I'm with McLaren on this. CFD plus wind tunnel is the way to go. Those teams who are wholly behind CFD are just that way because they can't afford a wind tunnel. Eventually, a wholly CFD approach may work, but we're not there yet.

cfd is a tool that can be used, but to rely on it completely (like manor) is silly. nothing can replicate actual track time. why do you think that so many cars were using flow paint, different wings , different exhausts etc? because the computers could not replicate actual running conditions.

i agree with Neale but it's only a matter of time. once they have enough real-world data from test days, under all conditions to feed back into the computational model then the wind tunnel will become more and more an unnecessary expense.

I can see that happening in the next four to five years if we have fairly stable aero regs.

CFD is the future, no denying that. It's a rather substantial jump from the traditional wind tunnel and may be the lack of a middle generation concept is what troubles McLaren, which is completely understandable.

@Marcus Garvey
It's not about collecting data from test days, it's about the computational complexity of it all. Not even NASA-grade CFD is powerful enough to model air flow in enough detail, at least in a viable amount of time.

@James Smith-Brown... and you think in four to five years CPU processing power, memory access time and computer architecture will not have improved?

@Marcus Garvey
Of course they will improve, I never said otherwise (and it certainly wasn't your initial point).
But even then, a huge increase in power simply isn't enough, I'm not sure if you appreciate quite how computationally difficult all of this is.

If supercomputers in 4-5 years can simulate this stuff accurately enough that teams forget about windtunnels then we should all be very worried- those computers would be able to crack any security on our financial transactions within hours...

Virgin is at the BACK off the grid ---No wind tunnel in sight

Kinda like running test in a lab. A bunch different from the real world.

@James Smith-Brown
i'm fully aware of the issues surrounding computational modelling it's part of my job. and i still stand by my original prediction that it's just a matter of time.