Interesting video from "The Race":
Why Mercedes has got F1's current era so wrong so farThey went through the same riff on a recent podcast, and it's in a long article on their web site:
https://the-race.com/formula-1/why-merc ... ng-so-far/The video is good for visualizing what they are saying, and the article gives detailed text so you can follow along.
The explanation makes sense to me as a non-expert, but the thing that sticks with me is:
A key part of Mercedes’ downfall was believing numbers in the virtual world that couldn’t be replicated in real-world conditions – something visible from its brutal first shakedown at Silverstone in February 2022.
“Mercedes’ simulation gave sensational figures,” Hughes continued.
“Had the W13 worked the way simulation suggested it would, Red Bull and Ferrari would have been left for dust – it was that clear.
“Mercedes was going into that season believing it was going to have an enormous advantage and there was an exciting buzz leaking out about just how good this car was before it ran.
“The real world inconveniently got in the way of making those simulations accurate.”
I think I'd state it more directly: Merc's simulations did
not model the real world, yet Merc kept believing its simulations for two seasons in a row.
The article tries to address the question of why they kept up with the concept for a second season:
“Still with those sensational windtunnel simulation figures tantalising them, for this year’s W14, Mercedes decided to keep that basic aero philosophy with a redesigned rear suspension and gearbox, which gave it the appropriate range that it didn’t need to be so stiff and it made it more driveable and also helped by the anti-porpoising floor regulation tweaks,” Hughes explained of Mercedes’ logic.
“They could finally make that concept work as intended but it turns out that those first-order porpoising and bouncing problems that they had last year had simply hidden a more subtle limitation.
“The aero concept just isn’t as effective as a classic undercut. The exposed floor and the wide sis bar just did not give adequate compensation for the lack of that undercut channel, especially when measured against the Red Bull, which has made massive gains thanks to the way it has configured its suspension to limit pitch and dive.
“So the Mercedes now seems to run out of airflow capacity at high speed relative to the Red Bull so it’s down on high-speed downforce, which is the opposite of last year.
“Although it’s now competitive with Ferrari on straightline speed, it’s well behind the Red Bull and its spectacular DRS stall.
So it seems Merc talked themselves into believing that the simulation was still overall accurate, yet it didn't show the issue with porposing. Then they talked themselves into thinking if they got rid of porposing/bouncing they'd find all those colossal gains and blow past RB and Ferrari. Yet now after two years we're being told the simulation has been wrong all along, and the "classic undercut" approach was and is better.
Again, without being an expert, this to me seems like a over-reliance on the tools, and insufficient vetting of the tools. We see they run the car around the track with all those instruments that can map the real-world airflows, didn't those things show them they were barking up the wrong tree?
I have used a lot of tools over the years to model things in the real world, but always have some way of comparing the models to the real world. When they don't agree, I pretty quickly decide the tools are broken. Again, as a non-expert in this field, it seems they've put way too much faith into their simulations.
I wish the article/podcast had addressed these aspects more directly.
