1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

PC Self-aligning torque (SAT) relationship vs slip angle and camber

Discussion in 'Assetto Corsa' started by David Dominguez, Jun 5, 2016.

  1. @kunos

    Hi all, I open this thread to discuss about the effects of self-aligning torque in AC tyres. For those who don’t know what it is:

    Self-aligning torque, also known as aligning torque, SAT, or Mz, is the torque that a tire creates as it rolls along, which tends to steer it, i.e. rotate it around its vertical axis. In the presence of a non-zero slip angle, this torque tends to steer the tire toward the direction in which it is traveling, hence its name.


    So, I made a test with the McLaren 650S GT3 and default setup, and recorded telemetry from it. The process was very simple; accelerate until 4th gear, hit the max speed, and start turning slowly. This is the result:

    [​IMG]

    So, few things to talk about. First and more obvious, the SAT curve (top left chart) goes below zero very early, even before reaching the slip angle peak where the lateral force is at maximum (top right chart, peak happens about 6 deg of slip angle). In this case, it reaches zero at just ~3.6 degrees.

    Next, here I compare two runs. One is with cambers at -1 degrees (white), and the other one with -3.5 deg (green). The difference in the buildup part of the curve is quite important, as it already produces significant torque even at 0ish slip angle, and peaks very soon, around 1.5 deg (and the FR tyre is already quite loaded, about 7 kN). Also, the run with lowest camber seems to produce higher total torque, which seems strange. Basically, compared to the low camber run, running higher cambers seem to shift the curve left and down. It could make sense to shift left as higher camber should increase lateral forces in the buildup area of the curve (although it's not very noticeable in this test...). But producing lower torque is weird I think.

    [​IMG]



    Here are SAT curves gathered from Internet, sadly there are no comparisons at different camber angles.


    [​IMG]

    [​IMG]


    [​IMG]

    [​IMG]


    [​IMG]

    [​IMG]


    One might argue that tires are different, and they might not be racing slicks like in the McLaren GT3, but the effect happens as well in the street tyres of the 458 Italia for example (same two options again, low camber vs default camber)

    [​IMG]

    So, in conclusion. I can’t say for sure that “this is wrong”, but to me it looks like the shape of the curve drops too fast. Most of the data shows an aggressive dropoff past the peak, as expected, however in most cases it approaches 0 way later, and in some cases it doesn’t even go below 0, depending on the load, etc. I think this explains a bit (personal opinion) the springy feel of AC FFB, making it go full force as soon as you start steering (though this also depends on the suspension geometry), and already drops an important amount of force when you are cornering at grip peak.

    I don’t know if Stefano reads RD anymore, so please feel free to open a thread in AC physics bugs and issues sub forum and copy paste this post, as I am banned from their forum. He confirmed when I asked him in the last streaming that the SAT curve shouldn’t go below 0 earlier than the lateral force slip angle peak, so he might want to check this.

    Sorry if you are totally confused with this thread, I am a bit as well, and I sometimes struggle to put order in my explanations :D
     
    • Like Like x 15
    • Love Love x 1
    • Wow Wow x 1
    • Agree Agree x 1
  2. kunos

    kunos
    Kunos Simulazioni

    you are right David, this needs a new formula. I am rewriting most of the tyre model in order to be able to be more "editor friendly" and to allow us to look at these stuff a bit more easily.
    In order to get the torque inversion the trail formula was wrong and the falling slope was too high... thank you for pointing this out.
     
    • Like Like x 23
    • Love Love x 2
  3. burrito

    burrito
    It's an opinion, nothing more, nothing less. Premium Member

    Does that mean more variables to control it or a rework of the code behind the current variables?
     
    • Like Like x 1
  4. kunos

    kunos
    Kunos Simulazioni

    no no, I've just fixed it in the formula, no additional parameters.
     
    • Love Love x 2
    • Like Like x 1
  5. burrito

    burrito
    It's an opinion, nothing more, nothing less. Premium Member

    Great. Thanks for this correction.
     
    • Love Love x 1
  6. larsbengtsson25

    larsbengtsson25
    Premium Member

    good response Kunos.
    And very good Burritos.
     
    • Agree Agree x 1
  7. Nice work, David. Thank you for bringing this information to the masses. I'm not going to pretend I totally understand all of the in-depth physics involved in creating a racing sim, but seeing data like in your post, with plain English explanations of what's going on certainly helps a half-wit like myself! :D
     
    • Like Like x 1
    • Agree Agree x 1
  8. @David Dominguez the SAT curves you are pulling from the web are all pneumatic trail only ? Zero caster?

    Generally how does the mechanical trail vary with increasing steering angle on a GT3 car?
     
  9. Qazdar Karim

    Qazdar Karim
    Premium Member

    This is from a F1 car suspension steering simulation:

    [​IMG]
     
    • Like Like x 1
  10. Thanks! Am I wrong seeing that the mechanical trail dominates here and the effect of the fix on the actual FFB will be quite small ?
     
  11. It depends on the geometries ... for example I know that some old racing cars that had no power steering ran very little caster/mechanical trail, that kind of cars should benefit in AC with a better SAT implementation.
     
    • Agree Agree x 4
    • Like Like x 2
  12. Indeed. Even way back in the 50's and before, they ran positive camber, partly because the wheel would be too hard to turn, lol.
     
    • Like Like x 1
  13. LilSki

    LilSki
    Premium Member

    Correct me I'd I'm wrong but I don't believe camber has any effect on steering effort. Caster is what does that.

    I like this thread. It shows if you present something with intelligence and with no snarky attitude, you get an intelligent, non snarky response.
     
    • Like Like x 2
    • Agree Agree x 2
  14. It does actually, since there's less contact patch on the outside tyre. Hence less "grip". If that makes sense. There are other reasons they did that, but steering effort is actually one. Also why they had ENORMOUS steering wheels.
     
    • Like Like x 2
  15. Qazdar Karim

    Qazdar Karim
    Premium Member

    Normally camber should affect FFB, but the mechanical trail's effect is much bigger :) (the way i say it is not 100% correct, because it's a bit more complicated than that :D like always)
     
    Last edited: Jun 7, 2016
    • Like Like x 1
    • Agree Agree x 1
  16. LilSki

    LilSki
    Premium Member

    Yeah OK that makes sense. In my own experiments on my real car, when I first put it on the road my TA had zero camber and about 3* of caster. When I got my alignment machine I added 1* of camber and there was little to no change on steering effort. When I got my new control arms which added 5* of caster I set it up with the same 1* of camber but it now had 7* of caster and the steering force was significantly increased. I know my numbers are low when compared to a race car but it all should work the same. So camber does have some effect on steering effort but very little when compared to caster.
     
    • Like Like x 1
  17. Stereo

    Stereo
    Premium Member

    On the other hand they ran such differently built tires that modern data can't even relate (for example, bias ply is known to peak at much higher slip angles, and has much closer to 0 ideal camber)
     
  18. Well, the cars that I had in mind already had slick radials :)