How well does the ardupilot handle off CG on trad heli?
For instance, if I wanted to use a larger than normal battery or a cam gimbal or both and it moves the CG forward some of the main rotor shaft how will the auto pilot be impacted? Will stability modes drift more or require more cyclic trim or just holding cyclic back? Will the autopilot just compensate on its own and the heli will more or less fly the same with maybe slightly different handling?
I have some larger batteries that I got for longer flight times when my heli used a beastx controller. With that it was not a very noticeable difference other maybe having to hold a little more back pressure on the pitch stick some times. I was pretty new to flying with a probably sorta poorly tuned heli when I tried it though. Just looking for any gotchas or good info on how this impacts the autopilot since I am converting this heli to arudpilot. I say this just to convey the heli was perfectly flyable with the amount of off CG that it has. Its just not balanced at the rotor shaft with this configuration but would be in the envelope if these RC heli had CG charts.
The flight controller will for the most part handle it well, until it can’t. However that doesn’t mean you should do it.
Every bit that the C of G is off you are compromising control authority. Even if you’re just flying around calm and level, at some point you need to stop or there will be wind gusts or something. If the C of G is off and the controller is compensating for it then you may be lacking the authority to make the required movements. The controller can’t override physical limitations of the servos, blades, swashplate travel, or physics.
Regarding tuning, there is also a potential for high levels of I-term that could result in poor performance.
Thanks! Sound like an approach with caution and do some testing type thing? Just gota remember if I load the other battery it wont fly the same or might even need to be returned?
I do get the lack of cyclic authority as its being used up to compensate.
Its not that far out. Less than 2cm forward of the main shaft closer to 1cm. Or hanging it from the blade grips it settles around 12-13 degrees nose down. I expect it will still have plenty of authority for non 3d flying.
I forgot to say this is on a Trex470L so MedSmall 3d heli.
It depends a little on your setup and what integrator scheme you are using. I will mainly affect the takeoff and landing. Mainly because you have to purposely keep the integrator from building too much. Generally, I don’t like to have any more than 0.2 on the integrator when you look at the PIDR.I or the PIDP.I terms in the log. If you use the leaky integrator, that means you will have to set ILMI to 0.2 so it limits the build up. But with leaky I, once you are off of the ground and in forward flight, the integrator is then allowed to use the IMAX limit.
So this means to you as the pilot that, as you pointed out with beastx, you will have to hold some stick into the wind or if the CG offset is too big, you will have to hold stick after liftoff to keep the aircraft from drifting. I think for manual mode flight like stabilize, althold or acro, this would not be an issue. It could become more of an issue with Loiter or auto modes because the position controller relies on the attitude controller to set the attitude accurately otherwise you could get poor performance. If the CG offset gets really big then you could run into controllability issues like @Allister points out.
It is possible that the tune could change depending on the weight that is added. However if you plan to use both larger and smaller batteries, it is best to tune it for the light battery and accept the slower response of the aircraft at the heavier weight.
I don’t think the pitch attitude hanging from the blade grips means anything. if it was a teetering rotor system then it may be a problem.
Most full scale heli’s hover nose up because of the shaft angle. It is affected by CG as well but the reason why all hover nose up is mainly the shaft angle. They are compensating for the thrust vector needing to be pointed forward for cruise flight while trying to keep the fuselage level (low drag).
@spova do short hover flight where you just want to see where the integrator settles out with the new CG. I think any more the 0.2 to 0.3 Integrator (PIDP.I or PIDR.I) would probably be too much.