Will Gray

End of the wind tunnel?

Wed Oct 21 11:06AM

Manor F1 designer Nick Wirth has revealed the new team’s car will be the first Grand Prix machine designed entirely using Computational Fluid Dynamics – but can it prove wind tunnels are a thing of the past?

Wirth confirmed last week that the new Manor F1 machine, which Wirth Research and sister company Digital Flow Solutions have been hired to develop, will be completed by the end of next month - and in doing so he confidently revealed the car will hit the track having never been near a wind tunnel.

The general long-held consensus of designers up and down the paddock, however, is that while CFD shows significant cost-effective potential and is moving forward at a rapid pace, there are still several years before a competitive car can be developed with confidence using solely CFD - and many believe taking that route could end in disaster.

Aerodynamics has long been one of the major performance differentiators in F1, but development has always required teams to create scale models to test aero modifications in wind tunnels week-in, week-out, costing millions of pounds in the process.

CFD instead uses computer programmes with mathematical equations to model how the air flows across the car. Although there is still significant costs in computing power and hardware, it’s cheaper to run tests and can also better measure the effect of a car in a ‘true corner’ and include multiple parameters such as fluid flow, heat transfer and fluid-structure interactions which, because they run cold, wind tunnels cannot.

The CFD technique has always been limited by computing power, with even the best supercomputers unable to cope with the amount of mathematical processing required. But things are changing and new methods of hooking up servers along with new hardware technologies are rapidly increasing the capacity of these supercomputers to cope with the intense and fast-moving world of Formula One.

Now, calculations that literally used to take days can be completed in a matter of hours.

Renault recently invested in a major new CFD centre rather than building a second wind tunnel, and confidently predicted that the CFD testing will deliver 50 per cent of the performance gains developed by using a conventional wind tunnel at half the cost. But they still wouldn’t trust it without testing those gains in the wind tunnel after going through the CFD work.

The nature of mathematical models is that they are never perfect – and so most teams believe that it is essential that results are cross-checked with a wind-tunnel model. That said, regulation-imposed restrictions on running time available in wind-tunnels and the cost required to develop the large models now used by most teams means going straight from CFD to full-size vehicle may be a viable alternative. But will the CFD results be good enough?

The aerospace industry relies a lot on CFD, but that has relatively simple geometry compared to racing cars. Wirth did develop the Acura LMP1 cars using only CFD, with great success, but again they are less complex than Formula One cars, particularly as they have covered wheels – and understanding flow interaction from open wheels is acknowledged as one of the most complex aspects of Formula One design.

There is no doubt that CFD has its strength in conceptual development, allowing designers to make quick major modifications - like altering wheelbase significantly - to get a baseline design understanding within a matter of hours, but many believe the complex detail on an F1 car is still out of reach for CFD modelling.

It’s a brave move, then, for the newcomers to attempt to avoid a wind tunnel altogether – although it could also be a gamble born of necessity rather than choice.

The new teams were enticed into the sport by the promise of a budget cap that never saw the light of day and while efforts are being made to reduce costs, it would still take a great amount of money to compete with the current teams on the same playing field.

Slashing all wind tunnel costs will certainly help Manor’s budget and if Wirth gets the CFD route right it could overcome the ‘newcomer’ disadvantage.

But if it goes the other way it could prove the CFD-only development concept is still a little ahead of its time.

Comments1 - 21 of 21

What happened to the font on this one? Almost impossibe to read
From enigmard on Wed Oct 21 12:14PM
Report abuse
Did I read this right:-
CFD testing will deliver 50 per cent of the performance gains developed by using a conventional wind tunnel at half the cost
So I can pay half as much and receive half as much gain?
Why don't I just run my wind tunnel half as much instead?
From snbudd on Wed Oct 21 12:38PM
Report abuse
lol...really? I found a place that many players are hoo king up with h ot mo dels, seems the club called: __Tallconnect Co M___, do you hear this before?
From samsos27 on Wed Oct 21 12:45PM
Report abuse
Good luck to all the new teams I say in whatever method they choose to design their new cars, might just get lucky you never know
From kga on Wed Oct 21 12:53PM
Report abuse
This is brilliant! Why don't we just do everything with computer processing? Let's let the driver do races with computer cars on computer tracks in virtual rooms and then we can create virtual fans to watch the race..that will be exciting! Think of it as progress.
From mickeygv8 on Wed Oct 21 12:54PM
Report abuse
The Acura LMP car was done entirely with CFD design and it worked pretty well, but still ended up with some issues. Not sure how it'd work out in F1 but it does have great potential.
From superjohn_rtp on Wed Oct 21 12:55PM
Report abuse
Most people get there Yorkshire Puds ready made from the super market why are Manor wasting all this time and money on CFD ?
From arttidesco on Wed Oct 21 01:17PM
Report abuse
is it not taking it to far look what happend to concord i think we should leave well alone man can only go so fast i think
From tobykaly on Wed Oct 21 02:11PM
Report abuse
This is a bold step, but that is the good thing about start up teams they come in with new ideas. I hope it works for them, and I hope the USF1 skunk works approce works too. It will be intresting to see who gets the narrow front tyres and larger fuel tanks right. RedBull, Mclaren, Ferrari, and Brawn should be real close next year.
From mcshambus on Wed Oct 21 03:42PM
Report abuse
well all f1 cars look the same, so they should just share a xchassis and reduce costs by 90%
From kcrally on Wed Oct 21 03:46PM
Report abuse
Come and meet me on mycams.asia -Lisa
From johnnysalaza on Wed Oct 21 07:31PM
Report abuse
i hope lotus wud do well next year
From honeybadger_17 on Wed Oct 21 08:29PM
Report abuse
If it was up to me I'd ban the windtunnels altogether, and much reduce the aero on the cars, so that we can get back to mechanical grip and better racing.
Cheers from Vermont.
From oatisbeansworthe on Wed Oct 21 09:50PM
Report abuse
It would be interesting to see the difference in CFD data and wind tunnel data for the same car. Then compare actual data (car on track) with the CFD and wind tunnel data.
From jdhanssen on Wed Oct 21 09:59PM
Report abuse
OH No...computer based design... I bet McLaren is searching the globa right now for the best hackers. SPYGATE II I can see it now.
From daycosucks on Wed Oct 21 10:10PM
Report abuse
No amount of computer modeling can substitute time in a wind tunel, sorry there are just too many variables and figures that would be at best just finger in the air estimates. Would be a good starting point before going in but other than that.....
From paulhancock1123 on Thu Oct 22 08:15AM
Report abuse
good for manor, they've been very sucessful in all the different racing championships they've entered I hope they do well, (but then they are only down the road from me so i'm biased) but surely if everyone went to CFD the computer would spit out the same info... all cars would look the same... :S
From jamessilvester on Thu Oct 22 08:37AM
Report abuse
quote "Why don't I just run my wind tunnel half as much instead?"
and do it in twice the time? CFD is as fast as computers and with clowd computing and modern processors that is more attractive. In the future this will inevitably becaome so fast that they will be able to try multiple simulations in the time it takes to make a model (1/3 scale?), run the tests, and analyse the data. Wind tunnels are a simulation after all. they even try running the wheels to make it more realistic. But like some drivers they have difficulty with the turns.................
CFD is the coming technknowlogy (sic) it is not a question of if, but when it becomes the first choice of design. The track is the the ultimate choice.
From mrred on Thu Oct 22 10:10AM
Report abuse
At last a decent discussion topic !! I'm also unsure as to whether CFD will completely replace traditional wind tunnel work, since the aerodynamics in play on an F1 car are so complex that they HAVE to be proven in the real world, and you can't build driver preference/ technique/ driving style into either work, or the exact influence of track layout or weather conditions either ..... so ultimately the proof is on the track with feedback from the man at the wheel (if he's capable) - but without good mechanical grip from the chassis, the aerowork is anyway redundant ..... but for sure it will make for another interesting comparison between teams once pre-season testing gets underway
From t4tigger1977 on Thu Oct 22 12:53PM
Report abuse
mrred, clowd computing??? you mean cloud computing and what has that got to do with this??
From paulhancock1123 on Thu Oct 22 02:48PM
Report abuse
It certainly should not come as a surprise that eventually all F1 cars will look identical....when you find the optimum shape....everyone will follow it.
It is the engines and gearing and chassis/suspension variations which will provide the extra edge as teams try to utilise the best of all these for each track variation....
So...unless the FIA demand everyone uses the same engine and chassis (dimensions,etc)....certain teams will still be able to do better than others...
and then ....of course....the driver plays a part....I believe.
From geoffthebull on Fri Oct 23 05:44PM
Report abuse