Make a functional wing with free plywood?

Matt said:

what effect do the endplates have on the wing? is there a design rule for them?

so, that old explanation that "the top of the foil is longer, so the molecule that goes over the top has to travel faster to meet the one from the bottom at the trailing edge"... is not true.  They don't meet.

it gets complicated from there... and there's like an hour long lecture video where an aerodynamicist goes through the various other explanations for "what's happening"... feel free to look it up, and decide for yourself.

The shape of the foil, and the angle means that the flow over the top (in traditional wings) speeds up.  This speed up causes lower pressure.  Things like to move from high pressure to low pressure.

All the analysis that we've been doing in the thread with xfoil and javafoil utilizes a simplification of... essentially... pulling the section of the foil from the middle of the wing.  What happens out at the wing tip, is that the suction on the top of the wing pulls the high pressure air from the bottom of the wing up onto the upper surface from around the tip.  Because of speed... it arrives on the top of the wing on the "back side" of the wing and creates a "spiral" or "tip vortex"... which is where you see those pretty pictures of a crop duster flying through a red cloud and creating a swirl.  This tends to reduce the amount of lift the "real life / 3D" wing produces, when compared to the "ideal" 2D results.

The shorter the wing is, in relation to it's area, the stronger/more the flow is pulled over onto the top.  We call this relationship the "AspectRatio" ( AR ), and calculate it by AR = ( b^2 ) / S   .... where: b = span, and S = wing area.
It's done this way, because airplane wings are frequently tapered... and using the area makes it easier to account for that.

Side note... this doesn't just happen with wings.  If you have a miata with a hard top, and leave it dirty, and drive it through the rain... you can see the same thing form from the flow coming up over the windsheild... with a path of dirt coming off the windowframe and tucking ~6inches in to the middle of the car.  Other cars probably do it too... but I know from experience with the miata.

so... how do we stop the flow from coming up from the bottom of the wing over the tip and onto the back edge of the top of the wing and kill our ideal lift?

1) make the wing longer.
2) convince it not to, by putting something in the way.
3) put a big-donkey propeller at the tip of the wing and spin it counter to the direction of this "downwash"*

*there's some debate about how effective option 3 is.  For fun see: Zimmerman Flying Flapjack.

cars are usually pretty limited, width (or span) -wise.  So, option 2 is the predominate option.

General rules of thumb?  Depends on how crazy you want to get.

stafford1500 said:

In reply to NOT A TA :

I am pretty sure I sent you some thoughts on this floor a while back, but I may be mistaken.

We talked about it during your busy work season before your racing season was about to start and you didn't have any free time to really study it.  Then I got busy with other projects and forgot to remind you. I didn't get  a message if you sent one. I'll sit back down and follow this thread while anxiously awaiting a separate splitter/diffuser thread..

You can see the tufts in this pic indicating what sleepyhead was referring to with his Miata roof example. Even with the side window open you'll notice the yarn tufts at the upper window edge under the sun reflection angling up. The effect is more pronounced with the windows up but I don't have a still pic with windows rolled up.

[URL=http://s240.photobucket.com/user/NOTATA/media/AERO/20160822_182556_zpsxk7owkl5.jpg.html][/URL]

In reply to sleepyhead :

I met the designer of the spiroid winglet, right there where the picture was taken, Oshkosh. 

Just mocked up and no end plates yet

it looks like there's some "twist" in it, left side vs. right side... but that might be the camera angle

Watching intently for.... reasons. 

If the endplates are wood you could router some grooves in the end plates to hold the curve.

sleepyhead said:

it looks like there's some "twist" in it, left side vs. right side... but that might be the camera angle

Maybe, it didn't really look like it in person. But I'll flip it over and check on flat concrete tomorrow. (If it is twisted, that wouldn't really be a shock since the board was twisted before I started!!)

NOT A TA said:

If the endplates are wood you could router some grooves in the end plates to hold the curve.

I was actually thinking I would just jigsaw all the way through, but yes the plan is for the end plates to help hold the shape too.

Robbie said:

sleepyhead said:

it looks like there's some "twist" in it, left side vs. right side... but that might be the camera angle

Maybe, it didn't really look like it in person. But I'll flip it over and check on flat concrete tomorrow. (If it is twisted, that wouldn't really be a shock since the board was twisted before I started!!)

So the twist is real, but only about 1 inch or less corner to corner. And I can easily "pull" it out by pushing down on both corners, so I think I'll be able to straighten it out a bit when mounting. 

I am able to stand on the top of the camber with the wing laying on the ground, so the glue is certainly strong (but I expect a wood-glued joint to be quite strong). 

And final measurement of the camber is a little less than 2.5 inches which does put me right at my original design cambered profile of about 14%. I may make another wing with more aggressive camber and test it too. 

In reply to Robbie :

I want to think my eyes are that good... but I really think the camera angle is what's "setting it off"

you might consider buying some 2x6x10' stock to make a four "doubled up beds" to set the wing into during the glue up.  that way it'll be a lot easier to check if it's twisting and correct it.  The beds don't stay on the car, so they're $challenge free... and you can make toys for the kids (or somebody's kids) out of them when you're done.

Now that I think about it, the correct way to build this would be to flip the wing over and let it sit on the flat ground while drying, that would keep both front and rear edges parallel. I could even "weight" the corners or something to hold it as it dries. 

I didn't get a chance to do any catching up on this thread over the weekend due to family stuff, but it looks like it is progressing nicely...

As far as end plate and 3D Vs 2D, there are some corrections that can be made to APPROXIMATE the effect of an actual wing compared to a 2D section. As mentioned the Aspect Ratio is a big one and the end plate is the other major item. One other that has been included but sort of ignored so far is the bi-plane effect. Using the low pressure side of a trunk mounted wing will also pull up on the trunk and that is impacted by the separation between the two. It can be modeled, but not with any simple /free/2D CFD that I know of. A more complete 3D model is possible, but much more complicated (and time consuming).

Creative end plate design could benefit a wing shape that twists from center to outboard using the routed/cutout method Robbie discussed. It could also be used to help keep the bottom surface flow attached in the corner of the wing/endplate where the boundary layer/pressure gradient is fighting to stall the trailing edge.

In reply to Robbie :

The easiest way to build a wing is use a series of wing shapes made from plywood  glued together with plywood spacers  to connect everything.  

Then buy a yard or two of Dacron ( about $5.00 a yard) and sew a cover  to fit over the whole thing.  Slide it over the wing and use a hair dryer to shrink it tight.  

Just watch someone cover a wing on U tube to get a clear idea.  Since you aren’t going flying with this wing you won’t need to use Sitka Spruce etc etc etc. 

keeping these imagegrabs for later 

Robbie:

More angle on the third element...

Try to reduced the pressure at the gap between the first and second element by manipulating the third element.

Also, try to increase the cell counts at the top right from 100 and 50 to 200 and 100 to see if the resolution gets better around the leading edges.

Robbie said:

Now that I think about it, the correct way to build this would be to flip the wing over and let it sit on the flat ground while drying, that would keep both front and rear edges parallel. I could even "weight" the corners or something to hold it as it dries. 

Keep the outside of the curve to the ground as it's typically higher humidity. Putting the concave side to the ground/concrete will make the board want to straighten itself.

stafford1500 said:

Robbie:

More angle on the third element...

Try to reduced the pressure at the gap between the first and second element by manipulating the third element.

Also, try to increase the cell counts at the top right from 100 and 50 to 200 and 100 to see if the resolution gets better around the leading edges.

holy moly, you aren't joking, you can really crank the angle on the 2nd and 3rd element! This does seem to make the wing quite 'peaky' however in terms of change in AoA. 

and just for comparison, here is a similar overall shape made out of a single element instead of 3. Drag is almost the same but downforce is TRIPLE on the multi-element wing. 

Robbie said:

 

holy moly, you aren't joking, you can really crank the angle on the 2nd and 3rd element! This does seem to make the wing quite 'peaky' however in terms of change in AoA. 

 

Yes it is more peaky, but how much do you expect the incoming air to change as it flows around the car. The general flow will be quite similar, you just have to adjust the entire assembly thru a small AOA sweep to find the the maximum  (or twist the wing to get even more out of it).

stafford1500 said:

Yes it is more peaky, but how much do you expect the incoming air to change as it flows around the car. The general flow will be quite similar, you just have to adjust the entire assembly thru a small AOA sweep to find the the maximum  (or twist the wing to get even more out of it).

This is actually something I've wondered about... especially considering the low Velocity / low Reynolds number range we're discussing with autocross.  Depending on the rear glass/hatch/trunk shape, I could see the flow being at a lower angle at lower speed, since it's probably not staying attached.  although, adding the low pressure from the wing might resolve that.

of course, we'd have to rely on the streamlines to tell us that, right?  which would be peeling them out of javafoil and calculating the inflow angle by hand/spreadsheet?

sleepyhead said:

stafford1500 said:

Yes it is more peaky, but how much do you expect the incoming air to change as it flows around the car. The general flow will be quite similar, you just have to adjust the entire assembly thru a small AOA sweep to find the the maximum  (or twist the wing to get even more out of it).

This is actually something I've wondered about... especially considering the low Velocity / low Reynolds number range we're discussing with autocross.  Depending on the rear glass/hatch/trunk shape, I could see the flow being at a lower angle at lower speed, since it's probably not staying attached.  although, adding the low pressure from the wing might resolve that.

of course, we'd have to rely on the streamlines to tell us that, right?  which would be peeling them out of javafoil and calculating the inflow angle by hand/spreadsheet?

On the topic of incoming airflow; last race of the year it was raining, so there are directional dirt lines on my wing that tell me the airflow was different over the wing. The outside sections on the top of the wing show the air being  directed to the center if the car, center of the wing shows straight back - but on the underside it’s all straight back,  is this the endplates working or not working?

it's interesting that the flow is straight on the bottom, and angled on the top... although, I wouldn't read too far into that.  the main thing I think I see from the top photo is that you have laminar separation bubble happening on the upper surface from approximately %10 to 70% chord.  that bubble is happening the whole way down the wing... although there might be a slight influence of cross-flow from the body at the edges vs the center... I don't think it's too much to be concerned about.

I'd have to go back and look at the mshd paper and results... but off-hand I wonder if you're running at a low enough angle of attack that the mshd airfoil is "overkill" for the downforce you need to balance your front splitter.  I know they talk about the likelihood of one of those bubbles forming in the report, and the relative lack of impact on the airfoils' performance.  but, my first instinct is to guess that you don't need as much lift as that thing is generating... which isn't too surprising.  most cars are front-downforce limited, especially compared to the relative ease/efficiency of creating dowforce out of a rear wing.