“ are rfactor2 physics broken” video

Discussion in 'General Discussion' started by GTClub_wajdi, Dec 29, 2020.

  1. Andregee

    Andregee Registered

    Joined:
    Jun 23, 2012
    Messages:
    927
    Likes Received:
    390
    Actual race tyres from Pirelli have 7 degrees optimum slip angle. I know it from Lord Kunos who told that they got data from Pirelli and the their old AC tyre modell with 12 degrees was totally off.
     
  2. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403
    Well, yeah, I just mentioned it because of the reasons I mentioned, as some posts above you can see in the Stephan post where he talks about coordinate of the car and stuff.
     
  3. Yzangard

    Yzangard Registered

    Joined:
    Dec 6, 2018
    Messages:
    351
    Likes Received:
    173
    Ok, here is my first batch of "results" (very small test with Audi R8 LMS 2019, a few laps), still over rotating my wheel to induce understeer, as before :

    upload_2021-1-22_23-45-43.png

    I added the "curve" that actually shows more or less that what we have is what we should get. Requires more data but results seem "legit" and close to real life tyres...
     
    Last edited: Jan 23, 2021
    Comante likes this.
  4. Yzangard

    Yzangard Registered

    Joined:
    Dec 6, 2018
    Messages:
    351
    Likes Received:
    173
    Sure but my previous results didn't took into account several things and only shown lateral G Forces of the whole car, not the actual stress on the tyre.

    I just added first batch of results with lateral force on the tyre directly and it is MUCH closer to what we should get...actually, I think it is clearly pretty much what we should get.


    PS : Lord Kunos is Aristotelis or another one ?
     
  5. Andregee

    Andregee Registered

    Joined:
    Jun 23, 2012
    Messages:
    927
    Likes Received:
    390
    I have to correct myself it was Aris V. not Lord Kunos. Lord Kunos is Stefano ...
     
  6. Yzangard

    Yzangard Registered

    Joined:
    Dec 6, 2018
    Messages:
    351
    Likes Received:
    173
    Aris is very interesting and a very nice guy.
     
    Nieubermesch likes this.
  7. Stefan_L_01

    Stefan_L_01 Registered

    Joined:
    Nov 18, 2012
    Messages:
    595
    Likes Received:
    386
    Nice graph. The problem is you find a doozen different opinions what slip angle is realistic. Most sources you find in internet are for street tires, and they vary a lot. However if we look at a great compendium of tire grip from michelin
    http://www.dimnp.unipi.it/guiggiani-m/Michelin_Tire_Grip.pdf
    we find the info for street tires optimum slip angle is 4-7°.
    Now your chart seems proper for street tires, not for race tires. Even if you add some high load etc.
    I have not seen a race tire chart however, probably that data is somewhat secret and hard to get
     

    Attached Files:

    Last edited: Jan 23, 2021
    Nieubermesch likes this.
  8. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403


    Take a look at this and then tell me it doesn't look any simillar. I am replying to you but it's not really directed at you as if you are denying anything, but some people will probably not believe it when they see this exploits are very simillar to the way PC2 drives fast too... Leaves me disappointed, but oh well.



    Can't say I feel confident about the behavior of the Oreca after all. That last chicane at the end really looks out of this world to be honest. Upon further observation I can't really think anyone who has watched racing think that lap would ever be possible on a car like this. The thing is that the setup exploits that certainly are at play here, are just an exacerbation of the problem that exists even if setups exploits are present. Setup exploit is not an exception, it's an aggravation of what is possible to be done at at least some cars.
     
    Last edited: Jan 23, 2021
    Slip_Angel likes this.
  9. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403
    I wonder what you really mean by your second theory about what might be happening with grip. I didn't fully understood it, and I doubt the tyre doesn't simulate slip angle...
     
  10. Comante

    Comante Registered

    Joined:
    Nov 20, 2013
    Messages:
    1,694
    Likes Received:
    1,218
    I want just to remember that the amount of steering turn means very little without knowing the steering ratio. I have driven custom mods where to be fast you had to have superfast steering, and mods where you had to turn it down to the minimum to avoid understeer and to ruin the tires. This obviously affect the amount of steering used. Having large wheel movements for little steering angle , help with controlling more precisely the angles , yes to be fast you don't want FFB to slow you down.
    This wall of text just to remember to not judge partial informations hastily.
     
  11. Yzangard

    Yzangard Registered

    Joined:
    Dec 6, 2018
    Messages:
    351
    Likes Received:
    173
    Actually there are several books dealing with this in which they describe the curvature of the graph without precise slip angle, only the shape. For example, Milliken & Milliken book or this website : http://racingcardynamics.com/racing-tires-lateral-force/

    The actual ideal slip angle was given by Aristotelis (from Kunos) in his youtube channel (Aris.Drives) : around 7°.

    This information is from Pirelli themselves but since Aris in under NDA, he didn't gave much precisions, only the ideal slip angle and the "good range" of slip angle. The shape is exactly the one described in books and the one I've found in my testings.

    All those information tend to prove that rFactor 2 tires aren't that bad at all when compared to their real counterparts.
     
    Last edited: Jan 23, 2021
  12. mechanic

    mechanic Registered

    Joined:
    Nov 7, 2019
    Messages:
    18
    Likes Received:
    17
    Good to see some data and telemetry based analysis in this thread instead of onboard videos and comparisons with other sims based on feelings.

    If it helps, Avon published Pacejka parameters for some of their race tires. Can't find them on my phone, but you can find them somewhere on their website.

    I not interested in gt3 and never driven one, but bases on feelings there is quite a big difference in how tires are modelled across various mods. It just shows how incredibly difficult it is, mainly because a lack of data.
    From a pure simulation point of view the bristle model of rf2 is pretty good and very advanced for a game which runs on comsumer hardware.
    It is however difficult to get it right, and while a have A LOT of respect for the developers, we have to keep in mind that they have to build these cars and tires with limited real life info and data and a very limited budget. Some people here are really hard on the devs, it still is a game, built for amusement which only costs 30 euros.
    I use professional sim software on a daily basis which costs literarely 1000 times more and has more flaws and issues than rf2.

    That being said, I do feel that the newer s397 tires are maybe a bit too forgiving. I mainly use the Caterham Classic mod with the 13 inch tires from the s397 tire database. Tires are forgiving (just like the car) but I think the are pretty realistic and have a lot of depth in their behaviour.
     
    Comante likes this.
  13. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,926
    Likes Received:
    3,876
    True, data and telemetry is best. But even before that we should be able to get a rough feeling of where is the ballpark by collecting observations. I think matching RL vs Simulation observations is a worthy practice to evaluate realism up to rather high level. Using reality footage as evidence and treating it by logic and combining it with knowledge should bring up useful conclusions, however by getting data and doing telemetry matching it could be pushed further and more precision achieved.
     
    Nieubermesch and Slip_Angel like this.
  14. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403
    Agreed 100%. It's as if some people forgot how so many of our advancements by so many great minds (Einstein, Newton, Galilleu...) started with observation and they confirmed it or actually just found the way to express it mathmatically.
     
  15. jGer0705

    jGer0705 Registered

    Joined:
    Mar 15, 2020
    Messages:
    4
    Likes Received:
    1
    All this because Ermin posted a video questioning the physics using exploitive setups. Which I have a feeling he only created because he was tired of being called an rfactor2 fanboy, watch any of his previous content. Regardless I hope some of these concerns are noted by the developers and can be looked into
     
  16. hitm4k3r

    hitm4k3r Registered

    Joined:
    Jun 15, 2016
    Messages:
    1,320
    Likes Received:
    3,121
    I have no idea why we are still just watching Henris videos and claim based on those videos that there is a major issue with the rF2 physics engine per se. Ofcourse the car will look out of this world going through the last chicane - it's a virtual car after all. He has unlimited virtual cars, tires and no fear of hitting a wall if he get's it a tiny bit wrong. That's what made people wondering why the Oreca went like 6 seconds faster during the Le Mans virtual. Those are pro sim racers who propably do nothing else than searching for exploits. There are factors like confidence and fear that play propably the most important role on how you approach a track.

    What do you guys think how big the impact of the mind set of a driver is, when he knows during a qualy run that he has to bring back the car in one piece to be able to race the next day, knowing that he has to give a bit of a buffer here and there to not f*ck everything up for his team and his teammates before the race has even started? Or that he could get seriously injured or killed when doing the stuff that people do in sim racing?

    Curbs can be taken very agressively in rF2 because the physics engine and damage model permits it. In other sims you can't do that because the physics engine goes nuts when tires hit a curb. So what do we take from that? Are all physics models wrong? Wich one is right? All we can take from this is that there are so many fundamental variables that make real racing as real as it is, while we lack "simple" factors like breaking suspensions due to overstressing them over longer times, tire failures or even just limited budgets. Should we really make sim racing cars, virtual fuel and tires as expensive as the real stuff to get the same sensation? I think sometimes you guys are overstressing how important certain factors are, while there are other potential areas that have a much bigger impact on the driving. Is one degree slip angle more or less really that important compared to let's say, the lack of a proper drive train model, tire failures or a suspension damage model that punishes hitting curbs hard for longer periods of time? How about to finaly start limiting tire sets, engines and cars per weekend for such high profile competitions with incredibly long repair times, no options to hit esc while being on track and real parc fermé rules? Isn't there more gained by such systems than building big databases with alot more human impact and sources for errors just to get slip angles not 95% but 99% right? AMS1 was the last sim that implemented limiting systems, like limits for tire sets, longer repair times and even longer times to toe back the cars back to the garage and I think it's time for S397 to take alook at this aswell.
     
    McKiernan and RaceNut like this.
  17. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403
    Would you live under a house whose pillars would crumble down by a gust of wind? And then go and add lots of luxuries just for everything to go to hell when the house would crumbling down? It's a bit like that. I believe we can have that but I think we should start with this aspect of physics first. Competition will become a joke when people finally see how some people are fast and next time we have online coverage of Rfactor 2 things will look ridiculous.

    Do you think people really don't sometimes push as hard with no fear of death or damage? Or rather, despiste those things? Humanity history is full of that, so cmon, sometimes world record laps would probably look like Rfactor2 exploitative driving if it was possible. But certainly tyre and suspension damage might be something that allows this too, but I don't believe it's only that that is creating such results.
     
    Slip_Angel likes this.
  18. hitm4k3r

    hitm4k3r Registered

    Joined:
    Jun 15, 2016
    Messages:
    1,320
    Likes Received:
    3,121
    I didn't say that those issues that I mentioned are the only factors, don't get me wrong. But going by your analogy we have some very different perception of what luxury and what a house pillar is. The lack of consqequences is what enables sim racers to do what they are doing. And in that regard infinite tires, fuel, budget, suspension lifetime, tire lifetime etc etc etc has a tremendous effect on how we attack a track. A real racing driver would never take the last chicane at Le Mans like it was taken during the Le Mans virtual by the majority of drivers during the whole race, simply knowing that this can have severe consequences in the real world. Let's be real, Henri does like thousends of laps propably just to patch that single out of this world laptime together having to worry about nothing but his laptime. None of the current real world drivers has that luxury. :)
     
  19. Nieubermesch

    Nieubermesch Registered

    Joined:
    May 18, 2019
    Messages:
    792
    Likes Received:
    403
    Yes, but when the handling isn't propperly dialed in and it may be because of that more probably that things can be driven a certain way and not because of consequences, don't you think it should the priority? I know it's not a perfect analogy, but it's priorities, and most sims don't really do great damage modelling.
     
  20. hitm4k3r

    hitm4k3r Registered

    Joined:
    Jun 15, 2016
    Messages:
    1,320
    Likes Received:
    3,121
    The question is how big the error in the model is and how much you have to invest to identify, quantify and fix a certain issue to get a few percent more realistic slip angles. You get to a point where you can ask yourself: is it worth it? Or might it be better to first fix more obvious issues, like the lack of tire damage, no downshift protection or the lack of a properly simulated drive train wich is genre standard and makes people downshift like crazy and leads to all kinds of stupid unrealsitic stuff. I haven't come across a single post yet wich indicates how big the problem really is and we are on page 40 now. So please explain to me, where would you set the priority considering that we are dealing with limited budgets. Everything I read is it might be this or that and here might be an issue - or maybe not - it's all very vague. Most people have a suspicion but nothing more. And it's another reason why I am not a big fan of those hotlap competitions. People drive thousends of laps - wich is pretty unrealistic by itself just to get a virtual car around a track in a very wierd way trying to exploit the system in every possible way, knowing all they have to do after getting it wrong is hiting escape and getting on track in a blink of an eye again. If this is less of an issue for you compared to something you can't even quantify for yourself, then why are we even discussing of what should be priority and what shouldn't?

    With the points I mentioned, it's clear that they will have an immediate impact because people are forced to take care of their car. Most of this stuff is allready available in other sims and proven to work well. And again, we are talking about a single driver who found a loophole while most of the coverage from the GT Pro series looks pretty damn good to me. Have you seen the big shunt in one of the last GT Pro races where half the field crashed at Radillion. Might not be as easy as it seems to keep the rF2 cars on their rails. Anyway, I am not trying to take anything away from your focus and wish to improve the physics engine. I am just questioning if there aren't more severe issues that will have a much bigger impact on how sim racers interact with their simulations in general. All this is much more of a philosphy issue with where this genre is moving towards more and more gamification. It's starts with people (even real racing drivers) being pissed off and not knowing what to do when they have to manualy use the pitlimiter. And the sad thing is that even some games and sims from 15 or 20 years ago did a much better job at punishing the player for setting a foot wrong. If you hit a tree in RBR your car was most likely toast. In most modern sims you can crash your car with 200mph and still get back to the pits.
     
    Last edited: Jan 24, 2021
    Nieubermesch likes this.

Share This Page