Re: [Jsbsim-devel] Prop aircraft yaws to the right on take-off.

[Alan T. <ajt...@v-...>]

On 02/09/2019 21:11, Bertrand Coconnier wrote:
> Hi Alan,
>
> Le dim. 1 sept. 2019 à 23:30, Alan Teeder <ajt...@v-... 
> <mailto:ajt...@v-...>> a écrit :
>
>
>     I have made some progress by adding the following code to FGLGear
>     bind(). (near line 798) I can now see these properties in Flightgear.
>
>        property_name = base_property_name + "/wheel-XForce-lbs";
>        PropertyManager->Tie(property_name.c_str(), &vFn(eX));
>        property_name = base_property_name + "/wheel-YForce-lbs";
>        PropertyManager->Tie(property_name.c_str(), &vFn(eY));
>        property_name = base_property_name + "/wheel-ZForce-lbs";
>        PropertyManager->Tie(property_name.c_str(), &vFn(eZ));
>
>        property_name = base_property_name + "/strutforce-lbs";
>        PropertyManager->Tie(property_name.c_str(), &StrutForce);
>
>     The result of this is similar numbers in the property tree for
>     wheel-ZForce-lbs and strutforce-lbs. With the throttle opened the
>     numbers are larger for the left gear than the right gear, as
>     expected.
>     The value gets smaller as the aircraft accelerates on the runway.
>
>     However  the properties wheel-XForce-lbs and wheel-YForce-lbs are
>     zero.
>     They stay this way  even with the brakes applied despite the aircraft
>     coming to a halt.
>
>     Is there any reason why vFn(eX) and vFn(eY) are zero?
>
>
> Friction components are assessed at the global level i.e. when the 
> sums of all the forces and moments acting on the aircraft are available.
> Therefore the computation of friction forces is taking place in 
> FGAccelerations::CalculateFrictionForces

Bertrand

Thanks .

I will look there. I am doing my latest tests with an artificially high 
yaw inertia - thus locking the yaw axis and avoiding yaw contributions 
due to the(unstable) taildragger wheel layout.  I wait until the tail 
has lifted before freezing the sim and making my measurements, thus 
removing any tail wheel contributions.

It would have been better to work with stand-alone JSBSim, but this 
particular FDM has many hooks to Flightgear systems which would all need 
sorting out.

Alan