Author Topic: Tire model  (Read 2106 times)

HornetMaX

  • Hero Member
  • *****
  • Posts: 5498
    • View Profile
Tire model
« on: June 07, 2014, 10:39:44 PM »
OK, I think I've finally found a reference that covers most of the parameters present in the tire mode of GPB: http://mech.unibg.it/~lorenzi/VD&S/Matlab/Tire/tire_models_pac2002.pdf

Also useful: https://www.tut.fi/ms/muo/vert/11_tyre_as_car_component/

Some things I still don't know or I'm not sure about (parameter names from the tire file are underlined):

  • Fz0 / SpringBase = nominal tire deflection (at optimal pressure) ?
  • SpringFactor: does this alter the SpringBase depending on the delta from the optimal pressure ? If yes, how exactly ? Just SpringFactor * delta_pressure ?
  • RadiusRPS: is this related to the variation of the effective rolling radius with respect to the tire rotation speed ? If yes how exactly ?
  • WearRate: out of curiosity, how the wear is calculated ?
  • NumKeys, Key0, Key1: what are these ?
  • dry_grip, wet_grip: do these scale the D coeff (peak value) in Pacejka MF ?
  • max_wet: what's this ?
  • curve_type: what's this ?
  • resistance_dry, resistance_wet: are these scale factors for q_sy1 ?

MaX.

HornetMaX

  • Hero Member
  • *****
  • Posts: 5498
    • View Profile
Re: Tire model
« Reply #1 on: June 08, 2014, 02:10:16 PM »
After closer inspection, the 1st link above seems to refer to a model less complete than the one used by Piboso: coeffs p_Cy2, p_Ey5 and p_Ky6 (al from lateral force at pure slip) are all missing in the link.

MaX.

PiBoSo

  • Administrator
  • Hero Member
  • *****
  • Posts: 2086
    • View Profile
Re: Tire model
« Reply #2 on: July 11, 2014, 01:48:41 PM »

GP Bikes uses Pacejka 2002 motorcycle magic formula.

HornetMaX

  • Hero Member
  • *****
  • Posts: 5498
    • View Profile
Re: Tire model
« Reply #3 on: July 11, 2014, 02:01:15 PM »

GP Bikes uses Pacejka 2002 motorcycle magic formula.

Thanks for the info. That's should then be chapter 11 (or 10, depending on the editions) of Pacejka's book "Tyre and Vehicle dynamics" (that I have at hand). Questions above still stands though ...

MaX.

LOOPATELI

  • Hero Member
  • *****
  • Posts: 606
    • View Profile
Re: Tire model
« Reply #4 on: February 06, 2017, 04:01:54 PM »
I think I found what is numkeys for, it's something related to the tyre grip and the wear.
I've been doing some test and this is what I found:
Code: [Select]
NumKeys = 2
Key0 = 0.900000, 0.980000
Key1 = 1.000000, 0.800000
The first value is the tyre wear and second value is the tyre grip percentage for the specific compound. Between Key0 and Key1 the grip value is calculated using a linear interpolation I guess. And by default if you don't add a key on 0% wear and 100% wear the values are 1 and 0 on each.

Here is a Graph:





So I decided to make a custom tyre for a WIP mod:


Code: [Select]
NumKeys = 6
Key0 = 0, 0.95
Key1 = 0.05, 1
Key2 = 0.4, 0.95
Key3 = 0.6, 0.9
Key4 = 0.8, 0.8
Key5 = 1, 0.75



Piboso can you confirm this?

PiBoSo

  • Administrator
  • Hero Member
  • *****
  • Posts: 2086
    • View Profile
Re: Tire model
« Reply #5 on: February 06, 2017, 05:22:24 PM »

Alone

  • Sr. Member
  • ****
  • Posts: 303
    • View Profile
Re: Tire model
« Reply #6 on: March 28, 2017, 09:30:03 PM »
Hi. Have some questions:
1: Is the springbase expessed in N/m?
2: That value is without pressure?
3: is the springfactor related to the increase of stiffness acording to the pressure? I.e.: 250(spring factor)200(pressure)=50000+50000(springbase)=100 kN/m
4: how can be the damper calculated?

PiBoSo

  • Administrator
  • Hero Member
  • *****
  • Posts: 2086
    • View Profile
Re: Tire model
« Reply #7 on: March 28, 2017, 09:38:11 PM »

1: Yes
2: Yes
3: Yes
4: A starting point could be 0.35 * critical damping, where critical damping is 2 * sqrt( wheel_mass * tyre stiffness ).

Alone

  • Sr. Member
  • ****
  • Posts: 303
    • View Profile
Re: Tire model
« Reply #8 on: March 28, 2017, 09:39:47 PM »
Thank you, really apreciate :)

Alone

  • Sr. Member
  • ****
  • Posts: 303
    • View Profile
Re: Tire model
« Reply #9 on: April 07, 2017, 03:34:14 PM »
I observed, playing around the radial stiffness, that the bike can handle the kerbs and track ondulation (a bump, i.e.) nicely while the bike is straight. But the problem apears while is turning. Is the Kl (lateral stiffness) simulated? If not, could be possible to add in the future?