Burnout physics/Exploits

[mantasisg]

@Nieubermesch Oh yes that is definitely an excuse. But it is also definitely how it is IRL, and it could be part of an equation. Surfaces has nuances too, not only tires.

The starts of the burnout are still happening at low speed, so tire is still able to make lots of revolutions on almost same place of the tarmac, accumulating more heat in there, that it would happen if car would travel a bit faster.

It would be interesting to find RL "brakey" examples that would be initiated at higher speeds, and see if car eventually halts to a stop with rear tires spinning with full power, or would it push locked front.

And digging holes is just extreme side of an effect, it would really take a moment for tarmac to reach melting point, and it would have less friction already. It would happen at its very surface and that is enough to have contact friction affected.

I wonder if you can do this test successfully in other simulations.

I am also not talking about track surfaces being dynamic simply. But them being so at each and every place, like registering and calculating temp gain within every contact with a tire, maybe even registering and calculating effect of being in sunlight or shadow, which would probably be something not to happen in simulations for ten more years if not more. Imagine wet tracks evaporating only where it is sunny, and remaining wet where it is in shadow, but that is another topic.

 

[Nieubermesch]

You're right, I'm a fanboy.. of scientific method, and I see none on those posts.

Yeah, I see a lot of the scientific method in your posts too, right? You know some things don't require much more then just close observation to rule out right? Not all science has to be crazy complex and extensive to at least be close to some precision and validity... You're posts are just demoralizing and this is just an interesting thread to discuss this. If it would be the opposite and would be a guy singing praise about a certain behavior you wouldn't be commenting and he could be just as much not being scientific about it. But whatever man. It has been seen that the actual people giving critics - wrong or right - are pushing the sim forward with S397 being very good at delivering updates and on a good path. This is not to bash RF2, just another detail on physics that might be or not be right and contributing to handling characteristics of cars.

 

[Nieubermesch]

@Nieubermesch Oh yes that is definitely an excuse. But it is also definitely how it is IRL, and it could be part of an equation. Surfaces has nuances too, not only tires.

The starts of the burnout are still happening at low speed, so tire is still able to make lots of revolutions on almost same place of the tarmac, accumulating more heat in there, that it would happen if car would travel a bit faster.

It would be interesting to find RL "brakey" examples that would be initiated at higher speeds, and see if car eventually halts to a stop with rear tires spinning with full power, or would it push locked front.

And digging holes is just extreme side of an effect, it would really take a moment for tarmac to reach melting point, and it would have less friction already. It would happen at its very surface and that is enough to have contact friction affected.

I wonder if you can do this test successfully in other simulations.

I am also not talking about track surfaces being dynamic simply. But them being so at each and every place, like registering and calculating temp gain within every contact with a tire, maybe even registering and calculating effect of being in sunlight or shadow, which would probably be something not to happen in simulations for ten more years if not more. Imagine wet tracks evaporating only where it is sunny, and remaining wet where it is in shadow, but that is another topic.

Oh, I see what you meant now, not about the dynamic temperature of the track, but a realroad for track temperatures on specific spots and all. I guess the rubbered in line would be hotter then? It gets really microscopic at that point, the dynamics of the track progression and I also would think it would be too much work for something that I imagine isn't really that impactful at all, since I am sure thermodynamic balance with the air and everything will leave it somewhat all homogeneous- I guess ahah. About wet tracks would be cool though, and also the same for dry tarmac under the shadows. Guess that will come some day to sims. More realism wouldn't hurt anyone - not quite true .

Well, I guess the car would have to be suspended at the rear and than lowered to make contact at the ground to see if the car would be pushed forward, but I would bet it wouldn't move with propperly locked brakes, but that's just me.

 

[Nieubermesch]

I just did some test with rear tires in the grass while the fronts are on tarmac and there it stays fixed, so it's definetly something about the grip difference between both front and rear.

I saw some videos of Assetto Corsa and the same happened in the limited cars I saw - fronts don't lock the car in place.

Another sim where the cars actually remained in place was BeamNG. That sim definetly doesn't mess around with the loss of grip from my experiences with it.

 

[Nieubermesch]

Could it be just a graphical bug/ glitch ?
The front tyres are probably moving under the physic engine but visually they seem static, this is my theory.
I still haven't checked it yet (I forgot lol).
Will try to do motec on this.

Hum, might be something to consider, but I wouldn't think the graphics would translate that wrongly, but a possibility. It sounds strange to me though, as it would mean you couldn't really lock the brakes of the cars in question even under 100% brake applied.

 

[Nieubermesch]

Took a while to get what you are saying ( not your fault), yea this would mean that even when you see visually brake locking BUT it isn't locking actually.
It was just a guess anyway.

Yeah, it would mean those cars under test can't lock the brakes, I think. If that was the case they wouldn't also lock during normal driving, as it would be harder to lock then.

 

[Nieubermesch]

This is BeamNG with just two randomly selected cars. Keep in mind that BeamNG doesnt simulate tire temperatures and the car still manages to stay in place - although with some big back and forth motion with the suspensions and would probably disappear with actual grip loss at high temperatures.

 

[Lazza]

There's another side to this test possibly being overlooked: static grip.

You're (as a group) comparing the rear spinning grip to the front static grip. It could be the front static grip at fault to some extent. Any ideas on verifying static grip? We know from past experiments and dev clarification that road surface angles approaching vertical don't work properly which rules out a steep hanging test.

Other peoples posts are just what I expected from them - diverging and focusing more on other mentioned topics wich aren't really the real issue, for whatever reason. I would actually like to see them actually give their opinion about it, but the silence kinda speaks for itself and I am sure they also find this behavior odd.

If Einstein is explaining general relativity and mentions the sky is green, that's wrong regardless of the rest of the paper. You CAN discuss specific topics without needing to give an opinion on the whole, and assuming no comment means agreeance is a logic error as well.

 

[Nieubermesch]

There's another side to this test possibly being overlooked: static grip.

You're (as a group) comparing the rear spinning grip to the front static grip. It could be the front static grip at fault to some extent. Any ideas on verifying static grip? We know from past experiments and dev clarification that road surface angles approaching vertical don't work properly which rules out a steep hanging test.



If Einstein is explaining general relativity and mentions the sky is green, that's wrong regardless of the rest of the paper. You CAN discuss specific topics without needing to give an opinion on the whole, and assuming no comment means agreeance is a logic error as well.

Of course it is, but the silence in this case actually speaks volumes when I see main focus of intervention on other side topics when the real one is thrown aside and ignored when I'd like that people like you actually gave this more importance then correcting someone about differentials, nothing more.

It could be also static grip, sure. One is either higher or lower, from first supposition when looking at this tests, at least to me. Coupled with the link (weak and conjecture of course) about some possible excessive rotation and forward push on throtle during slides, I wouldn't say it is the front static grip - if the link is there.

 

[Lazza]

I am also not talking about track surfaces being dynamic simply. But them being so at each and every place, like registering and calculating temp gain within every contact with a tire, maybe even registering and calculating effect of being in sunlight or shadow, which would probably be something not to happen in simulations for ten more years if not more. Imagine wet tracks evaporating only where it is sunny, and remaining wet where it is in shadow, but that is another topic.

Off-topic but perhaps of interest to you: ISI made noises about doing this (to some degree) in rF2. As you can already tell, it never happened. I'm sure if it had it would also affect the track wetness. Maybe we would have had standing water in that case too. What we don't know is whether their planned implementation (being ISI and taking everything to the next level, like the tyres) simply wasn't feasible due to calculation and bandwidth requirements, or they ran out of time to get the level of attainable implementation locked in and executed.

As in many areas I think you could have something quite coarse and imprecise, but good enough to convey a sense of reality. Anyway, sorry to extend the OT.

 

[green serpent]

You're right, I'm a fanboy.. of scientific method, and I see none on those posts.

If every post about physics had to live up to those standards, there wouldn't be many more posts left of this forum about physics, that's for sure!

Even the videos done by Niels, which are insanely in depth using data etc, fall short of the actual scientific method. I think we're all well aware here that none of this is actual science.

Scientific theorys start out as simply ideas, just a hypothesis. THEN the scientific method is used to determine it's merit, and even after that it is still considered an approximation.

Talking about ideas and possible explanations, I don't think there's anything wrong with that provided you don't say it's fact. I'll leave it to people smarter than me to do the actual science if they wanted to put that much effort in

 

[green serpent]

It is interesting test. I did this test to several of cars that I have been working on with physics. I was relatively confident they would succeed in this test, but they didn't. Perhaps it could be rF2 problem, but I doubt it a lot, it is probably about parameters of cars.

I have preliminary confirmation that the Mazda MX5 Mk1 mod can achieve line-locker burnout status at 10% rear wear with 90s street tire, and 50% rear wear with semi slick (turbo version). Awaiting experimental replication and peer review . No changes in setup. I tried making extreme changes in setup to camber and pressure and tried with no rear wear, but it didn't work. Only slight rear tire wear did the trick. Sorry for spamming this thread with low quality phone recorded vids. It's just easier as I can stay in sim and uplaod at the same time!

As others have stated, I think it is very possible that track surface and the interplay between the track and the tire could be part or all of the issue, rather than the tire.

As mantasisg stated, IRL the road surface would lose grip as the rear tires spin up. Perhaps what is happening is a) the track surface is not heating up as it would in real life and losing grip (as this is missing from the simulation I believe), and b) maybe it could actually be gaining grip as the track under the rear tires rapidly get saturated with real road, whereas the front tires remain on a non-rubbered in part of the track.

I think I might try this on a saturated track, so at least the real road variable can be removed. See.. removing variables is scientific, maybe this IS hardcore science after all

P.s this is an awesome driving mod @mantasisg , I absolutely love it!

Edit: fully rubbered in track did not really change things too much. Still front end pushes if rear tires have not been worn.

 

[Korva7]

Static real road didn't seem to remove the problem either.
Asphalt melting not being simulated sounds like a possible reason.
What about the molten rubber from the tire? It getting between the tyre and road would also reduce grip in this kind of slow speed situation, i think.

 

[mantasisg]

Speaking of static friction, it definitely is at play. Although I have not succeeded in several cars I have tried this test, but front engined cars seemed to do best, mid engine worse, and rear engine cars worst. Which of course happened because weight distribution, thus static friction at front went lower as engine weight moved more towards rear.

I have preliminary confirmation that the Mazda MX5 Mk1 mod can achieve line-locker burnout status at 90% rear wear with 90s street tire, and 50% rear wear with semi slick (turbo version). Awaiting experimental replication and peer review . No changes in setup. I tried making extreme changes in setup to camber and pressure and tried with no rear wear, but it didn't work. Only slight rear tire wear did the trick. Sorry for spamming this thread with low quality phone recorded vids. It's just easier as I can stay in sim and uplaod at the same time!

As others have stated, I think it is very possible that track surface and the interplay between the track and the tire could be part or all of the issue, rather than the tire.

As mantasisg stated, IRL the road surface would lose grip as the rear tires spin up. Perhaps what is happening is a) the track surface is not heating up as it would in real life and losing grip (as this is missing from the simulation I believe), and b) maybe it could actually be gaining grip as the track under the rear tires rapidly get saturated with real road, whereas the front tires remain on a non-rubbered in part of the track.

I think I might try this on a saturated track, so at least the real road variable can be removed. See.. removing variables is scientific, maybe this IS hardcore science after all

P.s this is an awesome driving mod @mantasisg , I absolutely love it!

Edit: fully rubbered in track did not really change things too much. Still front end pushes if rear tires have not been worn.

Oh thats cool. Thanks. I have tried Miata, and it didn't work for me. Although it was standard engine and with fresh tires. Turbo version with worn tires certainly seems like better candidate to success in this test. Though 90s tire not very grippy to begin with, and semislick with 50% of wear - massive wear. But it is nice to see it working at least like that.

I also thought perhaps people IRL wouldn't really do this thing with new fresh tires. But I guess it is irrelevant, because by doing this, tires would get very used in several seconds.

Good insights about surface getting rubbered, and thus even increasing friction at that place of a spin up. Tires have parameters for various things for rubbered groove, which some of them are for sliding.

I don't want to butcher tires just so this test would work, but I'd love to try to modify them to get better results at this, and not to make the car handling any worse, who knows perhaps it would be possible to achieve more realistic tire configuration by following this test. Will try it some day. I'd probably try altering some of those: altering wear at highest temps, altering highest temps friction loss, altering rubbered surface effects for sliding, altering high velocity sliding effect for sliding friction (I assume it counts velocity of actual tire rubber vs surface, not car velocity), it would probably be helpful to learn about WLF stuff for once since it is in play. I never played with gasses parameters, but perhaps getting more of pressure increase due to tire heating would help, as well as heat transfer to road surface parameter.... Well it sounds like this thing could potentially lead into remaking the tire lol But it wouldn't be acceptable to get worse results from the tires, just for getting better results in this test, which also would be a sign of something not working the correct way, as ideally everything should become more realistic.

Asphalt melting not being simulated sounds like a possible reason.
What about the molten rubber from the tire? It getting between the tyre and road would also reduce grip in this kind of slow speed situation, i think.

This is interesting idea, it must have a play I guess. Although iiric I have read that tires can wear differently, some can kind of melt-off the surface, and some tend to wear into kind of rubber dust or something like that. Anyway, I also remember reading that black marks that we can see on tarmac from tires, happens due to bonding of heated tarmac and heated tire touching, although I guess it also could be just pieces of rubber getting stuck in irregularities of the tarmac.

 

[mantasisg]

This is iRacing clip I found, I was almost sure it would work in iR since it is known to have very sharp high peak tires, while at the same time being pretty bad to slide, and also very hard to recover from overheating tires, so I don't think it marks iR as overall correct, but certainly for this test it works:

 

[green serpent]

Oh thats cool. Thanks. I have tried Miata, and it didn't work for me. Although it was standard engine and with fresh tires. Turbo version with worn tires certainly seems like better candidate to success in this test. Though 90s tire not very grippy to begin with

I actually made a mistake in my original post, 90s tires only had 10% wear, so these tires seemed to come very close to "passing" the test.

I don't think it marks iR as overall correct, but certainly for this test it works

Back when I was a weathly man and could afford iR, I would do line-lockers at the end of pit lane in the Mx-5. So definitely in this respect iR is "correct", but again none of this is to say that anything here is fake. I found another video with a Cobra pushing the front end at various times in this rather ugly burn out session (at the very beggining and at 40 sec). The surface looks pretty rubbered in.

Here is a clip though of a Cobra doing a text book perfect one with no front tire push at all (40 sec again).

 

[Nieubermesch]

Lol try FISI ->
View attachment 39727

Oh wow

 

[Nieubermesch]

I actually made a mistake in my original post, 90s tires only had 10% wear, so these tires seemed to come very close to "passing" the test.

Back when I was a weathly man and could afford iR, I would do line-lockers at the end of pit lane in the Mx-5. So definitely in this respect iR is "correct", but again none of this is to say that anything here is fake. I found another video with a Cobra pushing the front end at various times in this rather ugly burn out session (at the very beggining and at 40 sec). The surface looks pretty rubbered in.

Here is a clip though of a Cobra doing a text book perfect one with no front tire push at all (40 sec again).

Well, seems that under the right conditions it's realistic behavior then. Probably all the things like rubber and no tarmac heating are the culprit .

Still, has anyone seen the video with the BMW M2 CS lap? It's so strange seeing record times being achieved like that!

 

[green serpent]

Well, seems that under the right conditions it's realistic behavior then. Probably all the things like rubber and no tarmac heating are the culprit .

I think as @Sim_Player mentioned, the bobbying around and vibrating that the cars are doing is probably not helping the situation either.

Still, has anyone seen the video with the BMW M2 CS lap? It's so strange seeing record times being achieved like that!

The times he is getting (while very fast, and maybe too fast), are generally not the fastest lap times. Not actually "records or world beating" as far as I have seen. Henri is fast, but there are faster sim racers who are less slidey. So at best, sliding does not lose much (if any) time, but probably not gaining much time. I wonder if there are any other videos from others sim racers that are really abusing the car and getting the fastest time? From what I have seen the really fast guys like Nuno Pinto and Risto Kappet etc are not overly slidey (from what I have seen).

My own personal experience with many stints in the past goes something like this:

Try to drive using real world techniques, trying very hard to be smooth and flowing perfectly through corners as if I were driving the car IRL > get smashed by the AI, so decide to be way more aggressive and abuse the living shite out of the car > get a hugely faster lap time straight away, start thinking rF2 is fake > keep trying to get faster > eventually find that being very aggressive AND very smooth/not too slidy is the fastest.

Basically, although sliding and mashing on the throttle will be much faster than a timid and overly cautious yet technically perfect style, ultimately being less slidy, very smooth but still massively aggressive has always been faster for me. And I've tested this out quite a lot, and each time that is the result. Do I think sliding should be punished more, yeah probably, but proper driving style is still faster imo.

Edit: I still do feel though that something is NQR (in a very general and subjective way, i.e I have no actual evidence to back this and I'm not saying that it is fact). After watching dozens of hours of rF2 on board, the front end of the car seems to move in a very non linear way. The trajectory of the car is very darty and it lurches around too much imo compared to RL. If you drive down a straight in rF2 and go hard right to hard left in steering, you will notice a very non linear kind of "lurching" that the car does. For example the car will turn, then settle, turn, then settle etc. Could be totally realistic, but compared to RL onboard the trajectory of the car seems too hectic and not moving in smooth flowing linear arcs. You will notice A LOT in rF2 on boards little "gathering up" steering motion at the exit of corners, again and again and again. The car may not even be sliding and still these little gather up moments of steering input at the exits of corners. Just seems like the car is a bit floppy on it's chassis or something. This could or could not be related to the whole topic of this thread, IDK.

 

[Nieubermesch]

I think as @Sim_Player mentioned, the bobbying around and vibrating that the cars are doing is probably not helping the situation either.


The times he is getting (while very fast, and maybe too fast), are generally not the fastest lap times. Not actually "records or world beating" as far as I have seen. Henri is fast, but there are faster sim racers who are less slidey. So at best, sliding does not lose much (if any) time, but probably not gaining much time. I wonder if there are any other videos from others sim racers that are really abusing the car and getting the fastest time? From what I have seen the really fast guys like Nuno Pinto and Risto Kappet etc are not overly slidey (from what I have seen).

My own personal experience with many stints in the past goes something like this:

Try to drive using real world techniques, trying very hard to be smooth and flowing perfectly through corners as if I were driving the car IRL > get smashed by the AI, so decide to be way more aggressive and abuse the living shite out of the car > get a hugely faster lap time straight away, start thinking rF2 is fake > keep trying to get faster > eventually find that being very aggressive AND very smooth/not too slidy is the fastest.

Basically, although sliding and mashing on the throttle will be much faster than a timid and overly cautious yet technically perfect style, ultimately being less slidy, very smooth but still massively aggressive has always been faster for me. And I've tested this out quite a lot, and each time that is the result. Do I think sliding should be punished more, yeah probably, but proper driving style is still faster imo.

Edit: I still do feel though that something is NQR (in a very general and subjective way, i.e I have no actual evidence to back this and I'm not saying that it is fact). After watching dozens of hours of rF2 on board, the front end of the car seems to move in a very non linear way. The trajectory of the car is very darty and it lurches around too much imo compared to RL. If you drive down a straight in rF2 and go hard right to hard left in steering, you will notice a very non linear kind of "lurching" that the car does. For example the car will turn, then settle, turn, then settle etc. Could be totally realistic, but compared to RL onboard the trajectory of the car seems too hectic and not moving in smooth flowing linear arcs. You will notice A LOT in rF2 on boards little "gathering up" steering motion at the exit of corners, again and again and again. The car may not even be sliding and still these little gather up moments of steering input at the exits of corners. Just seems like the car is a bit floppy on it's chassis or something. This could or could not be related to the whole topic of this thread, IDK.

I commented under his video asking and his times are not the world record but close by ms's, so it kinda is fast, maybe the fastest. I mean, I had already tried with a Porsche GTE to drive more by sliding and abusing slip angles and setups and by just the first time doing it, I was beating my lap times at Spa and I'm a somewhat decent driver to be honest, at least to not just surpass those clean laps with just a crazy driving way.