Presented by Nine Lives Racing
1 2 3
nocones PowerDork
7/12/22 12:57 p.m.

So the rear wing is much simpler design.   To make it I took 1" conduit at the leading edge and a piece of 1*1 about 9" behind it.  I welded a center piece in and then welded some end pieces on.   Over this I bent a piece of .040 aluminum.   I riveted the aluminum down to the 1" tube on top and on bottom.  I then deflected the aluminum up to make a banana shape and clamped and riveted the trailing edge. 

 This was surprisingly effective.  It creates and airfoil shape, and was strong enough to get towed backwards at 75 MPH with no apparent vibrations.  

The wing is 69.5" wide with a 1.5" thickness and a 16" depth (8 sq-ft EMod/BMod limit).  

I'm somewhat concerned with the unattached nature of the surface sheets that there may be vibration induced boundary layer separation.   

The height above the car is way to high.  I was in a hurry and the center stantion started at the height of the roll bar (for the eventual shark fin).  I then didn't lower it enough so if needs to be lower.   

The angle of attack similarly was guestimated and it is to steep.  I plan to make adjustable mounts.   

The stanchion is a 3/4" tube with triangulation by 1/2" tubes.  I can hang on it, push the car around with it etc.  The end supports are 1/2 tube and the endplates are just 1/4" plywood.  Again they survived driving backwards at 70 so I know they will be good for the challenge and the design will be kept they will just lower with the wing when it adjusts.  

I am somewhat concerned about torsional stiffness relative to the car as well as lateral.  I'm worried it will rotate relative to the car.  Before I go crazy with more mounts I'm going to try adding triangular wires.   I have a old bontrager bike wheel with Aero spokes that is radially cracked and can't be used.  I will asign an FMV to it and use the spokes to make some bracing wires to see if that stiffens it.  It's not bad now but I think it could be improved.  


I will get some more pictures of the mounts Tonight as I don't have them.  I can hang from the wing so it at least can handle 240 lbs of Downforce.   It's not to heavy either at ~10lbs.   This one likely will stay.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 1:53 p.m.

So, one initial clarification is: are the "spars" for the front main element biased "up" or "down", to share a plane?

also, I suspect you're 15"  chord estimate might be off.

*drawing only to scale-ish (and scaled wrong... hang on)

nocones PowerDork
7/12/22 2:00 p.m.

In reply to sleepyhead the buffalo :

B is almost exactly right.   It is 15" from the leading edge to the trailing edge

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 2:03 p.m.
nocones said:

In reply to sleepyhead the buffalo :

B is almost exactly right.   It is 15" from the leading edge to the trailing edge

I'd think, though, that the front 24guage that was bent around 1" conduit would project forward from the spar.. ~0.6" ?

nocones PowerDork
7/12/22 2:10 p.m.

I'll measure the exact dimensions of the rear and angle that will answer the locations of the spars for an accurate representation of the current part.   I think the rough design works and If there is some slightly better geometry I could adjust it when I replace the bottom surface with another sheet of 24gauge that could remove the flat 9" part and result in a curvature.  It would also drop a hilarious amoubt of weight as the ~15"*54" 1/2" ply isn't light. 

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 2:21 p.m.
nocones said:

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. 

honestly... for the upcoming AutoX test, I'd just leave the fog lights there and not worry about fairing them in.  Yes, leave it open to allow flow to attach to the underside of the upper diveplane.  Eventually you'll want to add a smoothening panel there, which will take lots of high pressure air from the front outside of the car and compress it into a sheet that will help block the wheels from drag producing interaction with the 'normal' flow.  Go around the parking lot, looking forward through the front wheel well of many cars made after ~2018 and you'll see a similar element in the bumper.

one benefit will be that the compression here will mean that you can potentially really crank up the aft-end of the diveplane (and corresponding outer-splitter) and keep the flow attached.  Although, that will also lead to a trade-off between rear downforce balance and drag creation.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 2:34 p.m.

the parallax isn't doing us any favors... but this, after all, is less about 100% accuracy.  Instead, being more about "getting inside the ballpark":

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 2:37 p.m.


I don't know if I'll be able to do it (my being out there is still tentative), but I should probably figure out a way to "meet up" with this thing around the GridLife Mid-Ohio event.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/12/22 8:04 p.m.

I'm going to start with some analysis of the rear foil first.  I used the above outline, and a version where the outline was 'filled in' and sucked them into JavaFoil to generate two sets of coordinates.  I brought those both into XFoil and used the "interpolate" command to 'average' the data between them, to try and smooth out some of the 'roughness' that tends to creep in from tracing the outline with three-segment, four-node curved line segments in  This 'combined' foil was then expanded to 200 points via XFoil's PANE command.

I ran a viscous polar on it at Reynolds Number {RE} 7e5 (700,000) from AoA -5 to 20, in 0.1deg steps.  Then I took the foil into the 'MDES' design menu and used its FILT command to try and smooth out some of the roughness... with limited success.  Then I ran a second polar, which showed minimal (mostly in the drag curve) changes.  Please keep in mind, these results are 2D, only... and don't account for GroundEffect, or 3D Wing-Plan Effect:

I used the same 'filtered' 200 point file to pull into JavaFoil, and then went through setting up the coordinates in the 'format' that JavaFoil prefers.  With these JavaFoil-specific coordinates, I performed 3 Polars sweeps with JavaFoil from RE 600,000 to 1,000,000 and "-7" to "20" deg AoA.  The three setups were: 2D (infinite AspectRatio, AR), AR = 5*, and AR=5 and 250%chord above the ground:

*a word on AspectRatio = (b^2) / S  = ( wing span squared ) / ( wing planform area ) = ( 69.5^2 ) / ( 69.5 * 16 ) = 4.34
   but, there's EndPlates... so using the linked endplate effect estimate, and guess that the endplates are 6" tall/deep:  4.34*(1+1.9(6/69.5)) = 5

**a word on Reynold's Number (RE)
    this helps 'us' account for the nature of the flow (laminar regime vs. turbulent regime; or 'Low Reynolds Number' {e.g. less than ~500,000} vs. 'High Reynolds Number' {e.g. 1,000,000 and above}).  Calculated by:  RE = rho * V * L / mu   (where I'll be using 1976 NASA standard atmosphere values for Sea Level for rho (air density) and mu (air viscosity))
RE = 0.002377 * (88) * (16/12) / (3.737e-7) = 746,000
which is reasonable to round down to 700,000 for now.

I'm not real sure what the effect of the 'unsupported' low-pressure side of the skin will have on things.  The foil clearly has a pretty sizeable nose-radius... which in this case is good.  Plus, the conduit isn't too far off the skin, if drag at the stagnation point 'pushes' the skin back, and the low pressure suction trying to pull the skin "down".  For challenge... I'd guess it's not a big deal... especially if you shallow out the rear foil for the drag passes.

I think Cl.max of 1.75 is a reasonable estimate of this foil's capability, including 3D/Ground/Endplate Effects.

Rear Downforce, 60mph {88fps}, 1,112sqin = 0.5*0.002377*(88^2)*(7.72)*(1.75) = 124#s
Rear Drag,            60mph                               = 0.5*0.002377*(88^2)*(7.72)*(0.06+(0.0637*(1.75^2)) = 18#s

for ballpark sake... there's some public data for a cat-ish foil that's 70" long (9.25" chord) with 8" endplates.  At 60mph and 10deg AoA it pumps out a Cl ~1.76 for 76#s of downforce, with 7#s of drag.  So, you've got 2.5x the drag for 60% more downforce... mainly due to the longer chord and 'get it done' foil shape.  On the other hand, it's what... $5 out of pocket?

I've been thinking about the C.G. / Aero-Balance thing I brought up back a page.  I think it'll be better/easier to reference these to the 'engine side' of the firewall.  So, when you get a chance: can I get a distance from the nose of the foil to the backside of the firewall.  And, similar distance to the 3/4" square tube on the front main element?


I should note, when I pulled the sections and popped it into XFoil, it was estimating a maximum thickness that was 12% of the chord, at a point 20% back from the nose... or a 2" maximum thickness 3.25" back from the nose.  Looking at that a bit more, with squares scaled off the front conduit...

I think I might slim this down to 11% (i.e. 1.75" thick) and give it another go.  I don't think 1.5" thickness is right, though?

as for camber... the codes were evaluating it around 6%... which, eh.  I'll have to dig around a bit, and see if I can get you a 'height' above the 'flat-ish' "top" skin back at the 1" square tube 'rear spar', to check against.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/13/22 12:39 p.m.

Alright, lets get the Front Aero estimate out of the way...

I took the information gleaned from the last couple of days and made some modifications to the "flat board foil development" algorithms(?)... spreadsheet wonkery?  Anyways, I took what I'd built for stuff on the last page and used it to generate relatively reasonable approximations for the front foils nocones had built:

You'll notice that these foils are open at the TrailingEdge... which I found that JavaFoil really didn't like... especially at these TE thicknesses (it was giving me Cl  estimates of 0.14).  After adjusting the coordinate spacing to JavaFoil's preferred spread, I then added in two points to close up these foils with a 90deg corner, and then spaced the main foil up to GroundEffect for a 4in minimum spacing (iirc, that's the current working height between the front plywood fences?).  It's a ~15" chord, so again RE = 700,000:

Let's remember our 60% rule... 1.479*0.6 ~= 0.887

which, I think is pretty reasonable.

I then went through and added the second foil, spaced it's coordinates, resized it to 33% of the main chord (5" element compared to 15" main element), closed it's trailing edge and then moved it to a 0.5" overlap (i.e. 96.7%c) and 1" gap (which I took to mean from the top of the main element and bottom of the 2nd element, or 48.4%c from the ground... or aka 7.26inches.  I might have to go back and recheck that value).  Then I did FlowField analysis for the 2nd element deflected at 10deg, 20deg, 30deg, and 45deg.  I applied the 60% lift reduction, and summarized the results in this table:


There's two pretty important assumptions to this data, which go in-hand with nocones request for "simple things to do" for the current build.  The following two things I think are fairly important for making this data reasonable.  The first is to round over the front edge of the 2nd element with a 1/4" round-over bit, top and bottom.  This will help the flow over the top of the main foil follow the 2nd element (important)... and mean the real-world results more closely match these results (less important).

Second, I strongly suggest that before the autocross you tape the open joint between the two pieces of plywood, as well as the trailing edge of the 24guage "Leading Edge":

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/13/22 12:58 p.m.

That's cool and all, Sleepyhead... but, what's it all mean!?

So, why don't we drop things from these analysis into another worksheet that sums it all up, along with a couple of working guesses / assumptions about the LMP360:
Ax = 15sqft (cross-sectional area, which isn't the same as "frontal area"... although either could be {and often are} used)
Weight = 1800#s + 200# driver
Arm Length from the front element to the back of the firewall:  66inches
Arm Length from the back of the firewall to the quarter-chord of the rear wing:  54inches
Pre-existing span dimensions from the last page
20% across-the-board endplate adjustment to AspectRatios (although, I suspect I need to tweak the front element 'lift-induced drag' valuation on AR up a bit more, maybe?)

I ended up using the 10deg deflection on the 2nd element, based on the moment guesses for front and rear from the 66inch fron arm and 54inch rear arm.

This doesn't include the contribution of the diveplanes, nor the diffuser.

Personally, I think there's a lot of value in "running it as it is" coming up in 2weeks.  Although, I'd suggest that you add a couple of pre-built tension wires (and attachment hardware)... so that you can potentially tweak the rear wing foil trailing edge... "up"?  I.e. bend in more curvature/tension in the aft part of the rear foil section so it has more camber.

Alternatively (or, better yet additionally ), I'd suggest working up building a 4 or 5" chord 2nd element to the rear wing out of 1/2" ply that's supported at the ends and the middle (so the gap stays consistent).  It seems pretty clear to me that the front elements you've built so far can generate quite a bit more downfroce than the rear can (moment arms, depending).


Note:  when/if you add a diffuser, that won't impact this front-rear balance quite as much, because the force of the diffuser will impart its effect very close to the CG; so it's effect will primarily be 'overall downforce' increase; and drag decrease {iirc}.


That's about all the analysis I can due for the time being with the available information.  If you've got any questions, or you've got a scenario you'd like me to look into, chime in... and I'll see if I can analyze and/or answer them.

nocones PowerDork
7/13/22 1:10 p.m.

Oh man.  This is so much great information.  I have the background to understand it but don't immediately have the time.   But I will put in the time.

My initial reaction after reading is run what I've got, report back with some dimensions and then I can ask better questions when I get some time to digest the immense effort you've graciously put into this.  

I will do some reinforcement of the trailing edge of the rear and tape the bottom of the front for sure before next weekend.  I'm hopeful to add the 2nd element as well Is the current angle of the rear estimated as "good" or should Iower the angle when adding the 2nd element?  

Just to repeat and be perfectly clear thank you so much for the time/effort you've put in.  

nocones PowerDork
7/13/22 1:18 p.m.

Oh and a question to plant a seed.  

I've built the sides of the car to be able to accomodate a 4" deep x up to 8" wide tunnel down each side.  It will be a bit before I get to it but it's probably worth discussing what If any benefit there would be to running tunnels into a diffuser vs a flat floor with a diffuser 

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/13/22 2:45 p.m.
nocones said:

I will do some reinforcement of the trailing edge of the rear

Is the current angle of the rear estimated as "good" or should [lower] the angle when adding the 2nd element?  

Spaghetti, it's what's for lunch!

so, if the green lines are close to 'flat horizontal in the real world'... then I think that wing is set to 21deg?  And, if you put a digital inclinometer on the flat-ish bit between the mounts, I think it'd read 20deg?

if so, then you're 5deg high from the 'optimum' of 15deg.

of course, we also don't know how much "downwash" the cab/rear body work is going to influence the AoA into the rear element.  So, there's an argument to be made for 10 or 12 deg setting.

Although, if you go back up to the Cl-alpha and Cl-Cd curves from XFoil, you'll see that the big round nose, relatively low camber (i.e. average foil curvature) on your rear foil means that it's holding on producing lift (just at the cost of drag) even at 20deg.

so, if you can, back it off to 15deg... tuft the bottom of the wing, and point a GoPro at it when you're doing 60mph to give us an idea if/how-much separation is going on and where.

The reason to have the wires is partially to aid the rear of the rear foil from 'flattening out' at speed.  But, also, we could crank on it and raise the curvature and try and get more lift out of the existing rear structure.

Final thought for next weekend:   
I know you're on 200tw vs Hoosiers, and I think there's been some other minor suspension tweaks 'over the break'.  So, I'd advise you to try and get some runs (whenever possible/practicable) without the Aero installed, just to make sure you're not chasing a mechanical grip imbalance while fidgeting with the aero.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
7/17/22 6:33 p.m.
nocones said:

I've built the sides of the car to be able to accomodate a 4" deep x up to 8" wide tunnel down each side.  It will be a bit before I get to it but it's probably worth discussing what If any benefit there would be to running tunnels into a diffuser vs a flat floor with a diffuser 

Underbody design is definitely an area that I'm a little less familiar with.  I appreciate a clue about it that Steve gave me a couple years ago.  Additional, since this comment, I've gone back and read the section on rear underbody shapes in Katz.  I've also read Toet's article on his (?) site / RaceTechMag.  I've also combed through a couple of technical papers, and found one by Kevin Peddie & Luis Gonzalez "CFD Study on the Diffuser of a Formula 3 Racecar"... which was helpful for me (search around, there's a couple different links to download the pdf of it).

This big takeaway for me, which is similar to what Steve Stafford has stressed on the pages of GRM:

or, another way to talk about underbody aerodynamics, and rear exits of said underbody flow... is to think of these similar to one thinks about radiator ducts.  They end up being very similar, and heavily influenced by the relationships identified in Bernouli's principle:  That for a particular set of intial flow conditions, changes in area will have deterministic and correlated impacts on flow speed and pressure.  And, similar to radiator flows, putting the exit of the underbody flow in the presence of a low pressure zone, will advantageously improve the flow through the entire underbody.

There's some other details with respect to vortical flows and their ability to keep the underbody flow attached at "higher than normal" exit surface angles/areas.

This is a really long and convoluted approach to answer any aerodynamic engineering question with the classic engineering:
It Depends

Reading through, there's quite a bit of evidence that having a minimum clearance height at the "choke" point of the underbody being around 1" is beneficial (assuming you can keep it relatively consistent {see, Formula1 2022}).  It also looks like having some form of flow acceleration 'forebody' could be helpful in aligning the flow into the choke point.

Eventually, though, the question becomes: "how big can the exit area be?"
Which leads to two further items to consider:  a)  "do you want the underbody flow to intersect the lower a-arm {a flow disruption}?  b) based on the size of the exit area, was is the minimum choke area?   (i.e., you/I/one would probably end up nominally working backwards from the rear of the car to answer your question).

I suspect that the exit area available, and the angles involved might mean that you can't use the entire volume to feed the tunnels, similar to what's found on a Porsche 962.  But, we should probably start with exit area... and some thinking about how the tunnel 'wants' to interact with the rear suspension (I suspect for challenge, keeping the tunnel out of the suspension has lots of benefits).  Then we can think about ways to use the extra volume, if there is any, beneficially.

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
8/2/22 11:07 p.m.

so, underbody is in the process and  nocones posted up a couple images in the build thread...

off hand the broad strokes of it look good.  and I agree with the idea of a simplified floor shape that can be built quickly.  The exit looks pretty good, i like that it's expanding vertically and horizontally behind the tires.

I wonder about the ability to poke the exhaust cans exit out through the top of the "engine cover"...

I think this Peugoet picture from Mulsannes Corner 2011 news page explains what I'm suggesting... ish:


as for the planning drawing...

the main thing I'd like you to consider addressing is in "green", and that's to try and change the as drawn shape head of the main frame from a "concave" profile to more of an 's-bend', so the shape more closely follows the expected flow line you've drawn... without detaching from one section of the body to the the next.

The pink question mark, I wonder a little about what's drawn?  Although maybe that's because you're drawing a section that's 'flat' at the ~height of the underbody... and you're just going to have a flat plate that connects from the outboard aft portion of the front tire 'barge board' to the underbody section?  otherwise I'm a little concerned that you're putting an obstruction in there with the green triangle thing there?

so, the next part, I'm having trouble think about how to talk about.  The light blue line, I'll call: 'the inboard fence' for the side-exit flow... probably doesn't need to be a complicated curving shape.  Although, you might build in some ability to move it in/out side to side but holding the "as drawn" angle.
Here's where things get... harder... and I've made two attempts to try and highlight... neither of which I'm happy with:  being the "cross-section" from the yellow arrow.  The point is to encourage you to keep in mind the angle of the 'inboard fence' and the idea that some of the air going through the underbody is inherently not going to follow that fence.  But: that's ok!  In fact, I suspect it's actually really good.  Because, some of the highspeed flow coming into the underbody from between the wheel and the body is going to want to curl up into the main underbody that's going to head back to the diffuser... and in doing so while hopefully roll up into advantageous vortices.  The presences of these vortex-fences are pretty common on the current F1 cars.... and some of the reading I've done indicates that a lot of effort is going into make vortexes in the underbody, because they keep the flow attached and allow steeper than 'rule of thumb' exit area geometries.  So, you might want to build in some ability to angle up the last ~9inches of the diffuser ceiling in case "we"/"you"/"it" manages to roll up some vortex from that fence (or we go crude and fashion some simple bent aluminum vg's?) and use some tufts to figure out that more angle can be pushed.

clear as mud?

nocones PowerDork
8/3/22 1:25 p.m.

Quick thoughts

I'm going to investigate turning the exhaust up to free up lateral space tonight.  

The green outer area is the behind wheel board.  I'm planning to just have the floor be open between that and the "inlet" dividing element for the outwash area.  The floor will tie to the back of the fender so it will have some just flat profile between the outer fence portion and the back of the fender piece.   

It sounds overall like my plan is okay to move forward with.  My diffuser growth will be kept under 10⁰ until after the chassis just because there isn't enough space to do more.   After the chassis it can grow as radically as we want to try.  

Profiling the underfloor area for the inlet will kinda be a Second piece of this.   So I will make the tunnel inlets and outwash area and then deal with the underfoot after but I understand your recomended shape.  I will post some more pictures as I get to that end of the inlet.  

For the outwash area if that fence doesn't have to be curved it will be easier to construct.  I'm thinking I will taper the outer edge of the tunnel for a little more volume.  Or is a hard 90⁰ to the ground important?  If I taper the profiles will wind up nicer.  It would also then look a lot like a modern F1 floor with the "tunnel" volume tapering to the outer floor edge behind a pair of fences that divert some amount of flow to the outside.  

sleepyhead the buffalo
sleepyhead the buffalo PowerDork
8/4/22 2:12 p.m.
nocones said:

For the outwash area if that fence doesn't have to be curved it will be easier to construct.  I'm thinking I will taper the outer edge of the tunnel for a little more volume.  Or is a hard 90⁰ to the ground important?  If I taper the profiles will wind up nicer.  It would also then look a lot like a modern F1 floor with the "tunnel" volume tapering to the outer floor edge behind a pair of fences that divert some amount of flow to the outside.  

you could round/fillet the corners of the fence where it meets the "roof" of the underbody.  Where the fence is being used to created vortices into the 'rear tunnel' volume, you'll want the lower half to be "thin", with "sharp angles" where it juts down towards the road surface... that way the flow will "break" rather than try to 'stick'... and with the difference in pressure front side to back side (high vs low) it'll encourage it to 'roll up'.

if you're asking about 'widening' of the area/volume of the side exit.  It's probably a 50/50 proposition at this point... as is the question of should it have a "hard downstream corner, or a rounded one?"  For challenge it's probably somewhat academic... and the tradeoffs will be more important when you start using it for track work... and figuring out the right tradeoffs will probably involve some testing, imho (which is why you'll notice the large amount of variability in sidepod shapes between F1 and LMP {and different generations of LMP}).

1 2 3
Our Preferred Partners