Physics discussion thread

For your own sanity, do not also look-up the third version of this car--the Ginetta G55 GT3, which is not in the GT3 category, but instead in GT Open!!

This is what happens when race car manufacturers name their cars for various technical reasons and not for consumer sales/marketing purposes

P.S., it is a very fun car to drive...

 

This is my biggest issue with the game in terms of driving feel and audio - There's this constant scrubbing sound that's extremely present even in a straight line - I can have tire audio down to 30 which seriously mutes the squeals and makes slides, skids and such way harder to hear. However, it barely does anything for the constant Tie Fighter-esque scrubbing that is universally present.

It's extremely annoying and I don't think you should hear even half as much of the tire in a straight line as you do in AMS2.

 

Turn down surface/road noise to fix that

 

This has definitely helped! While I still find the scrubbing noise activates extremely early, this is a definite improvement. Thank you!

 

I already tested this a few weeks ago. (As a setup creator, i always try to find new meats to make the car go quicker)
In the GT3 at spa, i was about 7-8 tenths per lap quicker on 2.0 bar hot pressure compared to running the maximum pressures. You're definitely losing grip in the corners if you go too high on the pressures. It might be different if you go to really high speed circuits like daytona but for the most tracks, i'd say 1.9 - 2.0 bar is probably the ideal range for GT3.

 

Yep, you had clearly less grip in corners earlier as well, but the overinflated tires were giving you so much advantage on straights because less rolling resistance that it seemed to be worth it on some tracks. On some update during this year, Reiza changed the amount of rolling resistance you get to be less on higher speeds, so the difference shouldn't hopefully be as favorable towards clearly overinflated tires that it might you gain more than you lose on corners.

I personally don't notice any clear delay, but it could be also that my brain is somehow used to it. I guess it's also possible that you have some input delay that I don't. I'm generally quite conservative and smooth with my steering inputs, maybe it mitigates this as well. EDIT: have you tried to remove the interpolation filter altogether when playing AMS2? I think it's beneficial with games like iRacing that might not have fast enough refresh rate for the FFB signal from the game. I never ever have thought to use similar filter on my Moza for AMS2.

The GT3 understeer so much on high speed on default setup, that not sure how you can get that swinging pendulum effect with those on faster corners unless you start to lift too much and cause the engine braking to add load to front. Before the changes to the GT3 (including the change to aero loss on yaw), the rear did struggle more on faster speeds and you had more of that rear sliding going in the fast corners.

Just want to be sure, everybody who has issues with the GT3, are you testing with default setup and not some older or TT setups from the past?

 

You solve the problem. In a racing car, racing, moving, we never have 100% static friction between tyres and racing surface. Never. And AMS2 does It really nice. For me, it's better than other sims this transition in grip. Maybe some tweak in grip curves can make this feel better, but the physics is accurate.

AMS2 today is the only sim that I don't feel the "on rails" to "some slide" abrupt transition. And this is really accurate thinking in rubber tyres spinning.

 

Finally got to the GT3s which this thread seems to have devolved into (there are so many other cars). I'm not feeling the "floatiness" at all. I do recall this type of feeling when I first started playing AMS2 again, but I think it comes down to a few things (a) overbearing scrubbing sounds as mentioned above, (b) not enough power or too much damping in one's FFB setup and/or (c) weird setups, though the default wasn't bad in this case.

My most recent experience has been with LMU and these feel better to me (I'm not a pro driver and the closest I've come is being a passenger in a production 911 GT3...which has barely anything in common with the race version). The car is very predictable and little mistakes result in some scrubbing, not massive, uncontrollable spins. When I did push too far and get into an uncontrollable spin, it seemed to start and stop realistically if confined to pavement. Maybe I can stomp on the gas pedal a little too easily without TC and there's a corner in Mosport where I thought the ability to trail brake really hard was a bit much, but all just conjecture on my part. I'm really happy with how everything feels.

 

This is objectively false. Tyres do operate within static friction. As do all solids.
Static friction behaves like a force that reacts to whatever load is applied, up to a limit



As long as the required force is below the limit, the rubber does not slip.

A tire is a flexible, viscoelastic body. When it rolls:

• The leading edge of the contact patch is deformed and pushed forward
• The trailing edge is released
• Rubber molecules stretch and relax
• The carcass deflects

Because the surface molecules of the rubber "stick" to the asphalt during the deformation, the tire can support forces without sliding.

One of the main mechanism that creates grip is called adhesion. It's when rubber forms chemical-mechanical bonds with the road surface. The rubber keylocks into the tiny peaks of the asphalt. While bonded, it can transmit force. No sliding happens unless force exceeds μ·N. In order for sliding to happen, these bonds need to be broken and the intermolecular forces resist that.

Consider the following example: gentle acceleration.
The engine sends torque to the wheel. The wheel would have to rotate faster than what it would during free rolling, meaning it would need to slip backwards relative to the road. But static friction acts to prevent this, exerting a forward force at the contact patch. This is the acceleration of the car with zero relative motion between the contact patch and the road. The tyre does not slip.

So in conclusion, tyres absolutely do operate within static friction.

 

You wrote a lot, all perfect. And I never write that tyres in a moving racing car don't have any static friction, I said that never it's 100% static friction. And to be more clear, static friction is massive, more in slicks, as explained.

But even in controled acceleration, in contact we always have dynamic friction working. Racing surface and rubber are a lot imperfect regards contact, even macroscopic. I like how AMS simulate this, with a more subtle transition between no slide to some slide, and recovery.

For more reading about this mess:

Magic Formula - an overview | ScienceDirect Topics

Tyre dynamics

Sim racing games suffers a lot. Train sims are more simple, for sure!

I recorded a short vídeo using SimHUB to show telemetry regards slip ratio and slip angle, with a snap oversteer. I like AMS2 a lot! And bass shakers even more.

 

I agree. Like TC, it triggers too early and becomes omnipresent and annoying.
Perhaps it is just representing the tyre slip (from internal physics engine), but then we go back to the "slidey" effect...

 

This very interesting video popped up on my feed this morning, I thought I'd share it here. Wonder what the floaty physics mob might think

 

Thank you so much for this! Exactly this is what I've always been loving in AMS 2 - tyres are deforming and you can feel it.

 

After using SimHUB with bass shakers, I keep this audio effect really low. But It is good for some feedback, for sure.

 

You can hear the tires scrubbing and squealing even before losing grip?!? Oh, no, AMS 2 might be accurate in that way, too??

I think the area that confounds people, and it may be partly related to FFB, is the transition from manageable slip to loss of grip. It is tricky (IRL akin to driving on a very low-grip surface when you are at the edge of grip), but is the game properly communicating this transition? I have heard arguments on both sides, but enough people seem to have difficulty that it may not all be attributable to low skills.

 

I often wonder when people say AMS2 is easy or "simcade", if much of that may come from having more communicative feedback (FFB). I think the tire sounds still have room for improvement in terms of transitions from scrub to squeal noise but, the FFB provides great tactile feedback overall.

 

I had the great opportunity to be passenger to Oliver Gavin in a C8 Corvette on a track. The tires squealed in every single corner as he pushed the car. But were talking racecars in AMS2, they are so freaking loud from the inside that you definitely won't hear tire squeal over as loud over the engine and wind noise. I tuned down tire sounds to a very low value which is fine to me, so the default volume and the sound feeling like it's not dynamically changing but always sound the same could be improved imho.

 

Appreciate that video in the Porsche dealership. I always thought slip angle meant the tire was sliding slightly but still very much in control. I didn't realize it was actually the rubber twisting (right word?).

As for the tire sounds on AMS2, it feels like they're trying to give as much info as possible since we don't have g-forces. But they went too far. It's almost as if they make noise the second you hit .0005 degrees of slip angle and the noise doesn't grow proportionately.

 

The audio in a racing sim isn't meant only to be realistic, but rather to communicate things you would be able to sense in a real car. Of course, you can adjust it so it is muffled (as if you're wearing a helmet and ear protection) and hear nothing but the muted roar of the engine.

Same with FFB. Do you want to know what the car is doing, or feel only rack forces and imagine all the rest that is missing when you are sitting in front of a computer screen instead of in a real vehicle with g-forces, etc.? Luckily we can all chose what we prefer.

 

Exactly, the physics in the game are believable, IMHO. Could use some improvement but it's been getting better with every iteration. Even people who race IRL disagree on which sim feels realistic because the real world is way more complicated with all the options a driver has to setup a car.

I think a lot of people that are commenting on the "floaty" feel are actually talking about FFB not giving them the communication they are seeking, but are incorrectly using the term "physics" to describe it. :shrug:

"Physics", to me, is strictly is how the car and tires behave in different situations, and how the car behaves when a driver adds an input in an extreme situation.
Eg: If you do X with your inputs, then the car/tire should do Y in a specific situation. If it doesn't do Y (like IRL) but does Z, then the physics in the game needs work. This again has nothing to do with a "floaty" feeling FFB, IMHO.