Tire issue - Grip and relationship to load

Discussion in 'Car Modding' started by machine, Nov 14, 2013.

  1. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    I have been testing early versions of Nascar SCE mod on some ovals. The grip appears highly exaggerated on high banked ovals. Tested at Charlotte, Homestead, Dover & Calder Park Thunderdome. (Australia) All tracks in the 18-24 degree banking range. Massive grip, I can run flat all the way round. Tested at New Hampshire, Richmond & Kentucky, the grip felt pretty much spot on. I did notice that on all tracks, the grip appears to go away on the straights, specifically much more on the high banked ovals. With obvious early tyre physics I was wary, so I tested an ISI car, the C6R and got exactly the same result. It appears as if load on the tyres caused by centrifugal force on the high banking has way too much effect. Any one have any ideas on this?
     
  2. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    Just to add, I did see this in the tyre file, only thing resembling load.
    RubberPressureSensitivityPower=(-0.075,5000,500000,1)
     
  3. mschreiner

    mschreiner Registered

    Joined:
    Oct 5, 2010
    Messages:
    390
    Likes Received:
    262
    Could ISI reflect on this issue?

    I have tried a lot of different things to eliminate having too much grip on the high banked tracks. It seems like the car is super glued to the track going through the corners flat footing it all the way around the track. I'm hitting speeds of 185 mph on our Homestead track which is around 24 degrees of banking. While going through the corner the tire deforms a lot, and in the real world it would pop off the rim. I have even lowered the grip small amounts at a time and it just makes the car drive horrible on the straights.

    I actually took the formula1 tire ISI made and made it the size of a NASCAR tire figuring this would be the closest tire being a slick.
    The tire feels really good but it just gains a ton of grip going through the high banks as stated before.

    Here is what would be nice to help all modders out with tire design as we are not Tire engineers and it is very hard to decipher some of ISI's comments in the PDF.

    Could you add comments that we can understand better for each line in the TGM file?
    example:
    RubberPressureSensitivityPower=(-0.075,5000,500000,1) //(power, offset, nominal maximum, normalize (1=yes,2=no)).
    What the heck does that really mean? Power? What is it really and what does the 1st number do?//Offset? Offset what and what is it for? Makes it hard to adjust if we do not know what it really does. //Nominal maximum? Really? Who is really supposed to know what this is or does?, Normalize kind of makes sense but not really since we do not know what the other numbers really do.

    If ISI could help us understand their game better it would get more modding done in my opinion, and I have been Modding rFactor1 since 2005 so it is not like I'm dumb, or am I? LOL
    I think the future of rFactor 2 depends on modders making good content for everyone and right now I do not see that happening unless they help us understand their game better to put out Mod's that are quality and attract customers to rFactor2.

    Please please help us ISI because this could not have been made any more difficult.
    At this rate it is going to take us 6 months or more just to try and get the tires to work properly and understand them. Most people trying to mod this will give up after a few weeks and I can understand why at this point.

    I am willing to put a lot of time and testing into this if we can get some help. We really want to start running on the rFactor2 platform but right now we just can not make the move from rF1.

    Regards,
    Mike
     
  4. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    "RubberPressureSensitivityPower = Rubber contact pressure sensitivity (power, offset, nominal
    maximum, normalize (1=yes,2=no)). This adheres to the following formula,
    (BristleContactPressure+Offset)^Power, and if normalized (recommended) will provide a peak
    multiplier of 1.0 at 0 contact pressure."

    From the tyretool quickstart V3 pdf, available in the dev corner.

    *Hmm... you mention PDF... maybe you only have the original version, which I think had fewer comments in it? Anyway, contact pressure is obviously the input, the small negative power can only mean it's a decreasing value with increasing input (so can presume it'll start at or near 1 and reduce with increasing load - so then the normalize option makes sense, to force a 1.0 start), if you try a couple of series/graphs and offset the pressure you can see what effect it has, and there's a nominal maximum - so higher pressures have no more effect.

    Not one of the harder lines to work out, I thought.
     
    Last edited by a moderator: Nov 15, 2013
  5. Nils

    Nils Registered

    Joined:
    Jan 12, 2012
    Messages:
    61
    Likes Received:
    0
    from the KnownIssues.txt

    -Contact patch load distribution model is not final, which means tire pressure does not have the full effect that it should.
    -There are other relatively minor loose ends.

    I think that this also affects camber which in my opinion does also not work as it should, so if you add banked tracks these problems could add up. At the latest when the Indy stuff comes out they hopefully will have that fixed
     
  6. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    So you are saying that this example RubberPressureSensitivityPower=(-0.142,10500,550000,1)
    would go like this...
    (55000+10500)^-.142=.152652 as the first value, then normalised and it increases with increased downward pressure on the tyre either from braking - actioned as downward force on the front wheels, or driving around a highly banked corner until it reaches 1. That would give maximum grip. Makes sense, because the initial grip is .15 and increases to 1 in a 24 degree banked turn. Which is a bit much really. Explains why we are getting too much grip in the corners.
     
  7. Joel.Brown

    Joel.Brown Registered

    Joined:
    Jan 16, 2012
    Messages:
    107
    Likes Received:
    88
    We have the latest PDF with the formula you explained and have done graphs too.

    The problem still comes down to high banked tracks have way too much grip and the explanations of the parameters are terse. A couple more sentances explaining their effects would help. That's what the community ends up doing for ISI as its done in the past. I guess we're hoping we wouldn't have to go through that process again.
     
    Last edited by a moderator: Nov 15, 2013
  8. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Na, I think you missed where the input pressure goes. You've thrown 550000 (typed as 55000) in there which is the nominal maximum pressure.

    So, start with 0 pressure and your figures, then a few bigger pressures [rubber pressure, not tyre pressure!]:

    (0+10500)^-.142 = 0.269
    (5000+10500)^-.142 = 0.254
    (10000 + 10500)^-.142 = 0.244
    (15000 + 10500)^-.142 = 0.237

    etc etc.

    So with increasing input load, your grip return reduces. If you leave the normalize option off it'll give you grip values as above. If you put normalize on, the first figure (at 0 input) is made 1.0, and the whole series increases by the same magnitude. 1.0 / 0.269 = 3.717, so all grip values will be multiplied by 3.717, giving you 1.0 grip at 0 load and less grip as the load increases.

    Of course this grip multiplier is only a part of the equation, not the overall grip, so increasing load will increase grip - just not in a linear fashion. The question you're asking is probably whether the grip drops off fast enough with load.

    If you want to see what happens when you leave the initial grip about the same but it doesn't ramp up as quickly on banked ovals, make the power above a larger (negative) number. If that gets you roughly what you want but you need to tweak it a little, change the offset. Maybe that's the solution, maybe it's some other parameter, or maybe there's something else that is giving too much grip for you. Maybe not even ISI knows without doing some testing.

    Sure, ISI could throw in an extra sentence or two explaining in very simple terms what the whole line actually does (though not in any specific situations, because the game doesn't try to shape its behaviour to fit certain circumstances), but I would still suggest anyone reading the name of the line and the brief explanation of the parameters should probably be able to work it out - and if you can't, throwing numbers blindly at it if you don't understand it probably isn't much good either.
     
  9. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    Thanks Lazza

    This gave me something to test. I played with the figures to get extreme results in both directions which is displayed by this graph.
    View attachment 10805

    Both extremes produced exactly the same feeling as with default setting, exact same lap times also. Flat all the way round at Dover. Car tends to slide a bit on the straights, but grips like a F1 in the turns.
     
  10. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    double post... damnit
     
  11. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Hmm... interesting... lol

    What sort of tyre loads were you registering through the turns?
     
  12. speed1

    speed1 Banned

    Joined:
    Jul 26, 2012
    Messages:
    1,858
    Likes Received:
    0
    Couldn't the line ( RubberPressureSensitivityPower ) have done something with air influence ? Pressure ( efficiency of the compression ) on the rubber from inside of the tire, compression forces maybe ?

    With increasing load the air get compressed higher, therefore the pressure on the rubber and resistance again Impacts from the surface. The airtightness and resistance are important as well.
     
    Last edited by a moderator: Nov 16, 2013
  13. mschreiner

    mschreiner Registered

    Joined:
    Oct 5, 2010
    Messages:
    390
    Likes Received:
    262
    I'm working with ISI to help get some better explanations. I think if we can tackle this and help all modders understand the TDF file better then more will try to mod rF2 properly without getting frustrated with a guessing game on a lot of stuff. Just having one line explained to me in a PM helped tremendously to understand what it does better to allow me to test more accurately.

    I will share more as I get more information.

    Lazza,
    Thank you for your input as well. This stuff can turn the average Modder away from modding pretty fast. LOL
    I'm glad we have a community to help each other out as it will benefit rF2 in the long run.

    Off to bed and will be testing tomorrow!

    Mike
     
  14. speed1

    speed1 Banned

    Joined:
    Jul 26, 2012
    Messages:
    1,858
    Likes Received:
    0
    Unfortunately it is not in german. :(
     
  15. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    Doing some searching I found this article. http://isiforums.net/f/showthread.php/8328

    JTbo suggests in #9 post that tyres gain massive grip when used on a heavier car. This probably linked to this issue. Extra load on tyres is producing too much grip and from tests I have done, it seems impossible to change. If it is a setting in tyre files that can be changed, then every mod made so far is going to have to go back and edit their tyres.
     
  16. Rony1984

    Rony1984 Registered

    Joined:
    May 10, 2012
    Messages:
    500
    Likes Received:
    28
    interesting stuff. Please ISI comment on this
     
  17. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Ok, so I can see two potential issues here. I don't have time right now to do some testing on the TGM realtime line discussed above.

    1. The release notes say the "-Contact patch load distribution model is not final, which means tire pressure does not have the full effect that it should." This could be taken to mean that lowering tyre pressure and/or increasing its load, which will obviously increase the raw size of the contact patch, may not have the expected effects on the contact patch load pattern. If you double the weight on a tyre obviously in a very rough sense you have double the rubber on the road, but a lot of that rubber will actually have less load on it (than other parts) which will reduce the increase in grip. It's hard to know exactly how far away from target this area is because in the latest Q&A they made it quite clear that while it isn't final, there is a model in place. If that's the case you wouldn't expect some of the rather strange results some people seem to be getting around the place.

    2. Even if that's the case, it should be possible to offset this effect by playing with the RubberPressureSensitivityPower line above. But obviously if you play with these parameters to give very little grip with load it should be noticeable - and again we have people saying it doesn't make a difference. I would be interested to try a normalized curve that drops away extremely quickly from the start value (so that by the time load reaches values in the hundreds the [normalized] rubber grip multiplier will be in the order of 0.01 or lower), but also a non-normalized curve that starts at 1.0 and then drops away, to eliminate the possibility of something weird happening with normalization.

    Also I think it's important to do tests that are measurable. Having cars that can be run flat out around a banked curve, then changing some parameters and finding they can still run flat out around that curve, doesn't really reveal much. It's a bit like trying to pick up a car with your bare hands (and obviously not being able to) then halving its weight and still not being able to pick it up. There's a marked change but no difference as far as the measurement goes. So I think it's better to zero all aero effects on a test car and measure lateral g-forces when playing with these figures.
     
  18. machine

    machine Registered

    Joined:
    Jan 31, 2011
    Messages:
    399
    Likes Received:
    295
    With all tests so far, the line has been virtually flat by the time a large load is reached in the corners. So I devised a formula to get the maximum steepness on the curve at the top end of the curve. Still same, no dropoff in grip. 3 is the new test figures with a higher dropoff.
    View attachment 10854

    Yes, a high banked oval, but all the high banked ovals we tested at had the same issue.
    A car as heavy as a Nascar cannot possibly corner at 180mph at Dover. Not and still have the ability to turn the car down on to the apron at any time mid corner.

    More to think about anyway.

    EDIT:
    I tried reducing StaticBaseCoefficient=1.7285 & SlidingBaseCoefficient=1.1768 to half what they were before. The grip was definitely less at slow speeds and on the apron. Way less grip. But once up to speed on the banking, still same maximum, flat out grip. Once I touched a left front wheel on the apron there was an immediate loss in grip as 2 wheels rf & lr got light. This was now much more pronounced.

    ISI like a duck on water. We can't see anything going on. But those little legs...
     
    Last edited by a moderator: Nov 19, 2013
  19. speed1

    speed1 Banned

    Joined:
    Jul 26, 2012
    Messages:
    1,858
    Likes Received:
    0
    Honestly i'm not in this thing. I've read it here and have had the first thoughts written down. I thought maybe it helps someone somehow.

    I neither understand nor can I visualize something of the context of the sim engine. It's not really possible to understand the simulation and the reality will not much help further imho but when I read about, the presumption is perhaps not so wrong, at least it stays somehow in connection.

    How much of an influence has the air pressure on the overall performance of the tires in general ?

    edit: another thing comes in my mind when taking a tire compound as a starting base. It might be possible something collides with airvolume, compression magnitude when changing the diameter and such.

    The thing is if flexibility of a body is simulated which is filled with air then you must also include the air medium as such. The density, the oxygen content, nitrogen, ....... and their behavior under the influence of heat and pressure.


    As long as the medium is not treated properly, you can forget tires imho.
     
    Last edited by a moderator: Nov 19, 2013
  20. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    I'm not sure I follow your formulas... struggling to replicate your values on those curves. Have you tried a normal power (like your -0.142) and an offset of 1? For my thinking that'll reduce grip to 34% at 2000 load, and if I use those values I absolutely get a lot less grip at any speed (not surprising, since the weight of the car obviously has an effect).

    What sort of tyre loads are you getting during your banked corners?
     

Share This Page