LMP360 Aero Design Discussion Part 1: Front Element

so, let's go a "simple" step forward, and use a relatively easy to make foil section for the first element only.  The added thickness of this section (11.7% compared to 2.5%) means the nose has to be raised some in order to better match the minimum height clearance.  So the Leading Edge (LE) point ends up at 17.3%, giving a minimum clearance of 11.0% (i.e. ~2inches).

This is potentially possible, because the minimum clearance location is back from the very edge by 4 or 5inches;  depending on what kind of information we eventually get out of nocones... but also shows some of the iterative nature that this process most likely will require.

Normally a Clark Y will top out around a Clmax of 1.6 at 15degrees.  But proximity to the ground means it's maxing out around 7deg...

recall, that we've got to take that with a 60% grain of salt...

2.305 * 0.6 = 1.383

I'm not sure I've seen a span length estimated between the fenders, nor the first element chord length.  I'm going to go conservative, and guess it's 50inches in span and 18inches in chord (and I'll update if/when nocones gives an updated number).
thus, this splitter (which is 17% smaller than the last one), at 60mph {88 feet per second} would generate:  80#s of downforce

I'll have to dig through my data to get you the CG information and spring rate items.  I'll stiffen the front to keep the splitter off the ground either way.  

Frame Ground clearance is designed to be 2.5", I believe at the challenge I was closer to 2" because I was to time crunched to shorten the pullrods to raise it higher.   2" at the splitter is likely tollerable as the minimum.  

You guestimates for dimensions are reasonably close.  I pulled the following dimensions from the model.  The Inner tire width (~51") is accurate to reality but the Outer width is a little wider then the car actually is.  The car is going to be ~73" O/A width.  The 62" width of the black inward air deflector can adjust to be anything from 51" to 73".  The ~14" height is the rough height to the upper A-arm from the bottom of the chassis.  The plan is to have the entire splitter / first element removable for loading the car so Approach angle is not as big of a concern.  

This ones a little wierd because of the cutaway.    But ~17" from the front wheel centerline to the front of the nose section is accurate.  The fenders stick forward about another 5".    The model has a 12" deep lower foil which can go bigger if it would be better.  

I know it's like a Tertiary concern but after adjustments I like the look of the 3 element foil.  Again I know dimensions/clearances/shapes will change quite a bit but the general look of the 3 elements, the inward flow directors under the headlights and the high shroud definately looks "LMP".  

Also the "uninterupted" span lenght on each side of the nose would be ~13" for the 3rd element and up to 17" for the 2nd element.

nocones said:

You guestimates for dimensions are reasonably close.

woohoo!?

nocones said:

Frame Ground clearance is designed to be 2.5", I believe at the challenge I was closer to 2" because I was to time crunched to shorten the pullrods to raise it higher.   2" at the splitter is likely tollerable as the minimum.  

I pulled the following dimensions from the model.  The Inner tire width (~51") is accurate to reality but the Outer width is a little wider then the car actually is.  The car is going to be ~73" O/A width.  The 62" width of the black inward air deflector can adjust to be anything from 51" to 73".  The ~14" height is the rough height to the upper A-arm from the bottom of the chassis.  The plan is to have the entire splitter / first element removable for loading the car so Approach angle is not as big of a concern. 

I'll start working the models toward 2" of clearance.  Off hand, mainly from a looks perspective, I think 62" is as wide as you want to make that reduction cut.  62"x4" is 248sqin, and 51"x4" is 208sqin... which means inlet speed would theoretically be raise from 60mph {88fps} to 72mph {105fps}.  I'm not sure that inherently means it'll be able to make 20% more downforce... but it might mean element 2/3 can operate at higher than normal AoA's.

We're already going to be stuffing a bunch of air through the suspension area... so I think widening the front over 62", might overload the exit's ability to deal with the air we're feeding.

nocones said:

I'll have to dig through my data to get you the CG information and spring rate items.  I'll stiffen the front to keep the splitter off the ground either way. 

this isn't super necessary, right now.  But it'll be necessary when we start trying to balance the front and rear.

I've got one to three more JavaFoil setup outputs that I'll work to try and get out between now and ~Sunday-ish.  After that, I'll probably have to slow down a little bit, and take some time to learn/implement scripting in JavaFoil.

Nothing to add other than to say this is super interesting to read to and although its way over my head, I am definitely learning something.

sleepyhead the buffalo said:

so, [...]

Normally a Clark Y will top out around a Clmax of 1.6 at 15degrees.  But proximity to the ground means it's maxing out around 7deg...

recall, that we've got to take that with a 60% grain of salt...

2.305 * 0.6 = 1.383

I'm not sure I've seen a span length estimated between the fenders, nor the first element chord length.  I'm going to go conservative, and guess it's 50inches in span and 18inches in chord (and I'll update if/when nocones gives an updated number).
thus, this splitter (which is 17% smaller than the last one), at 60mph {88 feet per second} would generate:  80#s of downforce

corrected:   62"x12" element at 88fps:   65#s of downforce

I figured, since most of nocones' renderings include sections that look similar to what Robbie built for the AMC, I should take the time to run a JavaFoil analysis on a cambered plate made of 0.5" plywood.

This foil is easily generated with JavaFoil, and a quick check showed the best results for the 60mph corresponding ReynoldsNumbers being with the maximum camber located at 20% chord.

so, taking our 60% inefficiency salt dosage...    2.106 * 0.6 = 1.264

with the updated span of 62inches and 12inch chord, at 60mph {88fps} this would generate:   60#s of downforce

One thing that needs considering is two-fold:
the same foil would produce:  165#s of downforce at 100mph (i.e. far end of drag strip);  and 310#s of downforce at 135mph (i.e. RoadAmerica trap speeds)... while resisting drag forces, and holding the fenders in the right place, and the moment load that will result from the downforce being generated 15 to 30inches away from the mount... and that's without including the assumption that some of elements' 2/3 downforce will be added to that load and moment resistance consideration.

I suspect that a "new foil" should be considered for the main front splitter element.  Something that's around 15% in thickness, and that the main element's chord be expanded from 12inches to 15inches.  The added thickness will give room for internal structure for the foil to handle the loads, and the additional length will help bring the structure closer to tying in to the front subframe structure.  Plus, a 15% thick section on a 15" chord, would nicely accommodate at 2" spar of some TBD material.

Oh absolutely.  The structure shown in cad exists because it's easy to make in CAD.  A template of a "good" foil shape will be made to form the actual parts that will be manufactured.  There isn't much benefit without doing actual CFD from the STP files to making the CAD model more then a rough shape to get dimensions and clearances.  When In Real Life happens more complex geometry will be used.  

sleepyhead the buffalo said:

...with the updated span of 62inches and 12inch chord, at 60mph {88fps} this would generate:   60#s of downforce

One thing that needs considering is two-fold:
the same foil would produce:  165#s of downforce at 100mph (i.e. far end of drag strip);  and 310#s of downforce at 135mph (i.e. RoadAmerica trap speeds)... while resisting drag forces, and holding the fenders in the right place, and the moment load that will result from the downforce being generated 15 to 30inches away from the mount... and that's without including the assumption that some of elements' 2/3 downforce will be added to that load and moment resistance consideration...

 

This seems like an argument for active aero to me.

nocones said:

When In Real Life happens more complex geometry will be used.  

How complex you wanna get?

I guess, one of the other questions (and feel free to PM me if you want) is:  how are you planning to build these?

My current thinking is that 'element 1' should have: "a heavy/strong 'two spar' setup at 25%chord and 75%chord... with periodic 'anti-drag' chord-wise stiffening."
such a setup has a bunch of advantages... even if it means the aerodynamics are slightly compromised.  especially if it means that that main foil can be used for both GRMChallenge, and UTCC.  with the heavy being "relative", compared to the spars one would frequently see in an aircraft wing (box spar at 25% and c-spar at 75%)

Is there any specific reason for the fender to "frunk" connector/covers.. wouldn't those trap air in the suspension? I don't know much the technicality of aero but shouldn't it be more effective to run it over the roof or split it around the rear wheels?

In reply to Gambit0117 :

I'm including them because I like the look and I feel it adds legitimacy to saying the fenders are "attached" to the car so it's not an open wheeler at track days.  

My understanding of the Aero is that being behind the front element the shroud filling in between the fender and body will not really do much provided it's flat and not adding frontal area.   The backs of the fenders will be open to allow any air under the fenders to escape.  

In reply to sleepyhead the buffalo :

My evil plan to construct is 2 built up thin Plywood I beam spars with foam making the shape.  Then a leading and trailing edge of Cedar will be made and the foam will be covered with thin strips of Cedar like a Cedar Stip Kayak.  After some sanding I will put 1 layer of S-glass and Epoxy over it.  

That will most likely wait until the challenge.

I like the pictures in this thread

so, yeah, the last ~5 weeks have been a thing for me.

lots of changes and reorientation going on, some mostly done, other less done.  As such, I haven't spent a lot of time on this; but, it's a good thing to pick up and put down from time-to-time when I get frustrated on other things.  For now... I'll keep my comments short, and maybe weekly, until after OneLap.

re: thin ply spars
https://www.rcsoaringdigest.com/Supplements/ANC-18_1944.pdf

plywood isn't particularly used for spar "by itself" in most of the wood aircraft designs that I know of.  most of the time it is used only as the vertical shear web material, with spruce square stock glued top/bottom with some additional spruce stock glued to support the two spars from displacing from the shear-web under bending loading  {see kr-2 wing spar here, here, here.  sometimes this is a single ply shear web and "C" beam, or double ply shear webs (front/back) to make a 'box spar' beam.  and sometimes the beam will be box-spar in the central section (with the center-center mounting part being full-height spruce to help transfer the loads into the fuselage) and revert to c-beam for the outer portion... among other load tailoring/weight-savings strategies.

it's been 20+ years since I've done these calculations... so, I probably not going to be a great source for number crunching on that end.  just keep in mind, there's not a lot of benefit to skimping on the chord-wise thickness of spar.  and it doesn't have to be spruce, sufficiently-clear-of-knots poplar might be strong enough, and might be more commonly/cheaply available?

if you've going to put foam between the spars and the cedar LE and TE... then the strips probably won't be necessary for strength;  unless you're planning to cut out 1" or 2" thick foam ribs between the spars, only and leaving a 2-4inch gap between ribs.  then, having the cedar would help your s-glass not sag.  although, you could do the same thing with some of the real thin foam panels at the hardware store (1/2"?).  I think, though, having full foam between the front spar and the front cedar LE will help with mitigate cone-strike damage  (I've got a wing we built for an AIAA competition of somewhat similar construction that took no damage when a 12# model bent a runway light during said competition at ~15mph).

if that's the case, though, there's probably some merit to using something akin to the Clark-Y... since you can "jig" the whole thing up on the bench and make sure everything assembles square/tight.  Not the Clark-Y specifically, though, since I think it's a bit thin at 11%.

Ok so I finally got the fenders to the point where they where ready to have the front element mocked up.  I have more pictures in my build thread but for the Aero conversation I have a few that I want to discuss.  

Note: all elements are just flat 1/2" foam.  The attempt at any airfoil geometry is eyeballing something that looks "the part".  

The wing portion of this is stepped 2" above the outer sections and 4.5" above the ground at ride height.  This would be reduced if a curved airfoil profile is used for the bottom of the main element.   Is this to high?  To low?

Dimensionally the width is 54" average and the fences taper at a 5* angle front to back.   The height available at the back of the fence is 13" and depth is basically unlimited.  

This offers a little better view of the profile.  

The side plates outside of the fences are just flat with some pieces of cardboard I had lying around to make fake vortex generators.  Unless there is a reason not to I will probably put these on the real front element because they look really cool.   

I modeled a sloted front slat style shape.  The main element is flat with a kickup and the front slat is 6" deep.   I doubt I would do this for real unless there is an advantage.  It does look cool but I defer to working cool for the main element.  

After previous discussions for the "Max Downforce" build I like this general look.  I am doubtful that something like this would be a great idea on a track, but the structure is removable so an alternative more Track appropriate element could be built. 

I know little about aero so this may be a dumb question but is there going to be an issue with the air from the rearmost top element packing into the fender? 

I'm working on some foil stuff... but I've got a bunch on my plate, and it's going to take some time.  Nevertheless, it's worthwhile, and I'm learning some interesting things as-is.

First though, some structural thoughts, and an open-ended thought about the forward jutting sideplates:

first, I'd suggest adding a support on the outboard similar to the inboard ones, if possible.  It looks like you've thought to do that, based on the geometry of the cut in those side panels.

I remarked in the build thread that you might consider adding adding a truss structure.  The red above is what I meant.  There's a pretty good open area between the two spars, and I think it'd benefit from this, at a minimum.  ymmv.

Also, dunno if this was a design consideration for those side panels, but... with them angled in (and depending on the instantaneous yaw rate), those panels are going to act a bit like a "LEX" (i.e. Leading Edge eXtension.... similar to the curved forward part of an F-18 / F-16's wing, that generates a vortex at high AoA).  This might be beneficial, it might not... hard to say.  Just, something to be mindful of... and it might mean those side panels are on the list of adjustment to make for the track version, if drag becomes an issue.

Finally, I've been working through 11% thick foils that are flat on the "up" side (for a car flap)... have a constant slope from 70 to 100% on the bottom Trailing Edge.  Then I've worked from a 'circular' Leading Edge (i.e. reaches maximum thickness at 10% chord), to an elliptical Leading Edge that reaches maximum thickness at 35%chord.   These are "basic foils" that are easy to make.  However... it'd be good to know the following:

how far out the pink foam extends from the front face of the 3/4" bar.  The distance from the front face of the 3/4" bar to the rear face of the 1" bar.  And, how far the foam extends from the rear face of the 1" bar... 'currently'.

I don't recall if I've linked this here before... but since we're going multi-element and "High Lift", you might look through a slide deck from a class I took 'back in the day' about High Lift Aerodynamics:
http://www.dept.aoe.vt.edu/~mason/Mason_f/HiLiftPresPt1.pdf

I'll probably be referring to it later w.r.t. Gap and Overlap chord percents.

*personal aside:  Continued and eternal thanks, Dr. Mason!

so... perhaps a little less well known, but most of the airfoils initially developed by the N.A.C.A. generally had their maximum thickness at 30%, and their maximum camber at 40% chord (i.e. the '4' in the NACA 2412 and NACA 4412 airfoils indicates that the maximum camber is at 40%, while all of the "4-digit series of airfoils' had their maximum thickness at 30%).

While nocones has asked for maximum downforce... there's an element of "easy to build" and "expediency of build" that shouldn't be forgotten.  Thus, a foil that has large amounts of flat sections might be beneficial... certainly for the main foil.  Also, there was a suggestion that the main foil be "simply rounded over, like with a router"... so, why don't we investigate that a little bit.

So, I rigged up a spreadsheet that generates foil coordinates with a 1% thick section and a circular round "upper" surface, with a "lower" surface that had a constant retracting slope from 70% chord to the trailing edge meeting up with the 1% thickness from the maximum 11% thickness.  Then, there was a series of leading edges that went from a point of maximum thickness at 10% (i.e. circular) to 15%, 25%, and 35% chord following an elliptical coordinate distribution  (it should be noted here, most foils don't follow this distribution, because it tends to create too blunt a leading edge... it just happens to be easy/quick to generate these points.  I also reused the 4% thick "splitter" foil from earlier...

I ran these at 25% chord above the ground (i.e. 4inches of clearance at the nose)... which means a minimum clearance of 2.5inches in the middle... based on a 15inch chord.  I can tweak this later... remember, the raw numbers aren't the point on this, currently (since they haven't been reduced by 60%), mainly we're looking at trendlines.  I also did a second run with the foils inclined at 2deg trailing edge up, with the pivot pointed located at the point of maximum thickness so the clearance would be consistent across the foils.

here's the results:

so, some initial take-aways:
- There's a benefit to a main foil being inclined at an angle (this will probably be less of an issue as we as multiple elements)
- The data indicates that around 15%chord  (i'd guess around 18% or 20%) has a maximum benefit in lift capability.
- There's an L/D benefit to pushing the maximum chord/thickness point back towards 40%
- All of these thicker foils are signficantly better than the 1/2" thick, flat foil, in generating downforce

I've managed to make some time... which is no small feat, since I'm using 4 element sections that have 225points each.  So, JavaFoil takes a while to run... and it tends to freeze up after 2 runs.  Some initial runs showed some coefficient of lifts that seemed... "outside the ballpark", and had me going to see if I could find some relevant coefficients of lift for F1 cars.  After that, I raised the sections up to where the bottom of the 'main' foil was 4" off the ground, and got the following (which is still 'high', but potentially near the possible ballpark):

Don't get too excited, and remember to take your 60% optimism reduction.  Thus, Cl = 11.71
some background.  This is the 'main foil' from before, with a maximum thickness of 11%, which occurs at 15% chord back from the leading edge.  The 1st, 3rd and 4th foils are all 9" flat 'board' foils of ~3/8ths inch thickness... the are "quick and dirty" secondary elements for now.  They are set with 3% overlaps in the back, and they're all at 30deg angles to the foil adjacent to them  (30deg up for the front foil, 30deg down for the 3rd foil, and 60deg down for the 4th foil.).

With just the first and second element, we get:

1.865 * .6 = 1.119
yes, if you go up to the previous post, you'll see this is a reduction compared to the original foil testing... which was done at a lower clearance, which gave better ground effect.  However, it seems like the 4-element setup is overpowering this lower ground clearance.

although, this is with the 1st element reduced at 15deg, instead of 30deg  ( I need to go back and monkey around with this on the 4 element setup, and see about inclining the whole thing at -2deg.

But, If we plug all this in with some rough measurements from above, we get something like this:

So... the top line is:  lots of downforce.  However, there's lots of drag too... like 270#s at 60mph.

I've ballparked F1 front wing Cl's around 13... and supposedly their whole car are around L/D of 5.  This is at 2, right now... so that needs improving.

I could use some feedback on sizing of the elements.  Is 500#s of front downforce at 60mph a good benchmark to try and achieve  (this would probably mean whole car downforce levels in the 1000-1200lbf range)?   Has the Buffalo wondered off into the funky clovers again?

Got any input Steve?

This is some awesome stuff.  I will have to really dig into it to digest it.

One question that immediatly pops out about Drag though.  How does 270lbs compare to any car shape item with similar frontal area?  That wing will be about 1/2 the frontal area of the car.  That area of the car will always be there wing or no wing.  I know it's "worse" then the best possible but how much worse is it?

Otherwise thanks so much for putting in this effort.  I've had a few people comment that the wing looks to big.  But I think to make "huge" downforce at Autox speeds that is just how it's going to be.  Obviously based on Dowforce #s and drag I wouldn't be able to run it at trackdays but I can easily make a simpler smaller track wing.

In reply to nocones :

it's 5x worse than the car by itself.  Although, that's based on me guessing about the cross-sectional area at 15 square feet.  Could you calculate that from CAD, taking a slice of the model around the passenger area?

But, if we calculated the car's guessed drag, plus this wing's drag... added it together and used that to generate a "vehicle Cd"... it'd come out at 2.4... which is nearly twice as bad as the proverbial parachute (Cd ~= 1.3).

It could be a javafoil going nuts thing, since we might be outside it's comfortable operating zone.

But, I think that's pretty obvious that I should look at this more, and see what's influencing that and attempt to tune it down to a more reasonable 50#s of wing drag, not 250#s (at 60mph).

sleepyhead the buffalo said:

I've managed to make some time... which is no small feat, since I'm using 4 element sections that have 225points each.  So, JavaFoil takes a while to run... and it tends to freeze up after 2 runs.  Some initial runs showed some coefficient of lifts that seemed... "outside the ballpark", and had me going to see if I could find some relevant coefficients of lift for F1 cars.  After that, I raised the sections up to where the bottom of the 'main' foil was 4" off the ground, and got the following (which is still 'high', but potentially near the possible ballpark):

Don't get too excited, and remember to take your 60% optimism reduction.  Thus, Cl = 11.71
some background.  This is the 'main foil' from before, with a maximum thickness of 11%, which occurs at 15% chord back from the leading edge.  The 1st, 3rd and 4th foils are all 9" flat 'board' foils of ~3/8ths inch thickness... the are "quick and dirty" secondary elements for now.  They are set with 3% overlaps in the back, and they're all at 30deg angles to the foil adjacent to them  (30deg up for the front foil, 30deg down for the 3rd foil, and 60deg down for the 4th foil.).

With just the first and second element, we get:

1.865 * .6 = 1.119
yes, if you go up to the previous post, you'll see this is a reduction compared to the original foil testing... which was done at a lower clearance, which gave better ground effect.  However, it seems like the 4-element setup is overpowering this lower ground clearance.

although, this is with the 1st element reduced at 15deg, instead of 30deg  ( I need to go back and monkey around with this on the 4 element setup, and see about inclining the whole thing at -2deg.

But, If we plug all this in with some rough measurements from above, we get something like this:

So... the top line is:  lots of downforce.  However, there's lots of drag too... like 270#s at 60mph.

I've ballparked F1 front wing Cl's around 13... and supposedly their whole car are around L/D of 5.  This is at 2, right now... so that needs improving.

I could use some feedback on sizing of the elements.  Is 500#s of front downforce at 60mph a good benchmark to try and achieve  (this would probably mean whole car downforce levels in the 1000-1200lbf range)?   Has the Buffalo wondered off into the funky clovers again?

Got any input Steve?

I think those numbers are likely pretty realistic. The body may impact them some in the final tally, but I would not be surprised to see close to 1000# at 60mph in max downforce trim with decent ground clearance. Drag, well that is definitely an area that JavaFoil has some challenges.

https://youtu.be/U2PX1p7nbw8?t=1795

It's not particularly difficult to make pretty good foam airfoils with a couple of templates, a hotwire, and some sandpaper. It's not a big enough project that one should just give up on it and use glass-covered foam flat plate elements instead: If that's your construction type (fiberglass,epoxy,foam), you're already 90% of the way toward having nice airfoils anyway.

Ok so I delayed and then scrambled to get the car ready for SubieFest.  In order to do that the Aero stuff had to go from Foam to something semi structural.  I also scaled some of it down a little.   I tried to take input from people however building this stuff is an excercise in drawing a line between what is physically possible, theoretically ideal, and practically executible by me on the time and budget constraints.   So saying that I apologize in advance if what ends up built doesn't reflect the hard work people have done.  

Also this is prototype round 1.  It probably won't get used at the challenge but will get used at the Subaru Challenge Autox in 2 weeks.  

Ok so that out of the way this is what I came up with.   It was the intersection of easy to build, used materials I already had on hand, would handle use and would look cool because that was 90% of this rounds purpose.  

Remember from before that the frame is a piece of 3/4" tube in front of a piece of 1" tube separated by 9".  The front element is made out of 2 pieces of 1/2" Plywood the flat bottom is 9" deep and the rear is 6" deep angled at whatever angle makes it rise 2" over it's length.   This is the bottom of the wing.  The leading edge is a piece of 24 gauge steel hand bent over a piece of 1" conduit making a 1.25" diameter LE which gives the wing a chord thickness of 1.3".   Overall depth is 15".   The top edge of the wing is a sheet of 24 gauge riveted to the 3/4" leading spar and also riveted to the 1" spar so it rises 1/4" over 9".  Because the angle of the 1/2 Plywood results in a 2" rise the sheet also rests right at the plywood .7" above the top.   That was a long way to say..  the result is actually a bit of an airfoil shape.  

Pictures to help explain.

Top side

Bottom side (I had to make a plate to seam the rear piece of wood because I didn't have enough plywood on hand to make it 1 piece)

The 2nd element is 1" above and 1/2" overlap with the bottom trailing edge.   It is just a flat sheet of 1/4" plywood.   They are 5" deep and adjustable.   The adjustment brackets are roughly 10* increments from 10* to 45*.   I made the brackets out of the edges of a stainless steel grill door that I found on the side of the road.  Yay free parts.  

Outside of the main wing is a bit of a throwaway.  I really wanted to get the fog lights on so I made some flat splinter pieces then put endplates and a set of diveplains on.   I'm not sure if it would be better to let the air go around the flight under the diveplains or block it off.  

Thoughts?  Overall I'm pleased with the Aesthetic.  It looks the buisiness but will it work?   It also is heavy.  Not sure I'd say "very heay" but the whole thing with fog lights, and mounting brackets is 40lbs.   But I can walk on it.   I think round 2 I can get it down to 20ish.  

Next post will go into the rear wing and I'll just do it all in this thread.  

nocones said:

Thoughts?  Next post will go into the rear wing and I'll just do it all in this thread.  

I think one thread is a good idea.  I'd change the title, but I needed to back away from that role a couple months back.  I'll see about getting some initial thoughts jotted down later this afternoon.