4 ms·
Recently I've been doing a teeny tiny bit of CFD again. You might be amazed at how little CFD goes in to wings. It's still a huge amount, but tunnel testing sti
by bfieidhbrjr 7y ago
Recently I've been doing a teeny tiny bit of CFD again. You might be amazed at how little CFD goes in to wings. It's still a huge amount, but tunnel testing still wins. I was kind of surprised.
As best I can tell, CFD is still kind of like impedance-based home body fat scales. They're kind of directionally accurate but have a pretty high error.
I thought CFD would be a solved thing by now, but then I remembered how good (e.g. bad) weather prediction still is.
- LeifCarrotson 7y agoYou might be interested in this bet on the LongNow site: http://longbets.org/753/ http://longbets.org/753/ Some people think that wind tunnels will be obsoleted by computational fluid dynamics within the decade. Others think that the problems with CFD are not a lack of compute power but a lack of comprehensive understanding and modeling of the physics at play, and expect that wind tunnels will be with us for quite a while. (Personally, I believe in the progress of technology and expect an eventual if not near-term breakthrough in CFD software will eventually obsolete the wind tunnel, but I also think that institutional inertia will keep a lot of wind tunnels and businesses that purchase time in them in operation for a decade or more after CFD's big breakthrough...)
- Gravityloss 7y agoThere are sources of error in wind tunnel testing too.
- pimlottc 7y agoFor those who don’t know (like me), CFD stands for “computational fluid dynamics”. https://en.m.wikipedia.org/wiki/Computational_fluid_dynamics https://en.m.wikipedia.org/wiki/Computational_fluid_dynamics
- mhh__ 7y agoF1 teams spend a lot of time in wind tunnels looking for correlation between CFD and Reality. If there isn't any then the year is almost lost, usually.