Verus-Eric

300mm Chord, 1800mm Span, Rear Wing in Tunnel

Verus Wind Tunnel Results

A few months ago, Verus went to the wind tunnel with the goal of validating our CFD analysis. We tested at Auto Research Center Indianapolis (ARC Indy) using their wind tunnel for full-scale wing testing. Two of our rear wings were tested, along with 4 other wings from various other small businesses/privateers, using fixtures designed and built in house. Before releasing our 987 and 981 GT4 rear wing products; we wanted to share the results and findings from the wind tunnel.

Why Does Validation Matter?
Validating our CFD to wind tunnel data ensures that our test procedures within CFD are accurate and reasonable. CFD can be run incorrectly, easily, and we wanted to make certain that our results within CFD estimate real world numbers.

300mm Chord Wing in CFD on ARC Sting

How Did We Run CFD to Compare to the Wind Tunnel

Unfortunately, it is physically impossible to test a rear wing in free-stream, which is how we developed our wing profile. The wing needs a way to be held up in the wind tunnel; so to correlate our results from CFD to the wind tunnel; we had to model and CFD the entire sting at the correct height off the tunnel floor. The wind tunnel tests were run at ~80 MPH or 36.5 meters per second.

Mesh generation, analysis, and post-processing were handled by ANSYS. The mesh for this run was a bit over 40 million cells. The wing had 20 prism layers with a first cell height of 0.005mm to capture the viscous sublayer. We ran the wing through our main (3) turbulence models, K-Omega SST, K-Epsilon Realizable, and Spalart-Allmaras. We primarily use K-Omega SST in our analysis; but, presented a chance to improve our processes, we tested all three.

Verification and Validation - Downforce

The graph below shows downforce for the experimental data (wind tunnel) compared to our CFD with different turbulence models applied. The analysis using the K-omega SST turbulence model was the closest to the wind tunnel results. Comparing this data; our CFD analysis underestimated downforce by 3.4lbs at 80 MPH; 7.5lbs at 120 MPH, and 13.4lbs at 160 MPH.


Verus Validation - Downforce

Verification and Validation – Drag

The graph below shows drag for the experimental data (wind tunnel) compared to our CFD with the same turbulence models above. The analysis using the K-Omega SST turbulence model was again the closest to the wind tunnel results. Comparing this data, our CFD analysis underestimated drag by 0.7lbs at 80MPH, 1.6lbs at 120 MPH, and 2.8lbs at 160 MPH.


Verus Validation - Drag

Verification and Validation - % Difference

Verus Validation Chart

Conclusion

With under a 2% difference from our CFD analysis to the wind tunnel; we are happy with the results correlating so well. This really helped us solidify our analysis procedure and ensure what we do for each analysis is indeed correct and correlates to real-world results.

Please let us know if you have any questions or comments. We appreciate any and all feedback.

RennKit-Dave

Interesting write up and presentation. Your write up states the physical test at one wind speed but the graph seems to show it at several speeds for comparison to your various model calculations, can you explain this?

Verus-Paul

With coefficients such as cD and cL, forces can be extrapolated using Force = 0.5 * density * velocity^2 * Area * coefficient. We did the graph like this to show the force error will increase as velocity increases.

Note that this is just an extrapolation. Sometimes parts follow this well and sometimes they do not. It depends how sensitive the parts are to the Reynolds Number of the test speeds. Before getting too technical, when we tested the wing in CFD, it did not show this sensitivity at the speeds a vehicle will see. I can go more in depth on this type of explanation if others would like, but not sure if it is wanted/needed. A general rule of thumb is that this type of extrapolation is perfectly valid for most of our aerodynamic work.

lovetoturn

Did you test the stock GT4 wing, and if so, can you show a graph that compares the two wings. Curious to see what the differences is between yours and the OEM part under the exact same conditions. Thanks in advance if this is possible.

kiznarsh

As a former wind tunnel test engineer (and still heavily involved with it), nice to see some real data!

Seeing more downforce in that ideal scenario makes sense without the separated air coming off the roof on the actual car, but still very good comparisons. Any plans for repeating the test with a mockup of the car in front of the wing to better characterize its performance?

Also how close are you to the walls?

Verus-Paul

Yes, we tested this in CFD, but not in the wind tunnel. The graphs below compare our wing on car (downforce of the wing only) to the factory GT4 wing on car (downforce of the wing only). Let us know if that does not make sense or if we need to clarify further and we will do so.


Downforce Graph

Drag Graph


No plans currently for a full-car wind tunnel test. The cost to benefit is not there for us yet. Most tunnels we see used would be inadequate for our needs. We would need a full-scale wind tunnel with decent blockage ratio, it needs a rolling road to capture the tire/wheel interaction with the underbody, and needs to have a height able to handle decent wakes off the back of the car with a good wing. The only tunnel remotely close that meets the requirements is wind shear which is currently out of our budget.

We are not close to the walls as it is an open jet. Our blockage ratio was right at 0.1 for max angle of attack, including the test sting.

Reborn996

Nice data, and the GT4 stock wing reference is even in my configuration with Salter Aero riser and Gurney lip. The down force numbers for the Versus wing are crazy high as AOA is adjusted! I know Porsche reinforced the hatch to support the factory wing, do we know the structural limits of what down-force the hatch can successfully support? Also the risers?

RennKit-Dave

The GT4 wing has a variable AOA correct?

Reborn996

Yes, but only two settings.

kiznarsh

Yeah testing isn't cheap, but vital for verifying CFD results. Is that a boundary layer suction plate on the floor?

lovetoturn

Thank you for the comparison. This just puts everything in perspective as to how much down force your wing can generate. Quite amazing.

Dicknose

Now to see how to fit one to a Spyder!!

Soul Performance

Very cool, Eric. Keep up the great data driven work!

-Mike

lovetoturn

One more question. Is that the low or high position for the stock wing. Somewhere else it is stated that the low bolt is 5.3 degrees and the high bolt hole is 7.6 degrees. Which ever one that you did, can the other be easily calculated? Your wing creates so much more downforce.

Verus-Eric

We have had a few customers out running our rear wing at 9 degrees AOA at this point without issues on both 987 and 981 applications. The rear hatch seems to be made quite well on the GT4 which is great as installing the rear wing on a GT4 is quite painless and bolts on like OE. Below is a picture of Jeremy @Pr3z running at 9 degrees AOA and Aleks @doba_s running at 4 degrees AOA. Our shop car runs 6 degrees currently.


Jeremy at 9 degrees AOA


Aleks at 4 Degrees AOA


Shop 987 at 6 degrees AOA


Testing is most certainly not cheap. We will continue to verify our CFD results with wind tunnels and track testing (we have done this with our 987 kit and laser ride height sensors). We have also done coast down testing, flow vis testing, tuft testing, etc; all in the name to verify and improve our CFD as well as learn. Hopefully, the opportunity to go to a full-scale, rolling road, wind tunnel presents itself someday as that would be quite an experience. Yes, the ARC wind tunnel has boundary layer suction in place. The tunnel is scaled but very nice and even has a rolling road, which we obviously did not use. ARC provides support for IndyCar, NASCAR, and various other motorsports.






Thank you!

Thanks Mike; we'll do our best.

I would guess it is the lower. We estimated 6 degrees AOA on the factory wing based off of CAD. We do not have the car here with Salter risers so we cannot say for sure but 6 degrees is closer to 5.3 than 7.6. The other cannot be easily calculated without running CFD. Our wing is quite a bit larger than the factory unit and has a more aerodynamic profile; which is ultimately the reason for the improved performance.