Canards - Location, design, angle, and size?

Working on my 1981 Camaro CP project, and the rules will allow me to run a canard setup should I choose to do so.

Here is the chunk of the rules I'm focusing on:

The splitter and canards may have endplates. The endplates may connect the splitter and the canard. The splitter and canard endplate total surface area is limited to 100 sq. in. (645.2 cm2) for each side. Canards are allowed and may extend a maximum of 6” (15.24 cm) forward of front bodywork/fascia as viewed from above. No portion of the canard may extend past the widest part of the front bodywork/fascia as viewed from above.

Issue 1: Size

I'm already running splitter endplates that measure out to roughly 55 sq. in. per side, leaving me about 45 sq. in. of canard, which brings me to question 1: is a 45 sq. in. canard actually worth the effort?

But let's get into the other questions, because even if it isn't worth it for my application due to the limitation, maybe the info will help others.

Issue 2: Angle

Through "research" which mainly consisted of bugging people who are faster than me, I want to have the canard between 20 and 25 degrees for an autocross application.  Does that seem right?  I know it's hard without proper modeling and testing, but I'm looking for a good starting point.

Issue 3: Location

This becomes a bit of a multipart question. 

Part 1 - I'll start with "height above splitter."

Here is a 22 degree angle placed on the side of my car.  I didn't place it in any particular location, but I wanted to have it there for visual reference.  Most single element canards I've seen are placed closer, vertically, to the splitter.  Is there a rule of thumb for height?

Part 2 - Fore/aft location.

More assumptions.  I'm assuming it is best to have the canard be as long as possible - stretching from the front of the vehicle all the way to the wheel well.  Because I am size limited, I doubt this will be an option for me.  So, looking down from above, will it be better to have the canard as far forward as possible, or as far rearward?  Does the air dam created by my splitter and fender flare impact this decision?  Would it matter if the canard was further forward than the splitter?

Issue 3: Design

I've seen lots of different canard design - some flat, some curved, some with edges that turn up into endplates, some that are triangle shaped and follow the contour of the body on the outside edge, some where the outside edge is parallel to the centerline of the car.  Generally speaking (and here comes the loaded question), what is the most efficient shape that is easy to create in a garage?


I know there is a lot here, and a lot is vehicle specific, but I've gotten some great advice here in the past and I'm wondering what the pros have to think.  Thanks!

I will put some thoughts together while the rest of the family is watching sportsball later today...

stafford1500 said:

I will put some thoughts together while the rest of the family is watching sportsball later today...

You are awesome!  Thank you very much.

Just wanted to add a specific measurement for my car that might help with some of my questions above.

Real rough math says (20.5 x 5) / 2 is about 51 sq. in., which is much closer to 44 sq. in than I expected.

hopefully not stepping on stafford's toes... but some background for understanding "why are canards frequently triangle shaped":

https://en.m.wikipedia.org/wiki/Leading-edge_extension#Leading-edge_root_extension

Those 'canards' that are shaped that way are acting similar to LEREX.  I mean, LEREX are really just a 'particular variety' of VortexGenerator (VG); which often times are what canards end up functioning as.

edit:    
to be clear, there's no restrictions on 'shape'/'thickness' contour of the canard?  Like LeMans had a restriction that front elements had to be 'constant thickness'.  Similarly, I've seen verbiage that requires 'flat canards'... even though the series that has this requirement (via FAQ)  frequently has curved canards.  So, figured I'd asked about that... even though I suspect a relatively thin 'plate' material will be the eventual recommendation.

sleepyhead the buffalo said:

hopefully not stepping on stafford's toes... but some background for understanding "why are canards frequently triangle shaped":

https://en.m.wikipedia.org/wiki/Leading-edge_extension#Leading-edge_root_extension

Those 'canards' that are shaped that way are acting similar to LEREX.  I mean, LEREX are really just a 'particular variety' of VortexGenerator (VG); which often times are what canards end up functioning as.

edit:    
to be clear, there's no restrictions on 'shape'/'thickness' contour of the canard?  Like LeMans had a restriction that front elements had to be 'constant thickness'.  Similarly, I've seen verbiage that requires 'flat canards'... even though the series that has this requirement (via FAQ)  frequently has curved canards.  So, figured I'd asked about that... even though I suspect a relatively thin 'plate' material will be the eventual recommendation.

As far as I can tell, there is no restriction on shape. Just size. 

Gimp,

In the same order you listed your questions-

1 Size: As you noted with your dimensions, 45 sqin per side will get you close to the remaining area within the typical location and car measurements. More size is better for increadsed force, as with all aero devices. Max out the size/area.

2 Angle: Optimum angle is an "it depends" answer. As Sleepyhead noted, canards are mostly a horizontal vortex generator. Because of that the angle is most likely to be a curved surface that rolls up the strongest vortex in the length provided. For a single dive plane, my suggestion is to getbthe trailing edge to end at the wheel opening around 45degrees from wheel top dead center. Total angle from leading to trailimg edge should not exceed 35degrees or so, due to drag maxing out and downforce not increasing significantly. For multiple dive planes (I suggested these numbers for JGs corvette dive planes), top trailing edge at 30deg from wheel opening TDC, lower trailing edge at roughly hub height at wheel opening.

3a Location Fore-aft: You want to maximize area AND keep the trailing edge at the wheel opening. The reason is the vortex you are generating is going to apply its low pressure to the wheel opening and drive lower pressures under the front of the car. It will also generate some local force on the dive plane itself.

3b Location Vertical: This crosses withe angle discussion above. The leading edge should be angled slightly nose down (5-10 degrees). The height of the leading edge above other parts (splitter) should not be closer than 3-4 inches. This is to reduce biplane effects between the dive plane and splitter, which reduces the effectiveness.

4 Design: Overall design needs to account for the wide leading edge leverage and the drag along the length. The strongest section is going to be a T-shaped section with the fasteners on the upper part of the flange and the working area coming off the middle of the flange. If you need to tune the area, you could build to the triangular area you measured and radius the outside leading edge corner to dial in the area.

The outer edge detail needs to amplifybthe vortex spool up. In your case, a small vertical fence (upper or lower) that tapers to nothing by about 75% of the length, will leave the vortex a path out from under the dive plane.

Curvature is probably easiest to make a simple arc rather than try to define a spline, but elliptic is going to be close to ideal with increasing local angle at the trailing edge.

One area you did not mention is the lateral angle. Horizontal is going to get you very close, but some outboard down can spool the vortex harder. It also buys some actual area if the rules are defined as planform (top view).

Overall, the goal is to spool up a strong vortex that drives the front underbody pressures and helps evacuate the wheel opening. Keep in mind the vortex can last a long time and may impact your spoiler corners.

Post up any other questions that these notes generate.

Let me start by thanking both of you.  Not only do you point me in the right direction, you give me all the "why" behind it, which is invaluable.  Based on both of your comments, I made a quick mockup.  Please pick it apart and tell me how far off the mark I am.

For any of my competitors watching and getting ready to throw paper, I have not calculated the area yet.  Hold on to your protest fee.

I'll start with the profile since this shows a few things:

• I tried to get the curve mentioned, around 30 degrees where the canard meets the flare and softening as it goes forward.
• I implemented a fence that runs about 3/4 of the canard.  This will factor into the area of the canard, so I didn't make it too big.
• The outside front edge of the canard is lower than where the inside edge mounts tot he bumper.

Here is the straight on front view that shows the outer edge being lower as well as the outside edge curve and how it meets up with the flare. 
Based on this:

Curvature is probably easiest to make a simple arc rather than try to define a spline, but elliptic is going to be close to ideal with increasing local angle at the trailing edge.

If I'm not understanding that comment correctly, please let me know.

Here is a 3/4 view:

You can also see that the canard starts near the bottom of the bumper, but well off the splitter.

Starting there worked out well, as I was able to get the leading edge of the candard to match the leading edge of the splitter.  That probably doesn't matter, but it does look good.
 

So... how off the mark am I?

I think that will get you 95% of whats available withe the area you have to work with.

The elliptic comment refers to 1/4 of an ellipse with the long axis along the wheelbase direction and the short axis vertical. Basically it says to ramp the curvature up more as you move back along the dive plane.

Remember that if you need to tune the area by radiusing the outboard leading edge, trim the outboard fence as well.

Having the leading edge more or less at the same place as the splitter leading edge should not pose any issues at all.

95%? I'm stunned to think I could be that close on my first attempt. I think that's an absolute credit to your explanations.  

stafford1500 said:

The elliptic comment refers to 1/4 of an ellipse with the long axis along the wheelbase direction and the short axis vertical. Basically it says to ramp the curvature up more as you move back along the dive plane.

I think I understand what you mean now and that makes sense.

So, if that's the case, I could leave the outside edge (with the outside fence) straight, which would make it easier to build, correct?

Gimp (Forum Supporter) said:

So, if that's the case, I could leave the outside edge (with the outside fence) straight, which would make it easier to build, correct?

That was going to be my recommendation

sleepyhead the buffalo said:

Gimp (Forum Supporter) said:

So, if that's the case, I could leave the outside edge (with the outside fence) straight, which would make it easier to build, correct?

That was going to be my recommendation

Awesome! Thank you. 

Any thoughts, Steve, about adding a bit of thickness/curvature to the front 25% to encourage the bottom flow to stay attached and maybe induce more low pressure there... ala Liebeck "single element theoretical maximum"...

and Ananda/Selig's "Bird Foils" analysis?   

 

edit:   
dropping this link here so it's easier for me to read later....   
https://etd.ohiolink.edu/apexprod/rws_etd/send_file/send?accession=osu1586450691890636&dispo

I love this bar. 

Gimp (Forum Supporter) said:

I love this bar. 

Hell, I'm just glad this is the one place SCCA wasn't idiotic enough to say "{element} shall not act like a wing"

i really hate that phrase in their rulebook(s)

 Steve - can I get a little more clarification on this bit?

stafford1500 said:

Remember that if you need to tune the area by radiusing the outboard leading edge, trim the outboard fence as well.

Mainly what you mean by "radiusing the outboard leading edge."

I think this is what Steve is getting at...

That makes sense.  Thanks!

Not to distract from the deep aero discussion here, but we need more pics of that whole beast of a race car.

In reply to 84FSP :

There are a few out there.  This one is my most recent favorite.

More in this thread - https://grassrootsmotorsports.com/forum/build-projects-and-project-cars/1981-camaro-c-prepared-build/59501/page45/

Catching up after a few posts...

By breaking a flange on the outboard edge you will probably wind up with some curvature, unless you are using a brake. If it goes wrong way (upward flange will likely make the wrong curvature) a shrinker/stretcher can fix that.

Adding thickness to the front of the section will not appreciably change the results at the aspect ratio and speeds invloved. Also since this is a vortex generator the bottom surface will have the correct pressure gradient to maintain attachment.

Sleepyheads sketch summed up my comment about the outboard leading edge radius.

The basics of the aero stuff capture most of the target. It is always the last few percent that take the most effort. I spend my working days chasing tenths of a percent gains. How you attach the dive plane to the body will be the most stressed area, so you may need to add a fillet under the dive plane to thre body mounting flange. This can be done after the basic form is prepared.

Gimp (Forum Supporter) said:

I love this bar. 

Yep, my favorite hang out too, and the drinks are always free

Going to pick up some material and the shrinker/stretcher next weekend and start to get at some templates.  I'll update with progress when I have it!

Okay... I lied.  Curve ball time!

I'm not good at math.

I took the above canard mockup off the car and started to confirm the area.  While the simple "from above" view made it look like it would work, curving the canard, adding the angle to match the splitter, etc, created a lot of bloat.  Well over the 44 sq. in. I have in allowance.

So, using an online calculator because I can't math, I started trimming it down and things got a good bit smaller:

Now, this is still something I can make, but it leads me to a new question:

I am allowed 100 sq. in. (per side) to distribute between an endplate and a canard, in any ratio.  What, in theory, would yield the best aero?

• No canard, 100 sq. in. endplate
• 100 sq. in. canard, no endplate
• 45 sq. in. canard, 55 sq. in. endplate (what I have mocked up now)

If it was you, which path would you take?

Gimp (Forum Supporter) said:

I am allowed 100 sq. in. (per side) to distribute between an endplate and a canard, in any ratio.  What, in theory, would yield the best aero?

• No canard, 100 sq. in. endplate
• 100 sq. in. canard, no endplate
• 45 sq. in. canard, 55 sq. in. endplate (what I have mocked up now)

If it was you, which path would you take?

I mean, I rarely have "my stuff together" to say "I'd take the best path"... so instead:

The "standard" aero answer would probably be something along the lines of "that's a great thing to test all three configs at a test and tune, and go with the best"

although, I'd probably tweak the options to:

• No canard, 100 sq. in. endplate
• 100 sq. in. canard, no endplate
• ~60 sq. in. canard, ~40 sq. in. endplate (what's detailed in the picture below*)

we'll see what Steve says.  But, I'm inclined to add area to the canard over the endplate, and tweaking the endplate shape to enhance tripping the flow "over-the-top-the splitter"/"around-the-airdam" into a second low vortex, while also taking away area from the endplate to put into the canard.  Thus, bringing the end plate leading edge to back to line up with the rivet closest to the forward line of the airdam before it turns, and making the endplate more of a "triangle" shape... does a similar job to the canard/canard-fence that Steve's outlined.

this is kind of the "all in on vorticity" approach.

*those areas are just WAG's.  Cardboard up the rough outline of what I've napkined here to get an endplate area.  See what Steve chimes in with.  Go from there?

It's probably easiest to test the following two configs on a single test&tune day:

• No canard, 100 sq. in. endplate
• ~60sq. in. canard, ~40 sq. in. triangular endplate