Exploding physics with stiff suspensions

Associat0r · Jul 7, 2016

So I just tried the RSR Dallara F3 which is supposed to have physics from real data.
But when I increase the rear wheel rates in combination with high fast rebound dampers in the garage, the physics engine sadly explodes.
https://i.gyazo.com/c9b4289203f11227aefdc7754ae950f3.mp4

EDIT: See here on how to reproduce it http://www.racedepartment.com/threads/exploding-physics-with-stiff-suspensions.123651/#post-2269476

Radu Oros · Jul 7, 2016

I increased the rear wheel rates to max, then I increased fst rebound dampers to max. No crashing and I could drive the car. Did you change anything else on the setup? Can you share it.

ZethGAF · Jul 7, 2016

Exposed!

Associat0r · Jul 7, 2016

You're right, I changed other stuff too, but mistakenly thought the rear fast rebound was the main instigator, since they responded the most in my case. It's a combination of multiple things.
To reproduce it guaranteed, stiffen all the rear dampers, wheel rates and ARB, then work your down.
Stiffening the front seems to have an effect too, even when the rear isn't stiffened to max. Sometimes it seems almost fine in the pits, but when you drive a few tens or hundreds of meters it starts to explode.

Michael Hornbuckle · Jul 7, 2016

So you made completely unrealistic changes to a mod car until you got the game to act weird. Good job, I smell a reader submission brewing.

Associat0r · Jul 7, 2016

Michael Hornbuckle said:
So you made completely unrealistic changes to a mod car until you got the game to act weird. Good job, I smell a reader submission brewing.

I didn't make any unrealistic changes. These changes are done within the garage within the spec of the real car.

Radu Oros · Jul 7, 2016

Ultra stiffening everything together can be too much, however when you stiffen to max a parameter in the Dampers or a couple of them, it still works, but then everything together the software may not be prepared for it.
Is putting everything to max realistic to drive/race the car in those conditions? So maybe the software has boundaries with how much it can work or in how much realistic conditions the car could be. When you create the car you can write higher numbers than recommended. Perhaps in real life when you setup the components, to stress them out so much won't make the car race-able or will wear out the components much faster.

Associat0r · Jul 7, 2016

Radu Oros said:
Ultra stiffening everything together can be too much, however when you stiffen to max a parameter in the Dampers or a couple of them, it still works, but then everything together the software may not be prepared for it.
Is putting everything to max realistic to drive/race the car in those conditions? So maybe the software has boundaries with how much it can work or in how much realistic conditions the car could be. When you create the car you can write higher numbers than recommended. Perhaps in real life when you setup the components, to stress them out so much won't make the car race-able or will wear out the components much faster.

Stiffening your car for smoother tracks is a valid real-world setup approach. Excessive stiffness was the norm for early 90s F1 cars.
But even then, moderate setups in AC tend to result in excessive transmissibility over bumps and kerbs.

Michael Hornbuckle · Jul 7, 2016

If you can replicate your results on a vanilla content car then document what you have done and open a thread to report the bug.

Pablo Cortes · Jul 7, 2016

mmm vanilla my favorite rubber flavor

Eduard Mallorquí · Jul 7, 2016

It's not new that AC physics engine has flaws under very stiff suspensions, take for example the old kerb of death bug which was mainly fixed by editing tyre data eluding real values.

Take into account that when real data is used things like this can come up so the best is to move into a more serious simulation platform if you want to get a more accurate driving experience.

Radu Oros · Jul 7, 2016

Eduard Mallorquí said:
Take into account that when real data is used things like this can come up so the best is to move into a more serious simulation platform if you want to get a more accurate driving experience.

This doesn't make sense. You say you want to get a more accurate driving experience, but at the same time you're trying to use a crazy unrealistic setup for the car which wouldn't work to race the car properly.

mclarenf1papa · Jul 7, 2016

Eduard Mallorquí said:
It's not new that AC physics engine has flaws under very stiff suspensions, take for example the old kerb of death bug which was mainly fixed by editing tyre data eluding real values.

Take into account that when real data is used things like this can come up so the best is to move into a more serious simulation platform if you want to get a more accurate driving experience.

Tire damping values are more realistic now, not less (which was the only thing changed to fix the curb of death bug).

Regardless, it might simply be an error within the mod. I've worked on a number of cars with extremely stiff suspension and extremely stiff ARBs. I've never had this sort of issue at all...

RockyFra · Jul 7, 2016

[Ironic mode ON] AC tyre model lacks of plate tectonics variable, so if you're playing AC and you get an earthquake, you cannot simulate it. With RF2 you can simulate a fart from the pilot in front of you, and you get more dirty air on the wings. [Ironic mode OFF]

Jim Liak · Jul 7, 2016

Yep, he is done it again...

RMR1 · Jul 7, 2016

Jim Liak said:
Yep, he is done it again...

I just a bit teased what good ole Ed will have to say about super stiff suspension of the SF15-T. Because all he will be trying it and wouldn't explode

But for shure Kunos implemented a hidden cheat for that, can't be any other way.

Good ole Ed

Eduard Mallorquí · Jul 7, 2016

As far as I know some betatesters are already working hard behind the scene to make sure none of it happens on the SF15-T in order to get their free DLC copy.

Qazdar Karim · Jul 7, 2016

RMR1 said:
I just a bit teased what good ole Ed will have to say about super stiff suspension of the SF15-T. Because all he will be trying it and wouldn't explode
But for shure Kunos implemented a hidden cheat for that, can't be any other way.
Good ole Ed

F1s since 2014 are not really stiff, except on the front heave springs

mclarenf1papa · Jul 7, 2016

Oh forgot to mention, I was running a Group-C car with 7000lb/in front springs (an order of magnitude higher than anything in AC - as per the 1990 Nissan IRL...essentially a solid axle)...no issues whatsoever.