IR coaxial-rotor helicopter: pitching and resisting rotation

[ThinkInvisible]

So I made a coaxial-rotor helicopter with Infernal Robotics parts and TweakScale (as shown above), and it's being very contrary. Both rotors are directly above the center of mass; both rotate with constant and opposite speed, and lift is controlled by tilting the blades. As it stands, I'm just using reaction wheels to control rotation (and thus movement) of this vehicle, since I don't know of any other way to do this.

Something on this vehicle seems to be resisting any movement that isn't straight up or down. If I pitch the vehicle down in an attempt to move forwards, it works for a few seconds before the nose starts being forced back up (and slightly to the right, roll only). The wings and tail fin were added as a response to this problem, but they don't seem to help too much. With SAS disengaged, the vehicle also seems to be very stable with respect to yaw and roll, but it begins pitching uncontrollably and wildly (going from pointing straight down to straight up and back, repeatedly) if left alone. It seems like a similar oscillation can be induced with roll by moving left or right for more than a few seconds before trying to come to a stop.

What could cause a helicopter to act this way? Am I missing some kind of critical control surface? Also, is there any way to control a helicopter like this without rotating the entire vehicle?

[Snark]

1 hour ago, ThinkInvisible said:

So I made a coaxial-rotor helicopter with Infernal Robotics parts and TweakScale (as shown above), and it's being very contrary. Both rotors are directly above the center of mass; both rotate with constant and opposite speed, and lift is controlled by tilting the blades. As it stands, I'm just using reaction wheels to control rotation (and thus movement) of this vehicle, since I don't know of any other way to do this.

Okay, that's seriously cool!

1 hour ago, ThinkInvisible said:

What could cause a helicopter to act this way? Am I missing some kind of critical control surface? Also, is there any way to control a helicopter like this without rotating the entire vehicle?

Not sure to what degree KSP has the full aerodynamic complexity and/or physics modeling to accurately capture how a helicopter behaves.

Real helicopters are controlled by altering the pitch of the blades.

https://en.wikipedia.org/wiki/Helicopter_flight_controls

The pilot's controls are a cyclic (big joystick, this is the primary control) and a collective (a lever like the handbrake of a car).

The collective controls the net pitch of all the blades; "give it more collective" and all the blades are pitched steeper, = more lift.  This is used to control the up-down motion of the chopper (i.e. direction parallel to the main rotor axis).

The cyclic is more interesting.  It uses a clever mechanical arrangement to alter the pitch of the blades by different amounts based on what angle they're at, as they spin around.  So if you want the chopper to move forward, you push forward on the cyclic.  This causes the blades to have steeper pitch when they're at the back of the chopper and shallower when they're at the front, as they spin around, so that the rear part of the "rotor disk" is generating more lift than the front, the chopper noses down, and it scoots forward.

Helicopters also need a way to control yaw (i.e. roll around the main rotor axis).  This is typically a pair of pedals that you can push left-or-right to yaw left-or-right.  For conventional choppers, this controls the tail rotor (I assume blade pitch).  I dunno about coaxial choppers, but I would assume it would apply differential pitch to the clockwise-versus-counterclockwise rotors to achieve the same effect.

As to why your chopper isn't behaving as expected when you nose it forward:  no clue.  I would guess that it has something to do with the disparity between KSP's fairly simplistic aero modeling and real world aerodynamics.  Just out of curiosity, have you tried flying it with FAR?  Any difference?

Edited February 23, 2016 by Snark

[ThinkInvisible]

37 minutes ago, Snark said:

 As to why your chopper isn't behaving as expected when you nose it forward:  no clue.  I would guess that it has something to do with the disparity between KSP's fairly simplistic aero modeling and real world aerodynamics.  Just out of curiosity, have you tried flying it with FAR?  Any difference?

FAR seems to have solved the problem; there's still a little instability, but it's very manageable. FAR did steal most of my lift, but a redesign fixed that, including less of a massive nuclear reactor, comically oversized blades, and removal of the wings and tail fin. The thing isn't unbearable to fly anymore, and it goes surprisingly fast for something that looks so awkward - I managed ~30m/s before I found the ocean and decided not to fly over it.

It seems to be impossible to control individual blade pitch like that with just Infernal Robotics's default controls (each blade would need its own keys!); I'd need to use something like kOS and get a computer to do it for me, which I've been meaning to figure out someday (but not today). This particular helicopter just uses gyroscopes to push the nose around directly, and the addition of FAR seems to be enough to make that work.

[Snark]

22 minutes ago, ThinkInvisible said:

FAR seems to have solved the problem; there's still a little instability, but it's very manageable. FAR did steal most of my lift, but a redesign fixed that, including less of a massive nuclear reactor, comically oversized blades, and removal of the wings and tail fin. The thing isn't unbearable to fly anymore, and it goes surprisingly fast for something that looks so awkward - I managed ~30m/s before I found the ocean and decided not to fly over it.

It seems to be impossible to control individual blade pitch like that with just Infernal Robotics's default controls (each blade would need its own keys!); I'd need to use something like kOS and get a computer to do it for me, which I've been meaning to figure out someday (but not today). This particular helicopter just uses gyroscopes to push the nose around directly, and the addition of FAR seems to be enough to make that work.

Glad to hear it worked out!

And (belatedly, how rude am I!) welcome to the forums!

[Azimech]

Hi!

I recognize this behaviour with stock aero, yet the coaxials are a dream to fly compared with the singles. Thinner, longer blades which spin faster are better than fat, slow ones. Also, too much pitch might give better forward flight speed but it will also increase the risk of lift imbalance which is almost impossible for stock aero to correct ... it has a strange bug.

If you would like to get better performance in stock aero ... I'd like to take a look at it with you. I do believe a mechanical solution is possible for cyclic/collective, I did some experimenting with stock parts a year ago, with IR or some modified IR parts it would be even easier.

By the way, I'm the inventor of the stock turboshaft/turboprop engine, the first to build a turboshaft helicopter and the inventor of the stock coaxial turboshaft helicopter. I had coaxials flying in 1.04 but Squad needed to adjust jet engine parameters so they may not fly again without mods.

You might like the video.