2D Computational Fluid Dynamics simulation for our 818R - APR GTC-300 Rear Wing

[SixStar]

We had Zebulon Motorsports Consulting run some CFD on the APR GTC-300 wing to see what AOA (angle of attack) we should start with given our down force and top speed goals. They found out that an AOA of ~5-10 degrees would be ideal on our 818R since it has no windscreen or roof. The suggested starting point of the GTC-300 should even be set much lower than APR suggests given the design of the wing. Even at a 0 AOA there will still be separation on the sides since the wing isn't flat and exhibits flow closer to 15 AOA on the sides at a 0 AOA.

All simulations are at 100 mph. Vector traces are scaled
from 0 to 350 mph as seen on photo scales.

0 Degrees AOA
0AOA.jpg

15 Degrees AOA
15AOA.jpg

25.5 Degrees AOA
25.5AOA.jpg

From these simulations we can draw some interesting conclusions about the wing design.
First, remember that this is a twisted wing. As such, when they have the wings at 0 degrees
AoA, APR expects that the effective flow-based AoA is ~15 degrees. This can be derived
from the ~15 degree twist built into the wing. When the center is at 0 degrees AoA, the
outer edges are at 15 degrees. This makes the 0 degree simulation null, except in the case of
cars without windshields – like the 818R. At 0 AoA, you can see the extreme camber of the
wing is detrimental; the airflow cannot compensate and separates off the top. This is not
extremely bad, but it does limit the wing.

At 15 degrees AoA, or what the center section would effectively be at on a sedan, and
what the outer ends are at to start, the wing already separates off the back. In 3d, this would
not be as pronounced, but with this degree of separation it would still exist to some extent
and be highly detrimental to performance.

Hope this helps some folks when setting up their aero. Remember, tuning is a science and takes time and patience but having a good starting point never hurts!

[shinn497]

WOW how did they do these?

[Santiago]

You just need a sample wing, a 2-D rendering or 3-D model of it (easiest to acquire with a 3-D scanner), and software to run the CFD computations. Not common for sure, but the basic tools are out there and available to folks w/the time to learn the programs and money to afford the tools (which ain't cheap brother!).

One thing to keep in mind is that these are simulations, and like all such things their quality is subject to user input. While simulating wing profile on its own produces valuable information, wings aren't run in isolation. They're run on cars, and those cars have a significant impact on the airflow the wing sees. That's some of what Zebulon was referring to in their references to "null cases" you wouldn't normally expect to see (but can/do on the 818R) and their summary of the "sedan" cases (which would be the norm for most consumers). For my part, I prefer to see CFD analysis done on close approximations of their destined home, that is, on a simulation which has a car ahead of it. [Stay tuned...it's coming...]

Back to the value of a wing-only simulation - you'll get the most out of it if you are thinking "whole car" while trying to make sense of the data (even if you are not looking at a whole-car simulation). The APR published data doesn't discuss much of this, so they are able to throw (pretty big) DF figures at you in isolation. I really prefer to see a discussion of simulation results start the way Zebulon does, framed in considerations of application.

Thanks for posting this Six.

Best,
-j

[Wayne Presley]

I could have saved you time and expense. We ran the GTC 300 wing on the back of the 818R during testing and found that 0-1° AOA worked the best. At 8° we could not get the push out of the car. We also ran it without the wing and it was sooooooo loose that John took it out for only one lap in that configuration. At 0-1° AOA the car was near perfect, you could trailing throttle oversteer it, drift it on the exits and had just a slight push in the 100 mph corner.

[longislandwrx]

Thanks for doing this, very good to see what's going on back there. If you could do it again would you have chosen a different wing?

[Santiago]

So Wayne, am I getting this right:
- No wing: rear is uncontrollably loose
- At 0-1° AOA: rear is balanced
- At 8° AOA: rear is so tied down, you can't get enough front grip to compensate

If that's the case, sounds like only a relatively modest amount of DF in the rear is required to balance the car (w/the supplied front splitter). Adding more rear wing isn't going to get us far in terms of useable absolute DF values unless we've got more game in the front. We're just going to see more push.

Caveat: adding a more efficient rear wing may not get us more useable DF overall, but it may be worth quite a bit in getting us the same balanced DF level at a considerably lower drag penalty. So top-speed should benefit even if cornering prowess doesn't.

Wayne, in some of the videos John posted the car seemed to tap out at 130mph even on sections of track that seemed to offer more room to run...do you know if that was intentional, rev-limited, or aero-limited?

Best,
-j

[CHOTIS BILL]

I would try to increase front down force to make use of extra rear wing down force rather than down sizing the rear wing.
But then again I like lots of down force.

Bill Lomenick

[Santiago]

I would try to increase front down force to make use of extra rear wing down force rather than down sizing the rear wing.
But then again I like lots of down force.

Hell yeah!

But that was my point...folks interested in "more down force" need to look up front first. Folks looking for "less drag" (as a primary goal) might be content with a more efficient rear wing.

Best,
-j

[EODTech87]

What kind of angle recommendations are there for a GTC-300 with a front windshield?

[STiPWRD]

For my part, I prefer to see CFD analysis done on close approximations of their destined home, that is, on a simulation which has a car ahead of it.

Agreed. The car and windshield will have a major effect on the air flow over the wing. This could explain the differences between on-track experience and the 2D model. I've done CFD analysis before and it can be useful in making design choices but real world testing is always invaluable.

[Wayne Presley]

I was comparing apples to apples, sixstar said on an R with no windshield. The 818R tested had no windshield.

[STiPWRD]

The images they posted look like the analysis was being done on just the spoiler - this assumes the spoiler is at the leading edge of the air flow. In reality, with or without a windshield, the front end of the car will affect the downstream flow hitting the spoiler. The simple 2D model of the spoiler has the advantage of quick computation times but for accuracy the front end of the car needs to be modeled as well.

[Bob_n_Cincy]

Sixstar,
This is cool stuff.
At 100 mph do you have the calculated drag and downforce numbers at the 3 angles?
Thanks
Bob

[SixStar]

We've been working with Zebulon since before we had the cars. They wanted to run kinematics on the R and the S as well as entire vehicle flow data but were never able to gain access to those files. So we're working with them and starting with the basics. We will have a ton of track and autox testing once the cars are done but it's fun to see what we can learn before then. And it's all at a reasonable rate for us since I went to college with the Neff brothers.

I'll try and get some force data from him tonight.

Thanks for the interest and I'm glad to see this generated some intelligent discourse!!

[Bill Waters]

Wayne, SixStar;

A few really basic questions - but I don't want to presume that I know something I don't.

I presume the reference aspect of the wing in setting AOA is simply the angle of a straight line extending from the wing's leading edge to the trailing edge - is that correct? If so, with respect to level, or to the angle of the car itself?

The shape and airflow dynamics of the wing are what they are - that is, it may induce downforce while set at zero AOA literally, but it makes more intuitive sense that, if the wing is set at zero degrees with respect to the car, and the car is set up at a few degrees of rake, then you have an effective few degrees of effective AOA at the wing. Which is it?

Thanks, guys

Bill

[Wayne Presley]

The way APR measures AOA is to measure between the leading and trailing edges in the center of the wing. The ends of the APR wing have a higher AOA than the center. AOA is measure vs ground level and is independent of the rake of the chassis.

[Santiago]

...the car seemed to tap out at 130mph even on sections of track that seemed to offer more room to run...do you know if that was intentional, rev-limited, or aero-limited?

Wayne, my question to you above may have gotten lost in the mix. I was really hoping you could chime in on any knowledge you have of whether or not the test mule was topping out at 130mph for some known reason.

Best,
-j

[Wayne Presley]

When I drove it it had 270 RWHP and we were coming onto the straight at 40 ish and we ran out out straight before we ran out of motor. It would be fun to take to Savannah where you come onto the front straight at 100 mph and the straight is 2750 feet long and see where it would top out.

[kabacj]

Regarding adding front downforce. The 818r at SEMA had added small diffusers to the block off plate under the nose.



I have the same situation in the GTM. It's easy to crank up the rear wing and add more downforce on the rear but front downforce is at a premium.

In addition to adding a bigger front splitter, front dive planes and wheel arch vents, I am going to add the diffusers too.

Wayne do you know any more about the front diffuser?

John

[Santiago]

I think those diffusers are just built into the standard front race splitter they offer. It's hard to tell from the SEMA pic above, but they look like early versions of these:



Best,
-j

[Xusia]

One of the techs told me that front downforce is increased if you move the front splitter forward (meaning, the splitter protrudes more from front).

[SixStar]

Without modification you can only put the very back edge against the bumper. You could always cut notches and shift it forward, but then you would have a gap between the splitter and the battery tray tin, if you filled that in you'd be fine.

[Doowop]

Thanks Keith. very cool

[Bill Waters]

The way APR measures AOA is to measure between the leading and trailing edges in the center of the wing. The ends of the APR wing have a higher AOA than the center. AOA is measure vs ground level and is independent of the rake of the chassis.

Thanks.

[Scargo]

Is this still the definitive "wing thread"? I was contemplating ordering my wing. I have a few I'm looking at and wonder if any further analysis or results from real-world testing has been done, especially since a year and a half has elapsed?
One thing I feel I am taking away from this discussion is that, for an 8:1 P/W ratio 818R, the wing that FFR offers (the APR GTC 300 universal fit wing) seems to be a one-size fits all wing offering for all their kits, that may be too much wing for the car, especially at faster courses. If it is enough wing at "0" to one degree AOA at 100 mph then it seems there is little to no room for tuning with it and that it might be a detrimental, higher than desired drag at higher speeds.

[Bob_n_Cincy]

Is this still the definitive "wing thread"? I was contemplating ordering my wing. I have a few I'm looking at and wonder if any further analysis or results from real-world testing has been done, especially since a year and a half has elapsed?
One thing I feel I am taking away from this discussion is that, for an 8:1 P/W ratio 818R, the wing that FFR offers (the APR GTC 300 universal fit wing) seems to be a one-size fits all wing offering for all their kits, that may be too much wing for the car, especially at faster courses. If it is enough wing at "0" to one degree AOA at 100 mph then it seems there is little to no room for tuning with it and that it might be a detrimental, higher than desired drag at higher speeds.

Hi Scargo,
This was Jim's comments after wind tunnel testing of the 818R.

"
With all this added front downforce we then adjusted the wing angle and added the diffuser to balance out the car. In the end our best numbers came with the extra radiator ducts taped off, the new splitter with the air gap taped off, 5 degrees of wing angle, the rear diffuser in place, no side skirts, and a level ride height of 3-3/8 inch. The max downforce produced was 120 lbs front and 164 lbs rear at 100mph, this increases to 271lbs front and 369 lbs rear downforce at 150mph. "

so Jim is saying 5 degrees was best in the wind tunnel. That should give you some adjustment range.
Bob

NOTE: this is 2 year old data.

[C.Plavan]

As an added benefit of data. I have the front canards, door vents, forward shifted splitter and about 7* AOA. Too much drag on straights, but car does not push in corners. I'm going to try 5* then maybe 3*.

[Santiago]

Wow, it’s been awhile since I’ve been around… Work, kids, life are all conspiring against my “free time.”

Anyway, I’m glad to see this thread revived. I would caution against taking away firm conclusions from the results so far. I don’t think the GTC-300 is “enough wing” for the car; it’s simply “the wrong wing” for the car.

I think we need to appreciate that the GTC-300 is a twisted element wing. When you look at the CFD results at an AOA of 0 deg, you are looking at composite results that add and subtract different values for the main span and the ends. The ends are 15 deg twisted from of the main, so they’re already operating at a much more aggressive AOA (fwiw, APR is pretty transparent about this in their own published literature).

So when Chad says he feels there’s too much drag on the straights, a good portion of that is coming from the ends which are nearing their stall point while the main span is still well within its range. APR even acknowledges that the free-stream limit of this wing is around 10-12 deg., where the ends are now at 25-27 deg AOA (while the middle is nowhere near its max operating AOA). A straight element wing would likely allow him/us to run a wing for the same amount of downforce with far less drag penalty. It doesn’t help that we also feed the wing “dirty air” coming off of the roll bar hoop…

Using a twisted element wing is a great strategy if the vehicle body invites the airflow at the rear (along the centerline of the car) to approach at a lower angle than the vertical free stream. A “fastback” body works well here, so the new 818 Coupe roof might take advantage of the twisted wing strategy.

Anyway, back to the 818R. What strikes me is that you look at Chad’s car, which has every FFR option available to improve the front, and he’s still only using 7 deg out of the wing. I thought I read in his build thread that when he broached 8-9 deg he was getting high-speed push. So compare this to the original results we saw before where the FFR mule used 5 degrees. With everything that FFR offers, you can only get another 2 deg out of the rear wing and stay balanced; and Chad is planning on dialing back even more (though that seems to be driven by drag-related concerns). That’s not terribly encouraging.

So we have a few different (but interrelated) concerns being raised:
1. Drag penalties for a given amount of downforce
2. Front downforce capacity
3. Aero Balance

To stick to the theme of this thread, the important point is that not all downforce production is equally efficient. The GTC-300 wing is less efficient in this application than a straight wing. More importantly (to me at least), the floor seems underutilized – there’s probably no better source of efficient downforce potential than a well-executed floor but we’re not seeing this in what FFR offers. So for a given HP level, there’s a range of downforce levels that could be successfully used depending on how efficient it’s produced.

As for the front capacity and aero balance, we need a new thread…and I need more time to finish my floor studies.

Best,
-j

[C.Plavan]

I agree with everything you said. A straight Gt-250 wing would probably be ideal with this car. I can totally see FFR saying "We can just use the GTM wing we offer since we already have it".

You can see in my videos, once I get over 118mph- it takes forever to go any faster. Granted, I'm having some engine issues, but down low it does not seem to have the drag it does on the top end.

Take this into consideration, I just got done racing my vintage 911 at Buttonwillow raceway this weekend. We were running CCW #25 configuration which is a high speed configuration. My car has 211 rear wheel HP, weighs around 150 pounds less than the 818. It also has no real aerodynamics. I was able to pull off a 2:00.2. This same config in the 818 a month or so ago, I managed 1:58. A 2:02 is considered screaming fast. Just food for thought.

[SixStar]

We're running a custom width (67.25") wing from Zebulon, the company that did this analysis. It's not twisted (flat) and we run it at 0 AOA. So far it's done great, once the front end is finished we can add a degree or so and balance it out.

StROcMd.jpg


Ryan is racing at Solo Nats this week but I'll get the data on this wing ASAP.

[longislandwrx]

How wide is the FFR 350 wing? is 67 the best bet?

[SixStar]

We chose a 67" because it's almost the exact width of the 818.

Got a hold of Ryan @ Zebulon, he'll get us some flow data on his wing next week, for now he has this to add:

Car shape is obviously extremely important to wing performance - these 2d simulations were just to show what the GT300 profile does at different AoA's relative to the approaching airflow itself - the car will change the effective AoA by altering the airflow as it approaches the wing. This means that there is no end-all-be-all perfect solution; however, on a convertible or on the R, your safest bet is a straight profile wing. Especially one that is designed to work in "dirty" airflow. Forgetting about the Anhedral (twisted) design, The GTC300 is far too aggressive to work well in this dirty airflow condition. It just is not the correct wing for this application.