PDA

View Full Version : Rules of thumb in Wing design



awhittle
01-25-2003, 09:25 PM
Are there Rules of thumb in Wing design that teams are willing to share. Things like end plate design, flap gaps, that sort of thing...

I have always heard that if it is shaped like a banana, it will make downforce.

Close to the ground is good but what is too close. Does ride height change get to be a big deal at some magic dimension?

The bottom of the flaps should follow a smooth curve starting from the main element.

The smoothness of the bottom makes a huge difference. The top surface is seldom a problem. Mount from the top if possible.

Are there rules about slot width?

What about defuser and undertray design?

Most everything in books is based on airplane technology.

Andy

Mr2fastna
01-26-2003, 05:35 PM
Race Car Aerodynamics by Joseph Kats will point you in the right places

jack
01-30-2003, 03:34 PM
i have read "race car aerodynamics" it was a great book, but about all i learned was that downforce for a car going this slow is not worth the effort. if you want to pursue aerodynamics downforce for your fsae car for the sake of learning, that great--but i honestly dont think that downforce (unless power ground effects were allowed) will ever give a team an edge that would make the R/D worth it, or for that matter an edge over the compitition.

JACK@WWU

Dominic Venieri
01-30-2003, 03:51 PM
take a look at "Competition Car Downforce" by Simon McBeath. It's a good practical guide to racecar wing design.

www.formularpi.com (http://www.formularpi.com)

awhittle
01-30-2003, 07:19 PM
470 lb car with a 160 lb driver = 630 lb at the line.

add a 20 lb front wing that makes 40 lb downforce

add a 30 lb rear wing that makes say 85 lb downforce

add a creative floor that doubles as a tunnel making 100 lbs downforce, added weight say 20 lbs of carbon and cables. Most of you are building wide chassises anyway.

added mass 70 lbs or 700lbs at the line. 700mass+225 lbs downforce =925 lbs force

925 lbs force * 1.7 cof = 1572.5 lbs

1572.5 lbs / 700 lbs mass = 2.25 gees at full speed. A 32% increase. Plenty to be a big deal on the enduro section. The other thing is that you can tune in stability at high speed and oversteer at low speed.

Hope this helps

AW

ps loaded qeustion. How many of the FAST autocross cars have aero downforce... the answer Every One http://fsae.com/groupee_common/emoticons/icon_smile.gif

[This message was edited by awhittle on January 30, 2003 at 11:31 PM.]

Scott Wordley
01-30-2003, 08:08 PM
Thats right, trying to produce downforce at these ridiculously low speeds is a COMPLETE WASTE of your time. DO NOT even bother experimenting with wings, instead take mine and Jack's word for it, they will only make you slower. We only perservere with our wings because we like the high performance look they give our car. It took some work, I can tell you. At first we weren't sure which way the wings would look best so we tried them upside down, you know, just to stand out from the crowd. I put one of the photos in the signature if you don't believe me.... hahahaha

Nice post Andy, as you can probably tell, I just can't believe that people can still ignore the simple facts that you have just illustrated for the umpteenth time. If only every racing formula was like this.... hmmmm

Back to the thread:

The Competition Downforce book is an excellent summary of all the basic considerations for wings and underbodies. We designed our whole car pretty much on the guide given in that book, and can almost quote it figure for figure.

Our basic process (outlined in the book) was:

-Start by figuring out how much top speed you can sacrifice to rear wing drag (Front wing doesn't add too much). In FSAE you'll find this is heaps.

-Nominate a wing area (abap) and determine the drag coefficient.

-Relate this to a lift coeff via generic profiles

-Relate this to a generic profile

-Estimate centre of pressure and determine your cantilever about the CG.

- Determine what front wing you need to balance this moment, considering ground effect, increased cantilever, the nose in the way etc.

As for rules of thumb there are millions

- Endplate should be BIG specially on the sensitive LP underside.

- Slot gaps should always be convergent

- Leading edge slats for higher angles of attack without stalling

- Gurney height around 2-5% of chord, for a very cheap increase in DF

- Ground effect can as much as double the DF produced by a wing but watch pitch sensitivity

- Aspect ratio effects can be your best friend or worst enemy, in FSAE it tends towards the later. Realise that after a certain point (Mike, the cornell wing springs to mind) increasing the chord may actually decrease the downforce. FSAE cars would look and perform soooo much better without that shite outside wheel width rule.

- Always minimise bolts, mounts and surface imperfection on the wing undersides, you can lose a lot of efficiency. High side don't matter too much for high lift stuff.

- Keep an eye on your cooling airflows, front wings can be too effective at times.

- On diffuser tunnels look closely at your radii or lack there of, you want to encourage separation in some places and delay it in others. Otherwise skirt gap is all important, less the better.

Thats all I can remember off the top of my head right now.

Please ignore all of it, and remember that we only did it to look cool.
Scott

Regards,

http://www-personal.monash.edu.au/~fsae/dudsig.jpg
Scott Wordley & Roan Lyddy Meaney
Monash FSAE Wingmen
http://www-personal.monash.edu.au/~fsae

awhittle
01-30-2003, 08:54 PM
Remember that the aero drag is a vector pointing down and back. The sum of all these vectors must have a total moment of zero about the point ON THE GROUND below the cg to get constant handling. Move the vector sum back to get high speed stability. Make all this real adjustable. When I added a front defuser to my car, the top wing moved back about 10 inches. Belive me, this stuff works.

As far as the books go... some defusers have smooth gracefull curves, others have sharp corners. Whats up with that?

Ever think about running the exhaust thru the wing to energise the slots? What about unsprung skirts and spring tunnels with some sort of rubber connection. What about these vortisy generators to create air walls? As learning engineers, you will learn that lots of other people have some really cool ideas, Sharing these ideas is the key to learning fast. Learning fast is what will make it possible for you to be gainfully employed some day. Just a little food for thought.

AW

awhittle
01-30-2003, 09:37 PM
I have an areo question.

If you hang yarn wind tuffs in the tunnel but near the trailing edge. What direction will they hang at speed. Is the result different with a wing above and behind the trailing edge of the tunnel. I never done it on my rear defuser. I have no idea what the correct answer is. How can a person without a wind tunnel test these things? The people that know don't talk much.

AW

Al
01-31-2003, 12:26 AM
Hey Andy,


Exhaust blowing of diffusers and wings is a realistic way of developing more downforce at full throttle. F1 did it for years. The problem is the downforce changes with throttle opening. THis is not good for predictable handling, imagine having run wide in a fast corner, coming off the throttle and just making the situation worse. Whether this effect would be particularly large in this formula is a different matter, but hence probably not worth it. (you would also have to worry about melting your wings too!!)

As for underbody skirts, well the rules just about cover that one. No part of the car is allowed to touch the ground while on track. This rule is not a token one either, we at Monash nearly got black flagged for some slight grounding under heavy braking last year. Therefore, unless you can design an unsprung system that you can be sure of maintaining bugger all ground clearance under all operating conditions then you are forced to run I would guess at least 15-20mm clearance. This has a huge effect on downforce. However, there is a sneaky way of partially achieving tunnel sealing and you hit it on the head. By generating a vortex flow along the diffuser sides you can effectively provide a barrier to flow entering the tunnels. Easy to rig something up to do this but tough to visualise the actual result in motion without a rolling road tunnel. (NOte: Ferrari rigged up their bargeboards to do something along these lines in 2002).

As for flow directions in diffusers, Monash have a great wind tunnel for testing wings, but unfortunately it is not rolling road. SO even if we did try testing this I wouldn't trust the result. Maybe the answer is rigging a camera to your car to video some tufts during actual driving.

As for radii or no radii on diffusers, my understanding is that sharp corners will cause signifcant flow separation and hence ruin efficiency. But I have also seen some pretty sharp corners on F1 diffusers. Have you read any of Giorgio Piola's books on F1 cars? HIs pictures on aero stuff are real good, and since F1 knows more about ground effect aero than anyone, then maybe there is something in it. More Volume, flowing through perhaps? Or maybe since the driving effect of the rear wing makes flow separation less likely, then general principles can be ignored? It would be good to find out, anyone happen to know Adrian Newey??

jack
01-31-2003, 03:20 PM
scott & roan,

Fi cars go around turns at 100+ miles an hour. FSAE cars go around turns at -30. because downforce is increased exponentially with speed, FI downforce does so much more than FSAE downforce, you really cant compair the two. FSAE cars really only go fast when they are pointed in strait lines--when down force really isnt that important. i didnt mean to come off sounding like it isnt even worth the trouble, i did say, if you want to learn something, its a good idea. i have done the research my self (thinking in the begining that downforce would help out), and i found out that a car that is able to finish the endurance race will do better than one that is not, yet produces about 100lbs of force. I guess what im trying to say is, first make sure you have a well sorted out car, that is light, and that will run, before you put all your time and effort into DF. Another note about that--Cornell, Georgia Tech and Wollongong all have seemingly well sorted out cars that did pretty well at compitition. Fast, reliable cars make the difference--not wings (at least in at this point in the evolution of FSAE). the one school you could use to agrue my point with is UTA, and we will see if downforce does make the diffenernce for them on their seemingly well built car. If you do decide to undertake a DF progect, and you dont have access to a wind tunell, all your wing will do is look cool http://fsae.com/groupee_common/emoticons/icon_wink.gif

JACK@WWU

Al
01-31-2003, 04:01 PM
G'day Jack,

You are certainly correct in saying that a well sorted car will beat any unsorted car with wings. However, if your car is sorted and you add wings then you have an unbeatable combination in my view. I am expecting UTA to prove that this year in the US.

I don't agree with the requirement for a wind tunnel, it certainly helps but there is enough information out there to do it properly without. I am guessing that not many of the A-Mod guys have a tunnel to use and they do just fine.

As for F1 trends not being applicable I also don't agree. Sure, the flow regime will be different but the principles are the same, ie keeping flow from laterally entering the underbody, etc. The methods of achieving these things are also fundamentally the same.

J. Cheng
01-31-2003, 10:22 PM
[QUOTE]Originally posted by awhittle:
add a 20 lb front wing that makes 40 lb downforce

add a 30 lb rear wing that makes say 85 lb downforce
QUOTE]
A well built composite wing is way lighter than that.

Joe

Dominic Venieri
01-31-2003, 10:27 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by J. Cheng:

A well built composite wing is way lighter than that.

Joe<HR></BLOCKQUOTE>

Very true - our entire package with mounting was well under 15 lbs (just wings, no undertray). And we plan to take weight off that number this year.

www.formularpi.com (http://www.formularpi.com)

awhittle
02-01-2003, 05:28 AM
Joe

I was just tring to show how guickly the numbers can add up to a significant improvment. You don't need 100 mph turns and a 600 hp car. http://fsae.com/groupee_common/emoticons/icon_smile.gif Lighter hardware only makes the math better. As in all of racing, the devel is in the details, like "how do we mount it with out mounting thru the lower surface?" Or "how do we mount the end plates and still build this thing?" We never said this was going to be easy.

Andy

Rago
03-03-2013, 01:52 PM
I have planned a wind tunnel experiment for the scale down model of 10 % of the wings as the capacity of wind tunnel is limited which we have at our university.Does it help anyway to get a glimpse of real situation? I'm using STAR CCM+ for numerical simulation, also have alternative of ANSYS Fluent. which one is better out of these?

JulianH
03-03-2013, 01:59 PM
Well, wind tunnel testing is of course fun, but if you really can see anything real is questionable.

A 1:10 FSAE model has about 1.5cm chord-size wings and a very very small ride-height (0.3cm or so). If you have a good scale wind tunnel with rolling floor and boundary-layer control, you might get good results (although I think the manufacturing of these very small parts could be a problem...).


About CFD.. I don't think there is a "better" program. If you are good with Star CCM, use it. I know that a big bunch of teams use this program with good results.

We use ANSYS CFX and are also happy with it.

Murphy_49
03-11-2013, 12:21 AM
Rago,
I use Star CCM+ but that's just my personal preferance. I used Fluent my first year to simulate different wing profiles (2D Simulations). The past two years though I have used Star CCM+ for full car 3D models. From my experience Star CCM+ seems to have a much more user friendly interface with more visualizations to help find areas that need improvement. With that being said, those values and visualizations are pointless unless validated by combinations of track data, wind tunnel data, and other means of testing.

Out of curiosity, what models and sizes (RAM, # of Cores, Etc.) of computers do teams out there use for running CFD (Star CCM+ in particular) and how big are your end result files?