Average downforce your underbody diffuser generates..

[NjuL]

We are implementing aero for the first time this year and have started with an underbody diffuser. We are currently running CFD on it and only generate less than 40 lbs of downforce. I have searched the forums to see what is an average number but cant find anything

I guess I am just curious about the numbers produced by teams that have had aero for a number of years.

thanks

[Kirk Feldkamp]

40 lbs... on a 400lb car? Sounds like a fair amount to me. I'm no aero expert, but my impression is that diffusers are more often used to either alter the differential between the top and bottom of the bodywork, or to alter the conditions under the rear wing... rather than just by themselves.

More importantly, where is the center of pressure for that device, right? With the new rules for allowable wing locations, my guess is that the lowest hanging fruit are going to be maximizing (and balancing) the torque that the front and rear wings generate around the center of mass.

-Kirk

[Gruntguru]

Originally posted by NjuL:
We are implementing aero for the first time this year and have started with an underbody diffuser. We are currently running CFD on it and only generate less than 40 lbs of downforce. The downforce figure is pretty useless without the associated speed.

[NjuL]

O yea, I forgot to mention that. We are running our tests at 17 m/s

[tgman2]

40lbs at 17m/s is a bit less than expected, remember that the ground is moving relative to the car and that the surface roughness of the track is a factor. How large are your diffusers and what sort of angle are they relative to the ground?

[Lorenzo Pessa]

178 N @ 17 m/s (sorry, I don't like ancient measurement system)

Is it you bodywork only which generate this downforce or is your car?

Splitting the amount of downforce/drag generated by different elements is logical when you are analysing the whole car or different car configurations.

Evaluate if you really need an underbody to generate downforce.

If you need an underbody for other reasons, it is smart to try to get some downforce (just a little bit).
If you want an underbody (without wings) to generate downforce, I'm not sure it works. The weight do you add may be more the downforce you get. It would be great to put CFD data in a car simulator and compare different configurations by timings.

[Scott Wordley]

That sounds like a reasonable start, but given the new extended package space for diffusers you should be able to do a fair bit better.

When you say that you are running CFD on it, what do you mean exactly?

2D or 3D analysis?

Diffuser on its own or with the rest of the car and wheels? Moving ground? Rotating wheels?

Just curious because to design a diffuser properly in CFD it can get quite complicated very quickly. There are plenty of 3D effects and all the upstream wakes from the nose and wheels are very important.

As the previous posters have mentioned keep an eye on your front rear balance. Diffuser only tends to create a big rear bias which can make things somewhat understeer-ey at high speeds.

Even though diffusers can be used all on their own (given the package space to shift the centre of pressure forwards), for a FSAE application its probably easier to run a small front wing to get your balance.

Good luck with it.

Scott

[RollingCamel]

Will you verify your simulation with a scaled model test?

[Lorenzo Pessa]

Talking about FSAE only, if you don't have a proper wind tunnel at your disposal (for free) I think it's quite expensive to get a good simulation.
Quite expensive compared to current CFD.

[Ally_UGR]

We have a windtunnel at Glasgow that has been used for 3/4 scale F1 sims in the past for the old BAR Honda team. We asked about using it and were politely told to scram. We were told it's £2500 per hour of running and that's out of the question for FSAE. Also, getting a scale model with similitude to your end product at the design/build phase is unlikely, and thus any verification and justification for using aerodynamic surfaces can be highly flawed or invalid.

IMHO, you would be better to spend the time on other areas of the car, work on gaining more mechanical grip or losing excess weight. You'll be lucky to hit to speed for more than a few seconds, if at all at a competition. Any extra downforce you might get out of your aero will also carry a penalty on drag as well.

[Scott Wordley]

That wind tunnel sounds nice, shame you cant use it. It sometimes helps if you invite the wind tunnel manager to a drive day and put them in the car... a good way to make friends in high places.

Back on topic...

CFD is quite good and readily accessible these days. You can get reasonably accurate results using a single fast desktop computer with lots of RAM. It will take a little time but you will get there. Wind tunnel testing is nice but not always an option for most teams. On-track aero testing and validation can be done and usually yields good results (but fewer of them, and for more work).

Ally_UGR... there are lots of places where you can spend time and money to make a car faster. There are also lots of places were you can save money, do things simpler or even heavier and build a car more quickly and more robustly.

People will have different opinions about where their very limited resources are best spent. For some its might be a turbo, other a carbon tub, and still others aero devices.

I'm not looking to defend the virtues of aero in FSAE, as god knows I've done that enough over the years. All you need to do is run the numbers, and see how many seconds (points is better) a good aero package can be worth.

[Ally_UGR]

http://www.gla.ac.uk/departmen...erospace/facilities/ (http://www.gla.ac.uk/departments/aerospace/facilities/)

For anyone interested :P

[StephenAero]

200lb-400lb is achievable.

-Stephen
University of Maryland

[Mike Cook]

If it is made of depleted uranium.

[DFT]

Originally posted by StephenAero:
200lb-400lb is achievable.


Did you mean with or without a front wing? Also, are these numbers at 17 m/s and do they include the static downforce from the alleged uranium? I'm just trying to get more context for your statement.

[ZAMR]

Originally posted by StephenAero:
200lb-400lb is achievable.

-Stephen
University of Maryland

At 40mph? I call shenanigans. Don't try to tell a design judge that because they aren't stupid. 200lbs might be achievable at 40 but only if you have a fat rear wing. But then your center of pressure is well behind your CG and you are just adding useless drag.

Kinda like Maryland's '09 car. A small front wing and large rear wing is kinda pointless.

[StephenAero]

I always get hamburgers at our local shenanigans. Best BBQ sauce in the area!

What happens if it's not though?

[Mike Cook]

Hey Farva what's the name of that restaurant you like with all the goofy shit on the walls and the mozzarella sticks?

[Kirk Feldkamp]

I swear to God I'm going to pistol whip the next guy who says "Shenanigans".

[EHog]

Originally posted by Kirk Feldkamp:
I swear to God I'm going to pistol whip the next guy who says "Shenanigans".

Its official, the word shenanigans is hereby banned from the fsae world.

Anyway, who cares about downforce numbers? Any lift stated at any given speed is irrelevant. Why? Because we (SDSM&T) are going to have more than you anyway.

[Mike Cook]

No,

All your downforce are belong to us.

[ZAMR]

Yes but what is really important is downforce to weight ratio. I know SDSM&T is pretty light with the ape, I don't know if Maryland is running with the phazer and crazy diff/transmission idea, but we'll see who comes out on top. And who can hold their powertrain together.

Also,

Shenanigans Shenanigans Shenanigans Shenanigans

[EHog]

Originally posted by ZAMR:
And who can hold their powertrain together.


We wanted to join the parade of good teams failing endurance. You're all just jealous because we failed in style! I hope all the smoke and shenanigans kept everyone entertained! http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

[Lorenzo Pessa]

Aprilia never fails (to give problems).

[Mike Cook]

Originally posted by Lorenzo Pessa:
Aprilia never fails (to give problems).

LOL

Can we just take the apes out behind the shop and put a bullet through them. It's the biggest ponzi scheme the FSAE world has ever seen.

[ZAMR]

On a serious note, has anyone done any analysis on the effects the front wing has on undertray performance? We've been messing around with full body simulations for quite a while now, and pretty consistently, we are finding that the undertray performs quite poorly with a big wing in front of it. I'm talking maybe 15lbs at 35mph, while the front wing makes around 55lbs and the rear makes about 65lbs.

I noticed Illinois had a front wing pretty high off the ground last year, probably so they could feed their undertray. But then you lose all of your ground effect on your front wing, so why go with that approach?

I just don't understand where these high undertray DF numbers are coming from from the winged teams. Without wings I believe a very well designed undertray might make 40lbs at 35mph, but if you put wings on, you can kiss those numbers goodbye.

Anyone agree? Disagree?

[Lorenzo Pessa]

Ok, you confirm my doubts after my first simulations where I noticed very low DF from the undertray (with front and rear wings).

Not all teams with wings have an undertray. I'm wrong?

[EHog]

Our team noticed the same issues. We typically run front and rear wings with an undertray. After learning new fluent techniques this year, we found out that the undertrays we have run and were planning to run haven't done much of anything... And so... we're taking a slightly different approach.


Not all teams with wings have an undertray. I'm wrong?

I don't believe MS&T had an undertray with their wings previously. Correct me if i'm wrong.

We (SDSM&T) had an undertray with wings, but as I stated, it probably didn't do much...

[Lorenzo Pessa]

Maybe some F1 teams notice this kind of phenomena as they design so high nose and front wing with side flaps only.

Not easy to work in this way with the template.

[Will M]

MS&T ran a small under tray for Michigan in 2010 otherwise their aero package looked pretty similar to their 2009 set up.

[cwatson]

...if you put wings on, you can kiss those numbers goodbye.

Not quite.


Maybe some F1 teams notice this kind of phenomena as they design so high nose and front wing with side flaps only.

Rules.


But then you lose all of your ground effect on your front wing, so why go with that approach?

Why is ground effect so great?

[ZAMR]

Originally posted by cwatson:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">...if you put wings on, you can kiss those numbers goodbye.

Not quite.


Maybe some F1 teams notice this kind of phenomena as they design so high nose and front wing with side flaps only.

Rules.


But then you lose all of your ground effect on your front wing, so why go with that approach?

Why is ground effect so great? </div></BLOCKQUOTE>

1. I'm pretty sure air works the same in Italy and South Dakota as it does in Maryland.

2. F1 teams ARE raising their noses to aid their undertray. It is a rule now that your nose can't be ABOVE a certain height, read the rules.

3. Ground effect makes cars win championships. Lotus T78. Ground effect has a huge effect on your front wing's downforce.

[StephenAero]

Originally posted by ZAMR:

1. I'm pretty sure air works the same in Italy and South Dakota as it does in Maryland.

Elevation changes.



2. F1 teams ARE raising their noses to aid their undertray. It is a rule now that your nose can't be ABOVE a certain height, read the rules.

Safety.



3. Ground effect makes cars win championships. Lotus T78. Ground effect has a huge effect on your front wing's downforce.

1979.

[ZAMR]

Elevation does change, you are right. However, the Navier-Stokes equations still hold everywhere air is still a fluid.

Yes, they made the nose height rule for safety because everyone was raising their nose up high in order to extract extra downforce. But then the driver couldn't see the road, ergo the new rule.

The Lotus 80 was meant to be an evolution of the Lotus t78/79 ground effect concept, but it made too much downforce, and the downforce varied too much with ground clearance, so it was too hard for the driver to control, and it won 0 times.

Ground effect is still important today, as we all saw the "flexi" wings from red bull and ferrari in the 2010 season.

[PatClarke]

The Lotus 80 was meant to be an evolution of the Lotus t78/79 ground effect concept, but it made too much downforce, and the downforce varied too much with ground clearance, so it was too hard for the driver to control, and it won 0 times.


I just love the distortions of history :-) Especially when generational boundarys are crossed.

The best thing about the 79 was Mario and Ronnie. The best thing about the 80 was the porpoising.

Pat

[ZAMR]

Originally posted by PatClarke:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The Lotus 80 was meant to be an evolution of the Lotus t78/79 ground effect concept, but it made too much downforce, and the downforce varied too much with ground clearance, so it was too hard for the driver to control, and it won 0 times.


I just love the distortions of history :-) Especially when generational boundarys are crossed.

The best think about the 79 was Mario and Ronnie. The best thing about the 80 was the porpoising.

Pat </div></BLOCKQUOTE>

I'm just regurgitating what I read in books, that's all I can do. I'd be honored to hear about it from someone who was really there, during the golden age of racing. Or that's what I've been told it was http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Lorenzo Pessa]

Originally posted by PatClarke:
The best thing about the 80 was the porpoising.

I don't know if it was the case of Lotus T80.
Some years later someone noticed that lowering too much a front wing at high speed a shock wave was genereted. Was reported that the test driver was not so happy of that.

Not the case of FSAE, happy to work at constant density for now. Ops! I missed the changes in elevation.
If you are planning to run an aero car at Formula SAE Italy you have to evaluate the elevation and the closeness of a farm: there is a huge smell of cow shit that influence drammatically the performance of the undertray.

Interesting discussion around F1 rules: "which came first, the chicken or the egg?"
I have no absolute answers about that.

Sorry, only the first paragraph of this post was serious. I will try to improve.

[ZAMR]

Update:

So we've been running CFD sims around the clock to see the various effects the front wing has on undertray performance. What we are finding is that the wing has varying effects, depending on tray design. However, we are still unable to reach the insane downforce numbers claimed by some.

We have experimented with different windtunnel sizes, cell counts (from 1,000,000 to 10,000,000), boundary layer approximations, etc. Basically, your windtunnel size and cell count have more to do with your undertray performance than your front wing. Using the same geometry we can get almost 20lbs of extra downforce just by changing the mesh. Just goes to show you have to be very careful with your CFD procedure.

Hope this is more helpful than one word responses. If anyone has similar experience feel free to add to this...

[Lorenzo Pessa]

Originally posted by ZAMR:
We have experimented with different windtunnel sizes, cell counts (from 1,000,000 to 10,000,000), boundary layer approximations, etc. Basically, your windtunnel size and cell count have more to do with your undertray performance than your front wing. Using the same geometry we can get almost 20lbs of extra downforce just by changing the mesh. Just goes to show you have to be very careful with your CFD procedure.

Hope this is more helpful than one word responses. If anyone has similar experience feel free to add to this...

Interesting work.

Are you talking about tetrahedral or polyhedral cells?
From sensibility studies from other students, 1 million of poly for a simulation on a whole car (without simmetry plane) are too low.
On my preliminary simulation I worked at 2.5 millions of poly (with StarCCM+). Talking about grid on the flap part of undertray the elements are 20mm long (less than 10 in the refined zones).
At the beginning of March I hope to get more accurate results (now exams!).