Aerodynamic design AND optimization of an f1 rear spoiler

[wesley123]

in case of pressures: red=high, blue=low

streamlines are colored by id, so colors mean nothing

so if i'm seeing it correctly(due to the rear wing colours) the one without the slits has lower pressure then with the slits in the end plates?

[Slimak]

umm, no
in the picture you are talking about, colors of the wing do not matter as they are different just to distinguish models (blue was my basic model, yellow had much more complex geometry)
my investigation in this area was concentrated only on the pair of vortices behind the wing, not the pressures

[WhiteBlue]

Hi Slimak,

welcome from me to the forum as well. It is good to have someone with recent aero experience and possesion of models on board.

May I ask what kind of equipment and SW you use for CFD? Also interesting would be what kind of models you have. Do you have a diffusor in addition to your rear wing?

[bazanaius]

As slimak hinted, you'd need to clarify the aims of your project before starting to look at the complexities of modelling a wing in situ.

Are you actually looking to learn general trends about aerodynamics, possible efficiency solultions and other wing technologies, but simply using the example of an F1 rear wing to make it interesting (your supervisor can probably help with this). If this is the case, there is nothign wrong with examining the wing in isolation - in fact, I'd suggest it's the best place to start. Get the basic understanding of what's happening covered, then look at possible methods for improving the efficiency of your wing - and here comes in your aims from an aero point of view. Maybe max L/D for a given L? I'd start by not following the rules for F1 - start with a single plane, then move onto looking at the different ways you can acheive your aims. Maybe 94 planes is the best solution - why do F1 teams not run 94? regulations. What can you do to get the same results in a different way? Maybe end plates give you an improvement - make them flat,and see what they do - how could you improve them? Maybe try to equalise pressure to decrease form drag/Vortices? How would you do this? Maybe then look at constraining your wing in space, as the rules say - how might you get an efficient increase in lift? Gurneys maybe? Then maybe look at what YOU would do if you had to place the wing behind the car body.

This is the kind of process I'd follow - start with a wide view, a simple solution and build it up from there - use your ideas, with current solutions for guidance, and you'll learn a lot more than just copying current solutions (which are very specific and complex) and modelling them without an understanding of what's going on and why they are the shape they are.

[kilcoo316]

i'm studying aeronautical and mechanical engineering and i have to work on the thread topic. I have to create an formula 1 rear spoiler (a very simple one) with pro engineer, analyze it with CFD (gambit/fluent) and then i have to think about an optimization from this spoiler.

I'd simplify that... alot. If I had to do that task, I would follow this basic order:

1. Validate your code
2. Perform simulations on a baseline wing
3. Add vortex generators at 15% chord (or where-ever you feel is suitable) and evaluate the difference in seperation point and downforce/drag. Bear in mind the inaccuracies from your validation comparisons when doing so.


If you are doing 3D simulations within an undergraduate academic environment, don't overstretch yourself in terms of CPU time - one basic fundamental of CFD work... you will never have enough CPU time to do everything you want.

[Zweeper]

okay, thanks a lot for all the tips so far. you already helped me a lot with this task! i will met my tutor on wednesday and talk with him about all the new idea is got now from you and then i will give you feedback about the new situation

I think my biggest problem will become the CFD analysis because i think i will do the design with pro engineer, import and mesh it with gambit and do the analysis with fluent then.

[kilcoo316]

import and mesh it with gambit

It is probably better to actually construct the geometry within gambit itself.


Leave yourself alot of time for importing/making and meshing the geometry. Think of your worst estimate (for everything going wrong), then double it.

[PlatinumZealot]

i'm studying aeronautical and mechanical engineering and i have to work on the thread topic. I have to create an formula 1 rear spoiler (a very simple one) with pro engineer, analyze it with CFD (gambit/fluent) and then i have to think about an optimization from this spoiler.

I'd simplify that... alot. If I had to do that task, I would follow this basic order:

1. Validate your code
2. Perform simulations on a baseline wing
3. Add vortex generators at 15% chord (or where-ever you feel is suitable) and evaluate the difference in seperation point and downforce/drag. Bear in mind the inaccuracies from your validation comparisons when doing so.


If you are doing 3D simulations within an undergraduate academic environment, don't overstretch yourself in terms of CPU time - one basic fundamental of CFD work... you will never have enough CPU time to do everything you want.

I like this.. Keeping it simple and to ONE particular aspect, the vortex generators. I remember my project I concentrated on too many things, I still got an A but I overworked my self unnecessarily.

[Slimak]

Hi Slimak,

welcome from me to the forum as well. It is good to have someone with recent aero experience and possesion of models on board.

May I ask what kind of equipment and SW you use for CFD? Also interesting would be what kind of models you have. Do you have a diffusor in addition to your rear wing?

english is not my 1st language so by 'SW' you mean software, right?
i use gambit/tgrid/fluent for CFD, model was created in NX5
computer was the property of faculty and i dont know what exactly it was- just a good pc, 1kk cells were computed in less than 5 hours (hard to say since i went out)

the models i have are made by me in NX5. they are simplified a lot to enable successfull meshing and thus not always aerodynamically correct (but i fulfilled dimensional regulations), and yes- they have diffusers

when i find more time for uploading images i will try to show them, but probably in another thread, since this one belongs to Zweeper and his wing

[WhiteBlue]

We had some discussions here about the effect of the double deck diffusor. Basically a flat diffusor and a high rear wing should create a separation between the dirty air from the rear wing and the relatively clean air of a flat diffusor. This should make for a less turbulent wake behind the center of the car.

If one adds another 150% of diffusor height - like DDDs basically do - the wake should become much more turbulent. Would your models be able to show the different wake structure from the side?

[Zweeper]

but you think that vortex generators ON the wing could make sense?

because i thought the flow seperation is to neglect on the rear wings. doesnt make a huge effect!?


edit: okay i didnt saw your post in front of this one whiteblue.
one question to your aboves post. When we talk about lift from the wing, we talk about the downfroce it produces because its reverse mounted on the rear wing from the car?

and what do you mean with 94 planes?

[Mystery Steve]

I'm not sure if this is actual CFD data, or just a pictorial, but it does illustrate how the different elements act together as a whole:

[youtube]http://www.youtube.com/watch?v=AXc151KS ... re=related[/youtube]

What he means by 94 planes is the number of different "elements." The video above has a 3 element upper wing and a single element lower wing. The advantage of multiple elements is that when they are positioned like they are in the video, they act together as a single wing at a very high angle of a attack. The problem with running a single element wing at high angle of attack is it will stall out due to flow separation on the lower side. Using multiple elements allows some of the air on the top side to slip to the bottom side and "fill in" this separation. Also, the upwash from the lower wing helps to reduce the separation as well.

[Zweeper]

Very good, thanks for the explanation!

my first step now is to research on the influence of gurney flaps on a wing i will use. maybe i will also make a general research about which flap is the best for my wing.

can you suggest some typical sued NACA aerofoils for formula 1 wings? i read that every team designs more or less their own wings, but they are all based on NACA wings and this wont be part of my work. So would be nice to determine at first the drag/lift ratio of some common used wings and then determine the influence of an gurney flap.


Wow, i start to love this forum in general

[Slimak]

If one adds another 150% of diffusor height - like DDDs basically do - the wake should become much more turbulent. Would your models be able to show the different wake structure from the side?

sorry for the delay, but i have hardware problems

unfortunately not, my ddd was so simplified that the final solution proved to be designed wrong, there even occurs separation in the upper channel
i tried to compare them by tracking pathlines set from wing and diffuser, and results are quite similar

[flynfrog]

Very good, thanks for the explanation!

my first step now is to research on the influence of gurney flaps on a wing i will use. maybe i will also make a general research about which flap is the best for my wing.

can you suggest some typical sued NACA aerofoils for formula 1 wings? i read that every team designs more or less their own wings, but they are all based on NACA wings and this wont be part of my work. So would be nice to determine at first the drag/lift ratio of some common used wings and then determine the influence of an gurney flap.


Wow, i start to love this forum in general

start off with
NACA 63(3)-618
wortman FX 63-137