[ODE] Car Physics

[Steve Eckles]

> -----Original Message-----
> From: ode-bounces at q12.org [mailto:ode-bounces at q12.org] On Behalf Of Brent
> Burton
> Sent: 19 April 2004 17:08
> To: 'ODE Mailinglist'
> Subject: RE: [ODE] Car Physics
> 
> On Mon, 19 Apr 2004, Steve Eckles wrote:
> > Thanks, I'll give it a try. Any other suggestions? Do you have any code
> I
> > could look at?
> 
> (jumping in here...)
> 
> Other suggestions for roll-over:
> * CG is just too high for the car.  Place it about as high
>   as the top of the tires.  That is close to where many sedan
>   CGs are.


I tried this but I made very little difference! I'm using
dMassTraslate(0,-2,0) - that's right isn't it?


> * Springs are too soft.  Increase spring rate, and increase dampers.

Again, I've tried this too, with very little success! I'm just using the
suspension parameters of the hinge2 joint. Should I bee doing it a different
way?

> * Any anti roll bar simulated?  Are they strong enough and do
>   the forces act in the proper direction?

No I don't have anti-roll bars. How do I do this? Where can I find out more
info?

> * Too much tire grip.  Real tire grip changes with the camber and
>   slip angle.  Grip (holding forces) increase up to some threshold,
>   the decrease again after that threshold.  "Slip Ratio" and "slip
>   angle" are good terms for a websearch.


Again, where can I find info on this? Currently, the only way I can get it
anyway stable is to have the surface very slippy (more like an ice rink!),
otherwise it just flips over all the time!


> The source code for an older version of Racer (0.5.0f) is on
> the Racer website.
> 
> Another simulator is Vamos.  I've not tried it, but it's an
> open-source project and you can get the source from
> http://vamos.sourceforge.net/
> 
> Both Racer and Vamos use the Pacejka tire force model.
> 
> The "pendulum model" isn't a good way to stabilize the car.  It
> adds considerable unrealistic weight, and in turns the car will
> lean opposite reality (it will lean like a motorcycle).  A
> pendulum car will also nose-dive during acceleration and ride
> nose-high during braking...  Hmmm, sounds like helicopter dynamics.
> 

Adding a pendulum has actually been the best idea so far! It is obviously
very wrong and a dirty hack but at least the car stays upright most of the
time! The biggest problem now is that the front steering wheels get knocked
about all the time so it's impossible to get it to go in a straight line and
is generally very unpredictable! It works OK if I make the terrain
completely flat.

Steve