PhillipM wrote: ↑20 Feb 2019, 20:31
TimmTurbo wrote: ↑20 Feb 2019, 18:54
holding back because of internal reasons I agree but whenever there is components available which makes you faster it is a no-brainer to put them on right away in order to verify data with reality.
No, it isn't - if some designer says "look, I've found half a tenth with this bit, and it works with everything on the car" - you don't bolt that part straight on the car. You look at how much time and effort it's taken to refine that part and how much more it might take for more optimisations vs the runtime in simulation/windtunnel/durability testings.
If you can get another half a tenth out of it with more work, and you think your car is already fast enough, you'll leave it and let them optimise it more before you run it, whilst your production works on parts that are already far enough into diminishing returns that they're not worth more time and effort.
You'll probably have quite a few parts that show an improvement but are also so sensitive that they sometimes work in sims and sometimes don't, or they're failing your durability sims, if your car is fast, you'll let them have more development time on them to try to mitigate the risks more. If you
know your opponent is faster, you'll stick them on the car and risk it.
I don't know why some of you seem to be believe you just come up with an idea for a part and make it, development works nothing like that, it's a continous, massively iterative process.
First of all the process is different, the designers typically have a given target given by CFD, Manufacuring, or Validation dept. when they touch on a certain component. They don't just sit around and brainstorm. Nobody thinks that the car is fast enough, and that's exactly my point.
I fully agree with you referring to this process, but the initial statement was there is no sense behind sandbagging or hesitating putting something good on the car asap.
I know how hard it is to get a concept to pass internal gates as im an Engineer myself. We don't use CFD but FEA. We run several iterations always under extreme time pressure 24/7. My corporation spends millions / year in improving simulations (FEA) but now matter how hard we try, at the end one thing is for sure, you can be close but never reach 100%.
