Hello,
We have a tiltrotor VTOL with 8 motors. The aircraft is of a wing/canard design - there are 4 motors/props along the rear wing (10" diameter), and 4 motors/props along the front canard (7" diameter). All motors are tilting for forward flight. It’s about 30lbs.
This aircraft has been flying as an “Octa H” frame (Q_FRAME_CLASS=3 and Q_FRAME_TYPE=3) with moderate success. We have been able to run a quicktune and it flies reasonably well in low wind in Q_HOVER and Q_LOITER.
I want to be able to refine the tune further, and I believe it will require a custom frame class. The “H” frame is not representative of each motors roll/pitch contribution in this configuration. Being a wing/canard aircraft, the CG is aft, closer to the rear wing. The front motors/propellers along the canard are smaller, with less thrust contribution (about 80% of the rear).
In developing the normalized roll/pitch contributions for the custom frame class, should I account for the reduced thrust contribution of the front motors (i.e scaling them down by 80%)? Or, in this asymmetrical configuration, should the pitch/roll contributions be purely based on geometry (scaled lever arms from the CG)?
We have modeled the aircraft in RealFlight and are able to leverage SITL. With our first hack at a custom frame class, a quicktune produced much higher gains, which is promising, however there seems to be some roll/pitch coupling and loss of altitude when maneuvering, so we suspect something is wrong.
Thank you!