Best way to make an autopilot turn?

[skogsvargen]

I am trying to figure out how to make my autopilot make my plane turn in the best way?

The plane is stabilized with IR-sensors and i have control over rudder, ailerons and elevator.

Will it be enough to roll the plane and keep the nose at the horizon or do i have to keep the nose slightly over the horizon to get more turn rate?

How advanced aileron turns are other autopilots doing?

/Magnus

[ALtitudeap]

What are you trying to turn it in to?

[skogsvargen]

Well, turning it to a new direction is enough...

[Connexxion]

If your plane is stabilised by a separate IR device (Co-Pilot?),you could use your rudder to point the plane into the new direction.

No need for roll just let it yaw.

[dmgoedde]

I am trying to figure out how to make my autopilot make my plane turn in the best way?

The plane is stabilized with IR-sensors and i have control over rudder, ailerons and elevator.

Will it be enough to roll the plane and keep the nose at the horizon or do i have to keep the nose slightly over the horizon to get more turn rate?

How advanced aileron turns are other autopilots doing?

/MagnusMagnus - yes, you are 100% correct. If you roll the plane away from level and keep the nose merely on the horizon, the plane will turn at some non-zero rate. There is absolutely no need to hold the nose high above horizon. This method is 110% effective, and how the AttoPilot works.

Forget about plane-centric yaw for now. You can add a bit of rudder to make the turn more effective, but honestly you can lock the rudder at neutral, and use only ailerons. The yawing and turn will happen just due to roll angle being skewed to the right or left while keeping the nose up to the horizon. I am not a big fan of using rudder-only to turn. The resulting flat turn (if roll were held flat) are very draggy and inneficient. BTW, I am not a stud aerodynamicist, but I am 90% correct on what I wrote, and it is backed up by 100+ MB of logged flight data on 20+ planes. Some purist might add or partially correct what I wrote.

[skogsvargen]

Yes, rudder turns are one way to do it. But this requires an airframe design that response well to rudder. A flying wing, for example has no rudder at all...

/Magnus

[airmcn_3]

Yes, rudder turns are one way to do it. But this requires an airframe design that response well to rudder. A flying wing, for example has no rudder at all...

/Magnus

What aircraft are you flying?

Chris

[skogsvargen]

Thank's Dean. This confirms what i thought. This means that with IR-stabilization i control turnrate by controlling the rollangle.

Chris,
I am mainly flying electric glider with very limited rudder response. However my question was more about turn-algorithms in general.

/Magnus

[Buzz_Lightyear]

The most "efficient" turn that I know of is a roll based turn with little to no yaw. That means that you need to know yaw, which IR does not give you. Assuming you have an IR-only system, then your best bet is just to follow Dean's advice, as this will give you your best bet for good turn.