Poll: rF1 vs. rF2 tyre-model

Discussion in 'General Discussion' started by peterchen, Sep 27, 2016.

?

should rF1-tyre be possible (for modders)?

  1. only rF2-tyres please!

    67 vote(s)
    57.3%
  2. rF1-tyre should be possible

    50 vote(s)
    42.7%
  1. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    I was wrong, with current version camber affects rightly distribution of temps across tire.

    I tried again influence of camber on tire temps, I took the ISI C6R GT2 at Portugal, and made 5 laps with the same setup but camber on front wheels (-4,8 first stint, second stint 0.0).

    On first stint temps on long right before straight were 149, 79, 38 (inner, center, outer).
    On second stint temps on long right before straight were 81,105,124 (inner, center, outer).

    The last time I tried this on introduction of the new DW12(with CPM) I was unable to see this, center temp was always higher (even lowering pressures), and outer temps didn't follow.

    Now a very important feature for tuning cars works better, I tuned camber and tire pressures and improved lap times.
    I still find strange that on a long right curve which heats a lot tires, the three temps (inner, middle, outer) raise in paralell, I would expect the outer temp of outside wheel to grow quickly more than center, and inner to disminsh in relation due to the effect of roll on the contact patch.

    Sorry for the wrong statement, cheers.

    P.S. Checked DW12, it now behaves differently too, distribution of temps along thread is now different and more balanced (I don't know what version of RF2 did the trick).
     
    Last edited by a moderator: Oct 6, 2016
  2. green serpent

    green serpent Registered

    Joined:
    Feb 2, 2016
    Messages:
    704
    Likes Received:
    713
    I personally think the game would suffer from two different tyre models from a consistency point of view. It would seem odd if some cars handled in an rf2 style manner, and other cars handled in an rf1 style manner. They have such a different feel, and within the same sim it would effect the overall continuity, especially for an outsider who would one minute be driving an rf1 handling car, then the next minute have to adapt to the rf2 feeling. It would be confusing imo.

    I've driven AMS a fair bit lately, and despite trying to find a flaw in tyre behavior, I can't (not saying it's perfect, just saying I can't spot any real flaws). In rf2 however, I feel like there are flaws (at least, what my brain perceives as flaws.) The main thing that annoys me is when you breathe on the throttle mid corner, and the trajectory of the front of the car excessively dives toward the inside of the corner. I would love to know exactly what is causing this, but to my mind I rationalize it by thinking that power is applied to the outside wheel (which has more load on it), and a kind or torque vectoring occurs rotating the car. I think it's realistic to a degree, but just excessive - the front of the car is too unstable and it's hard to cut a smooth line under power. In saying that, I find that when I play other sims (which barely seem to display this behavior at all), it bothers me that I can't adjust my line with the right foot. I actually get a bit bored and find myself going back to rf2, which may not be 100% perfect, but is much more detailed, dynamic and rewarding.

    Not sure if anyone else can relate, just my thoughts.
     
  3. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    Automovilista has RF1 tire model perfected, and cars very well done.

    In the present state RF2 has already two different tire models, CPM tires and non CPM tires.
    And in both flavour of tires and cars, they can behave quite differently.

    For instance ISI BT20 and Flat6 don't have CPM tires, and they have nothing in common.

    If they were both models, and you dislike one of the tire models, just avoid playing with cars fitted with it.

    Cheers.
     
  4. Comante

    Comante Registered

    Joined:
    Nov 20, 2013
    Messages:
    1,694
    Likes Received:
    1,218
    If when you raise the throttle, the front close the turn, this has more to do with differential setup than with tires.
     
  5. Euskotracks

    Euskotracks Registered

    Joined:
    Nov 29, 2013
    Messages:
    1,092
    Likes Received:
    191
    I would say that the change in weight transfer from rear wheels to front wheels provokes the tighter turning radius when lifting throttle pedal.

    I think the behaviour is quite more noticeable when the slippage of front wheels is higher which occurs when you are close to the grip limit.


    Enviado desde mi ONE A2001 mediante Tapatalk
     
  6. stonec

    stonec Registered

    Joined:
    Jun 19, 2012
    Messages:
    3,399
    Likes Received:
    1,488
    CPM is just an incremental improvement of the existing tire model, it's only needed to simulate the tire better under extreme loads (you can read more about it in this interview). If rF2 moved to rF1 tire model, we can forget about realistic tire deformations and flatspots, those all come from the physics-based tire model, which ISI spent years researching.
     
  7. Will Mazeo

    Will Mazeo Registered

    Joined:
    Sep 12, 2015
    Messages:
    2,220
    Likes Received:
    1,578
    It's such a great thing Marcell never said rF2 is moving to rF1 tire model, isn't it? ;)
     
  8. green serpent

    green serpent Registered

    Joined:
    Feb 2, 2016
    Messages:
    704
    Likes Received:
    713
    The problem arises when I accelerate, not decelerate. However I did play around with some diff settings, and I found that the effect is reduced with an open diff. This demonstrates to me that it must be a kind of 'torque vectoring' occurring at the rear axles. Also, I'm quite sure I'm dialing in too much steering lock most of the time when I turn (hard to know how much to turn in with the G27), and this just makes the effect worse.

    Turned a few laps in the Megan last night, and once the tires were heated evenly and the track was rubbered in, the level of grip and feel was amazing. Not sure if the CPM had anything to do with it, but the placebo effect was certainly working it's magic.
     
  9. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    You say we can forget about tire deformations, I disagree on realistic.

    Yes ISI with a few tecnnicians (no doubt very skilled) spent years trying to design a very simplified model of what armies of high level specialized engineers have been perfecting with no limits on expenditure, computer power, and data available at tires manufacturers for various decades. This models represent the most valuable asset they have as a tire company.

    The fact that the tire doesn't react correctly to camber changes indicates in my view - once more I can be wrong - that the contact patch doesn't have a correct geometry under load when cornering.

    Did you notice if the steering was heavier with increased camber?
    It should, I can only notice the effect of caster.

    Cheers.
     
  10. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Doesn't or does? You seemed to say above it reacted correctly when you tried with different camber.

    As for your 'very simplified model' stuff, I don't get how a simpler model (rF1 style) is better than a more complex model that might still have some faults. If nothing else I don't know how you can measure it and come to that conclusion. It feels a bit like getting some 'noise' with a more precise instrument and seeing that as inferior to stable results from something less precise.
     
  11. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    What I say, is that RF2 tires react to camber, there is a difference when cornering in lateral spread of tire temps by varying camber.

    Normally spreading heat across thread should improve grip (it proves you have more thread applied on the track), in RF2 it is not the case, if you try to raise outer temps, you end running very very low camber (0.6 front) figures, and grip decreases.

    Try by yourself, take a good ISI car, C6R GT2 and experiment by yourself.

    What doesn't work in my view, is the fact that the three temps available raise in paralell when looking at the loaded tire across a curve.

    In real world you would expect outer temps of thread to rise faster than inside (due to weight transfer and hence roll, that is exactly what camber was invented for, to maintain a correct contact patch when cornering).

    On every car, camber is a compromise (if optimized for strong cornering situations, you will shave the inside of your tires in long straights, or mild cornering), on a road car that you bring on a track, you 'll end shaving very quickly the outer parts of the thread of your tires, because on sustained lateral loads, your car will roll too much, and have insuficient camber values then (even if you raise strongly tire pressure as you should before entering track).

    If I focus on this point of temps, it's because if we accept that temps on a point of a tire are consequence of the load at this point, it could mean that the geometry of the contact patch is not accurate, and hence steering is also off, and that is an very important matter for a car.

    I someone can correct me on this with facts I will gladly accept his explanations and stand corrected.

    "It feels a bit like getting some 'noise' with a more precise instrument and seeing that as inferior to stable results from something less precise."

    It'simply that I dislike figures with five decimals, that are may be 30 % off.

    Cheers
     
    Last edited by a moderator: Oct 11, 2016
  12. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Ok, so earlier you were saying camber does affect temperature correctly (at least somewhat), but the effect of those temperatures on grip doesn't seem right. Understood.

    The precision thing isn't a great example from me. To give a more relevant example: in rF1 there is an optimum camber angle, which gives x% extra lateral grip. There is also y% longitudinal grip loss determined by another parameter. If you currently have half that optimum camber, you can halve both the x and y values. Double, double. Negative, negative. The relationship between effective camber and grip gain/loss is either linear or follows a simple curve. rF2 presents the opportunity to produce more realistic results across a transition (from negative camber pushing in one direction, to negative camber pushing in the other, for example) by modelling how the tyre will behave in various scenarios, and interpolating between those known values given the current situation. If the grip curves should be quite non-linear, and even oscillatory in nature, rF2 could do that while rF1 couldn't.
     
  13. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    "If the grip curves should be quite non-linear, and even oscillatory in nature, rF2 could do that while rF1 couldn't."
    I have no doubts that a construction model can be superior, the question is wether such a simplified model can work well enough and how long will it take to perfect it.

    "but the effect of those temperatures on grip doesn't seem right"
    I would expect temperatures to be a consequence of the load a tire is subjected to, so what I suspect is off, is the geometrical shape of the contact patch when cornering.

    The armature of a real world tire is very complex and sophisticated to be able to maintain a wide tire thread evenly in contact with the road.

    Cheers.
     
    Last edited by a moderator: Oct 11, 2016
  14. rer8

    rer8 Registered

    Joined:
    Dec 19, 2010
    Messages:
    492
    Likes Received:
    23
    The tire model is one of the most important elements that distinguishes rF1 from rF2, and for that matter from most of the current racing games... It keeps drivers honest because the flat spotting of the tires is similar to real racing, and limits the capability of the car to be driven at speed. I for one will not download or use mods that have regressed to the old model... it is that simple. If the intent is to make a game and not a simulator, it is truly a sad day for rFactor 2.
     
  15. Guimengo

    Guimengo Guest

    What he is talking about is not about it being bad or something like it, it's that in the construction of the model, somewhere there is something that is miscalibrated, that is impacting tire behavior. It's like a F1 team having a wind tunnel vs a team with only cfd. Wind tunnel on top of cfd is better, but you don't always get the exact equivalent of real life.
     
  16. Domi

    Domi Registered

    Joined:
    Oct 11, 2011
    Messages:
    747
    Likes Received:
    44
    Compared to modding, it's more like a physics student running a F1 car in a wind tunnel and in cfd. Wind tunnel is better, but god knows what the student will do with it.
     
  17. TJones

    TJones Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,074
    Likes Received:
    257
    @Lgel - I understand your concerns about tyre temp readings.

    But i think we should take some things into account. First in RL motorsports it's not unusual to work with tyre temps around 10°C higher on the inner than on the outer, at the front axle, as long as the inner temps dont climp to high. To my knowledge these days (race)-tyre manufacturers try to get the tread as stable/stiff as possible, which reduces the effect of pressure on tread. Also the tyres, used today on modern GT-cars, with that very low sidewall compared to width ratio, might also flex not that much in lateral direction than for example a F1-tyre.

    But, all this above in mind, i still think flexibillity (tread and sidewall) is indeed pretty low. I'm not the guy to evaluate it's still realistic or not.

    You shouldn't be to concerned about it. :)

    BTW: There's a new Blog entry by Micheal Borda!
     
    Last edited by a moderator: Oct 11, 2016
  18. Euskotracks

    Euskotracks Registered

    Joined:
    Nov 29, 2013
    Messages:
    1,092
    Likes Received:
    191
    Michael's blog is very interesting. A good student with good real data should be able to correlate with the tool results. The procedure is very well explained and the tool to perform virtual tests is already there.

    The only thing I miss is to clearly distinguish the procedure to be followed in the two basic scenarios I can think about...

    1. Real data.
    Explaining what set of data would be need from real tests at each step to properly create the virtual tire model.

    2. rF1 tire model data.
    In this case we would have no real data but we have the rF1 tire model that we would like to replicate.
    The idea would be the same. Explain what data is needed to correlate at each step. It should be noted that this should complement the first scenario to create the TBC values to be used by the IA.

    In both cases, the procedure should explain how to check that both models correlate for each test in the procedure. It should also give some tips of the parameters to be touched in the rF2 tire model to achieve the goal for each test.

    I would add that after fulfilling the whole procedure, it will probably be necessary to repeat the procedure to fine tune all parameters. Despite some variables have significantly more influence in some results, their complex interdependency might provoke a worsening of the correlations after adjusting other parameters in later steps of the procedure. This iterative method usually does not need to be done more than one or two times. If the tire correlation procedure is properly done, it should converge to a unique solution. If not it will diverge and the parameters would ask for very different values at each iteration of the procedure.



    Enviado desde mi ONE A2001 mediante Tapatalk
     
  19. Lgel

    Lgel Registered

    Joined:
    Jun 24, 2014
    Messages:
    1,267
    Likes Received:
    365
    Had more time to test with DW12.

    Actually was able to obtain a quite good spread of temps across tire on front wheels, with outer temps of loaded front tire increasing faster than inner along curve, with reasonable values for camber. Rear tires weren't as good but much better (center temp still too high even with minimum pressure).

    So I suspect that suspension geometry or it's implementation on certain cars may be inacurate.

    For the moment I believe this car has one of the best combination of suspension geometry and CPM tires.

    Cheers.

    P.S F1RFT 2013 1.35 with FISI 2012 CPM tires, ISI FISI 2012 seem also to work quite well. URD C7R even if at too low values of camber, with some suspension tuning (ARB affect also spread of temps) behaves very well.
     
    Last edited by a moderator: Oct 26, 2016

Share This Page