rec.autos.simulators

very interesting infos here !

Let me summurize, correct me if i'm wrong

Okey so for a couple tyre/surface we have 3 parameters

Grip
Longitudinal viscosity
Lateral      viscosity

Longitudinal_Force = F_long( slip_angle, slip_ratio ) * Grip * Load
                     + Long_visc * Long_speed

Lateral_Force = F_lat( slip_angle, slip_ratio ) * Grip * Load
                + Lat_visc * Lat_speed

Of course long_visc is much smaller than lat_visc and close to zero except
for surface like sand/water.
We can also say that those viscosity coef depend of tyre depth in the
material.

well thanks everybody of those very usefull informations !

regards,

--
Seb
Game Developer
http://www.racesimcentral.net/
GPLRank -41.68
http://www.racesimcentral.net/

yup, that's a simple theory I would start with if I did this. Although,
what you name Long_speed should not be the speed of the tyre but the
slip speed (i.e. Long_Speed - Ang_Vel*R), the same that enters the
slip_ratio equation. But you probably know that already :).

Also, I have absolutely no idea how Lat_visc should change with Load,
but making it proportional to the Load on the tyres is probably a good
initial approximation (much like the usual grip).

-Gregor

  Good thinking, Dave.  My hack attempt at a brush model doesn't let the parts
of the ground surface move.  This would probably warp the curves more towards
what the original poster was indicating.
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com

Well,i really meant longitudinal speed. long_visc will be a resistive force
. It will be Null on tarmac, and very high on deep sand. It's a viscosity
and not a friction force. I mean if you'r driving in sand/water without
spinning there is high resistive force because of fluid's mechanics issue.
In fact its more complicated because long_visc must be dependant of wheel
rotation. It's obvious that longitudinal viscosity is lower when the wheel
is rotating.

Again, because we are talking about viscosity, i thought that lat_visc would
be more proportional with the tire depth in the material than the load.But
it's obvious that tire depth in material is proportional to the load, but
the funcion depth = F( load) is different for every surface.

So, the funcions could be this :

Longitudinal Grip Force      = F( slip_ratio, slip_angle ) * Grip * Load
Longitudinal Viscosity Force = ( F( slip_ratio ) + long_speed ) *
                               Long_visc * F( Surface,Load )
Longitudinal Force = Longitudinal Grip Force + Longitudinal Viscosity Force

Lateral Grip Force      = F( slip_angle, slip_ratio ) * Grip * Load
Lateral Viscosity Force = lat_speed * Lat_visc * F( Surface,Load )
Lateral Force = Lateral Grip Force + LonLateral gitudinal Viscosity Force

So, what do you think about this ?

--
Seb
Game Developer
http://www.eden-studios.fr
GPLRank -41.68
http://magicfr.multimania.com

I would introduce two components, one (typically smaller) that is
proportional to the longitudinal velocity itself, and another one that
has a higher viscosity coefficient and is proportinal to the
longitudinal slip velocity. I think this is something you took into
account in the following equation

but I would actually write it as

Longitudinal Viscosity Force = ( long_slip_speed * Long_visc +
long_speed * Long_visc_small) *
                               F( Surface,Load )

where

long_slip_speed = long_speed - Angular_velocity * R

-Gregor

hehehe, it's very intersting because this is just what i had modeled after
sending my previous post :o)
Moreover long_slip_speed info is very usefull for all Graphicals effects,
tire skids, tarmac burning smoke, gravel projections,etc.

Thanks for helping me,

regards,

--
Seb -
Game Developer
http://www.eden-studios.fr
GPLRank -41.68
http://magicfr.multimania.com

Be sure to let me know how it turned out!

-Gregor

I'll surely post again about car physics on r.a.s :o)

--
Seb -
Game Developer
http://www.eden-studios.fr
GPLRank -41.68
http://magicfr.multimania.com

Is it V-Rally 3 you are developing now?

/Carl

V-Rally 2 (utter garbage) and NFSPU (excellent game) from the same
developers? Nah, can't be..

Andre

NFSPU PC was made by EA ... wich "borrowed" our body deformation alogorithm,
and NFSPU PSX1 was made by my Co.

VR2 PC is an adaptation from the DC version, and the kernel was'nt made at
all for heterogeneous platform as the PCs.. with got lot of problems to
making works dynamic kernel . And i agree with you VR2 PC is crap. But we
sold 1 million of VR2 PSX1, wich i'm proud of. The only Racing game on PSX1
with modelised driver and co-driver .

By the way, i think that NFSPU PSX1 is more interesting to play that the PC
versions , tracks tighter, more actions. In My Humble Objective Opinion.

And about VRally3 ... no coment .. ;o)

regards,

--
Seb -
Game Developer
http://www.eden-studios.fr
GPLRank -41.68
http://magicfr.multimania.com

/Carl