Is CFD Flawed?

There is a high profile debate going on in the motor racing media and motor racing discussion forums that Computational Fluid Dynamics (CFD) is flawed. The proof cited is the poor performance of Virgin Racing (VR) Formula 1 racecars in the 2010 F1 championship and repeated so far in this year's championship. Why is CFD deemed responsible for this sorry state of affairs? Because VR have proclaimed that the aerodynamic design of its cars forgoes wind tunnel testing and only uses CFD. What if the CFD-only strategy is merely a symptom rather than the cause of the problems at VR?

Open Wheel Race Car CFD SimulationIs CFD Flawed?

Background

VR entered F1 in 2010 with the stated aim of working within a self-imposed budget cap (around US$60 Million), whereas other established teams continued with their high spending ways - reputed to be in excess of US$400 Million in one case. To be fair the provisional F1 rules issued in 2009 for the 2010 season included a budget cap which Manor Grand Prix Racing (later becoming VR) signed up for. However, money does indeed talk and no less so in F1 - the big teams vetoed the budget cap. This left VR with a dilemma - withdraw or continue with a puny budget. It chose the later.

CFD or Wind Tunnel

Like any cost-conscious company concerned with aerodynamic design VR looked to CFD as a cost-effective method to design their racecars. Where they departed from the F1 norm is that they decided to use only CFD for aerodynamic design, and forgo wind tunnel testing completely. Most F1 teams (and certainly the successful ones) have dedicated wind tunnel facilities that come with hefty building and running costs. Also all F1 teams have dedicated High Performance Computing (HPC) centers for running CFD. A traditional F1 team is the perfect model for CFD working in complimentary harmony with wind tunnel testing. Now in VR's position what would you do? Spend your entire budget on building a wind tunnel with nothing left for anything else? Or forego the wind tunnel and bulk up your CFD capabilities leaving enough money to go racing? Exactly, so with its CFD-only strategy and restricted budget VR went racing.

Results

The 2010 F1 season was not kind to VR, resulting in no championship points and many retirements. A mid-season redesign was needed after it turned out that the racecar's fuel tank was too small to finish races under full power. So far the story of the 2011 season has followed a similar script with the team struggling to qualify within the re-introduced 107% cutoff relative to the fastest qualifier.

Cause = Lack of Money, Symptom = CFD Only

It's easy to blame CFD for all the troubles of VR. But surely its CFD-only strategy is merely a symptom of the real problem: money or more accurately a lack thereof. It appears, over the long run, the better funded sports teams tend to be successful and conversely the poorer funded teams tend to struggle. In F1 that means the better funded teams have the best facilities, employ the best engineers, attract the best drivers, and so on. I believe VR is simply another example of this fundamental law linking sports finance to results.

Comments

VR F1

I don't know the specifics of VR F1's situation, but in general, and in my opinion, the biggest issues are a lack of aerodynamic experience, lack of validation, and the validation that is done is done on complex configurations. Sometimes there seems to be a lot of arm waving. Without experience, both wind tunnel and CFD results can lead one astray. In the right hands, a CFD code can do wonders. The key is to know where the CFD results are uncertain, and the interactions and interferences which amplify those CFD uncertainties.

Martin Hegedus

CFD Smarts

Yes, I agree a lack of aerodynamic understanding/experience and a lack of model validation whether it is for CFD analysis or wind tunnel testing would be a handicap for a racecar team.

However, I don't think the question centers on the level of CFD experience or even the strategy of only using CFD for aerodynamics evaluation in VR F1. Nick Wirth, technical director of VR F1, has seen notable successes in applying CFD to racecars at Wirth Research. I believe Wirth and his CFD engineers have the talents to know how to best use CFD, but that doesn't compensate for the general lack of funding at VR F1.

VR F1

Not sure if I totally understand. You state that the issue was not about CFD (and I assume you also mean aerodynamic) experience or about using CFD for evaluation of VR F1. Did poor aerodynamic prediction and/or misunderstanding of the uncertainty of those predictions play a part in the issue or was it just a general lack of funding handicapping the overall design and testing process?

Also, do you have recommendations for papers/articles/books that compare CFD results to experimental F1/racing car data and/or offers advice on best practices for using CFD for such analysis?

Lack of Funding

I think it's a general lack of funding handicapping the overall design and testing process.

Rigorous comparisons between CFD results and experimental F1/racing car data if it exists in actual teams will be confidential. You might have more luck approaching racecar CFD consultants such as Total Sim or Wirth Research.

Data

Thanks. What I am trying to find is data for basic configurations. Such as 1) spinning wheel with and with out ground plane, 2) spinning wheel in front of another, with and with out ground ground plane, 3) body without wheels with moving ground plane, 4) etc. Data that can be used to verify that CFD captures the basics at the component level. The key aspect, and this makes the search difficult, is to find data which people have confidence in. Frequently (depending on the source of the data), the data is contaminated over a portion of the test matrix by flow physics that the experimenter was unaware of and therefore not documented in the WT report. Such WT data can be misleading when comparing to CFD.

If you come up with any ideas, please let me know. I'll also try contacting Total Sim and Wirth Research.

Spinning wheels & moving ground planes

Martin, as mentioned above most of the info you seek will be proprietary! I do not know your level of expertise or experience but here are some suggestions: [In order] Race Car Aerodynamics: Designing for Speed by Joseph Katz; Competition Car Aerodynamics by Simon McBeath; Aerodynamics of Road Vehicles 4th Ed by Hucho et al. These will give you a feel for the directions if not the exact magnitudes - everyone forgets that there are no exact answers! Also peruse the www.sae.org site for pertinent CFD aero papers. Good luck!

The Saga Continues

For an update on the latest developments in the Virgin Racing F1 saga try "New Strategy = CFD + Wind Tunnel".