepods of the Mercedes are surreal, the detail of the tiny winglets at the sides of the diffuser on the Ferrari is also crazy.
How do you think they design them? Is it even humanly possible to tweak all those minuscule "feathers"?
Probably some sort of automated development is employed. Any insight on what exactly is used? Genetic algorithms? Machine learning?

hardingfv32
32
Joined: 03 Apr 2011, 19:42

Re: Flap development in F1

Post

Could their adjustment be automated in the wind tunnel?

Brian

Edax
47
Joined: 08 Apr 2014, 22:47

Re: Flap development in F1

Post

timbo wrote:
26 May 2017, 18:59
Today's generation of F1 cars has ridiculous flap detail. Deflectors in front of the sidepods of the Mercedes are surreal, the detail of the tiny winglets at the sides of the diffuser on the Ferrari is also crazy.
How do you think they design them? Is it even humanly possible to tweak all those minuscule "feathers"?
Probably some sort of automated development is employed. Any insight on what exactly is used? Genetic algorithms? Machine learning?
For mechanical and thermal structures topology optimisation software is getting closer towards practical implementation (At the moment it is more providing design inspiration than providing real designs). But I doubt it is mature enough to work for aerodynamics.

I am no aerodynamicist but If I look at the cars I see the same design solutions appearing on different part of the cars. Like the serrations and the cascades. My guess would be that they develop a number of geometric flow solutions and then copy-paste them on parts of the car which have a similar function.

Like we use isogrid designs, fixed geometry adapted to location and load.

Cold Fussion
93
Joined: 19 Dec 2010, 04:51

Re: Flap development in F1

Post

Is topology optimisation even a realistic option for aerodynamics given the low computing power and time requirements imposed by the FIA for CFD?

timbo
111
Joined: 22 Oct 2007, 10:14

Re: Flap development in F1

Post

Edax wrote:
27 May 2017, 04:26
My guess would be that they develop a number of geometric flow solutions and then copy-paste them on parts of the car which have a similar function.

Like we use isogrid designs, fixed geometry adapted to location and load.
That's an interesting idea!

timbo
111
Joined: 22 Oct 2007, 10:14

Re: Flap development in F1

Post

Cold Fussion wrote:
27 May 2017, 07:11
Is topology optimisation even a realistic option for aerodynamics given the low computing power and time requirements imposed by the FIA for CFD?
Dunno, depends on how the time is counted. Maybe the process of structure generation would not be included in the CFD time allocation?

Cold Fussion
93
Joined: 19 Dec 2010, 04:51

Re: Flap development in F1

Post

timbo wrote:
27 May 2017, 08:41
Cold Fussion wrote:
27 May 2017, 07:11
Is topology optimisation even a realistic option for aerodynamics given the low computing power and time requirements imposed by the FIA for CFD?
Dunno, depends on how the time is counted. Maybe the process of structure generation would not be included in the CFD time allocation?
My understanding of the implementation of topology optimisation is that there are lots of small incremental changes as it converges and it would be the testing of all those small changes that wouldn't be feasible under the FIA regulations. I could be totally wrong though and perhaps topology optimisation is a lot more advanced than I realise.
Last edited by Cold Fussion on 27 May 2017, 09:19, edited 1 time in total.

timbo
111
Joined: 22 Oct 2007, 10:14

Re: Flap development in F1

Post

Cold Fussion wrote:
27 May 2017, 08:51
My understanding of the implementation of topology optimisation is that there a lot of small incremental changes as it converges and it would be the testing of all those small changes that wouldn't be feasible under the FIA regulations. I could be totally wrong though and perhaps topology optimisation is a lot more advanced than I realise.
AFAIK you are right, but maybe they can focus on a specific part and therefore CFD test would be faster.

User avatar
Vanja #66
1350
Joined: 19 Mar 2012, 16:38
Contact:

Re: Flap development in F1

Post

Aero departments are largest in F1 teams, and they are split in 4 groups - CFD developers (programmers), CFD users (aerodynamicists), CAD designers (tasked with making perfect surface models for various uses) and wind tunnel operators/model makers. They all work closely together and teams use huge staff number in aero departments to overcome FIA limitations on teraflops - or at least this is my understanding... Those flaps and winglets have in general the same purpose and the same concept is behind all of them - flow control without too much energy dissipation. Wings with flaps and slots are good for this, so that's why you can see airfoils everywhere.

With a large number of CAD designers/modelers you can make a large number of different designs for fast-paced CFD development in a relatively short amount of time. Race car aero - and especially new-era F1 aero - is still very intricate and requires human touch to develop. In my view, we are still far from topology-like computer optimisation of aerodynamic (race) car design...
And they call it a stall. A STALL!

#Aerogimli
#DwarvesAreNaturalSprinters
#BlessYouLaddie

User avatar
DiogoBrand
73
Joined: 14 May 2015, 19:02
Location: Brazil

Re: Flap development in F1

Post

Vanja #66 wrote:
27 May 2017, 10:37
Aero departments are largest in F1 teams, and they are split in 4 groups - CFD developers (programmers), CFD users (aerodynamicists), CAD designers (tasked with making perfect surface models for various uses) and wind tunnel operators/model makers. They all work closely together and teams use huge staff number in aero departments to overcome FIA limitations on teraflops - or at least this is my understanding... Those flaps and winglets have in general the same purpose and the same concept is behind all of them - flow control without too much energy dissipation. Wings with flaps and slots are good for this, so that's why you can see airfoils everywhere.

With a large number of CAD designers/modelers you can make a large number of different designs for fast-paced CFD development in a relatively short amount of time. Race car aero - and especially new-era F1 aero - is still very intricate and requires human touch to develop. In my view, we are still far from topology-like computer optimisation of aerodynamic (race) car design...
That's probably accurate.
Newey, for example, is always praised for his hand drawings. Unfortunately the other aerodynamicists aren't as high profile so that we could have people showing us what they do.

In my (very uninformed) opinion, F1 designs are getting more and more restricted, so they know they need a front wing, turning vanes, cascades, tea tray, barge boards and so on, and unless somebody has a revolutionary idea (For example Ferrari's new sidepods), then they probably start with a very basic design within the rules, kind of like the RB13 from the start of the season. From then on it's just seeing where you have flow not doing what you would like, and changing/adding surfaces to try and make them do what you want.

That's a very simplified description, and likely to be all wrong, but that's how I'd imagine they doing it.

Post Reply