Erik wrote:

> Thinking about it a bit more this makes sense.
> Calculating every wheel separately isn't the whole story. In the end
> there is the friction caused by the complete landing gear which isn't
> wheel spin dependent.


I've considered this. It's possible I didn't code it right. I got the idea
from this paper (AIAA 4303), excerpt:

"Probably the most difficult ground dynamics phenomenon to simulate in
real-time is stopping the aircraft. When the aircraft stops the friction
forces disappear, and the aircraft cannot be moved or turned by using the
brakes or nose wheel while it is stopped.  If the usual equations are
applied continuously the simulated aircraft backs up or turns
unrealistically when stopping.  Several ground dynamics  functions,
especially the skid angle, depend on dividing by the speed of the aircraft.
If the speed is allowed to go to exactly zero the program will crash due to
a 'divide by zero' error.  Even if speed only decays to a small number,
unrealistic seemingly random results occur.

The classical way to solve this problem is to set the velocity directly to
zero inside the equations of motion when the velocity approaches zero.
Since the LaSRS++ architecture does not allow the landing gear model to
modify the velocity of the aircraft, it can only output forces and moments
that are integrated to produce translational and rotational velocities.

In the LaSRS++ Generic Gear Model the axial and lateral friction forces are
decayed to zero by making them proportional to forward and lateral velocity
at low speeds.  This prevents the aircraft from backing up, sliding
sideways, or turning after it stops.  If there are crosswinds producing
aerodynamic side forces, the gear model will produce enough side force to
keep the aircraft static, up to the limit of side force friction.  Only when
the forward thrust of the engines exceeds the static friction will the
aircraft start to move again."

I have the email address for the guy who coded the LaSRS++ gear model.  I
may try to look him up today.

Jon


_______________________________________________
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel

Reply via email to