Quote:
Originally Posted by Casper
I am still amazed that Honda did not run the motor in a test mule before it was an official F1 homologated PU. The thinking behind that would be an interesting read.
|
A test mule may or may not have helped. These days, I think, that on track testing is more about correlation and validation of your prior test regime and less about being the primary way to expose issues.
What engineers need is a way to do very controlled testing BEFORE showing up on track (either at a race or a test session). Find issues before putting in the real car (or a mule). Engine dyno testing is an example. But, the dynamics of the solution is more complex these days. That is why most everyone do much more than just engine dyno testing, or even engine dyno testing that try to simulate g-loads. That "something" is a chassis dyno (such as what AVL provides). It is an absolute requirement these days. This combines everything (engine, gearbox, drive shafts, suspension) together and tests the whole integrated unit. If you are able to get good correlation between your internal testing regime (engine dyno, chassis dyno, computer simulation of the power unit, etc.) with the actual on-track performance, then you can do powertrain development nearly 24/7. This works because you can trust your test regime to provide reliable data. If the data tells you that the performance and reliability is there, then it probably is.
I haven't looked at the regulations, but is the CFD limits just on aero development? If so, power unit manufactures could run as much computer simulation on the engines as they wanted as well as nearly unlimited physical testing (dyno, not track testing). It would just be a factor of how large your budget is.
The HUGE miss by Honda is that they seemed to rely purely upon a combination of things like single cylinder test rigs to validate combustion concepts, classic engine dynos and probably computer simulations. They had no capabilities for chassis dynos that would fully integrate the entire powertrain. So they integrated it all together at the very last minute thinking they had it all covered.
I can imagine that McLaren probably had a combination of CAD files, expected performance data plus some type of boilerplate engine mockup that allowed them to design and test fit stuff everything else (gearbox, plumbing, etc.). And then as the season start approached, Honda shows up with a real engine, the put it all together, fire it up, do some quick shake down testing, put it on the truck and drive it to the first test session.
Then they found they had very poor correlation between their pre-integration testing and actual performance. Something was screwing up their combustion process, they had weird power train vibrations/oscillations that didn't show up until on track (those two issue may be related), the had unexpected lubrication issues, reliability issues on various stuff, etc. There was a long list of stuff that just didn't show up on the dyno back in Japan.
Eventually they figured out (probably from external experts) that how they were testing was wrong. I don't know how long it took for Honda to realize this, but then they had to both fix their process and fix the engine at the same time. Without being able to replicate issues on dynos, they had to rely upon forensic analysis to figure out what was going on. Even then, that may say "what" is happening, but not "why". Too many issues to resolve and not enough time. So that was a recipe for a downward spiral in the relationship between Honda and McLaren.
While the slow motion dissolution of the Honda/McLaren relationship was going on, Honda was likely figuring out how to address their testing issues. This probably accelerated once they had an agreement with Red Bull. They probably saw the light by then, and then Red Bull was able to offer additional help and Honda was probably more than willing at that point.
So... I tend to think that a test mule "might" have exposed some of this, but using a test mule is the wrong strategy these days. While some of the rules make it hard to use mules, the real reason is that there are better solutions. So that is why nobody else does it. The real miss by Honda was getting back into the game without having testing capability parity with the other manufactures.
There remains a number of untold stories that I hope come out some day.
Quote:
Originally Posted by GingerPixel
Ferrari ran a La Ferrari with a big wing and a roof scoop, but I'm not ever sure it was confirmed what they were testing.
|
I kept thinking it was F1 related, and I think there is a narrow way this can be legally done, but Ferrari says "no" and I think I have seen a photo of that car on display somewhere with specific details as to what it was and those details didn't match the F1 specs. But who knows if that was accurate or not.
Quote:
Originally Posted by steve_r
McLaren have paid the pain, and RBR will get the gain.
|
Yes
Quote:
Originally Posted by steve_r
If Honda engines start going bang or they lose out on straight line speed at some tracks this year, it will be very interesting to compare RBR's reaction, bearing in mind the silly finger pointing against Renault of recent years.
|
Yes again!
Quote:
Originally Posted by chillibowl
i just hope Honda dont screw it up by thinking they can go full works again!
|
I agree. I suspect the current scars run very deep. That Honda has learned a very hard lesson. I can only see them owning a team in the distant future and the regulations have resulted in radically simpler cars. To my point earlier... If we are starting to see hints of power unit parity, then we can also conclude that there remains plenty of secret sauce to making a fast car beyond having a good engine. And so far... nobody seems to be better than Mercedes at both. And that is not for a lack of trying by their opponents.
Richard