How do I cancel axial spinning on a helicopter? [Breaking ground DLC[

[xendelaar]

I've been trying to make a small helicopter using several very cool tutorials I found online. I'm currently experimenting with dual rotor helicopters. I use a second identical yet mirrored rotor, which is attached to a motorless servo, to negate axial torque. The thing is.. the craft will eventually start spinning if I don't use any SAS.

If I do use reaction wheels, the rotation almost halts to a NEAR minimum.. but still..the vessel rotates. Like one rotation per minute.. Adding more reaction wheels makes it turn slower but it never stops.

this makes flying long distances a pain in the poopenshaft.

Is this minimum rotation velocity (with sas ) a common thing? Like, is it a bug? Or is my helicopter design just flawed.

[Guest]

I think it's a thing: the torque transfer between the rotors isn't 100% perfect because there is also some friction between the servo and the helicopter body, so some of the torque will bleed into it. IME using two identical rotors stacked on each other with the lower one unmotorised works better.

If you don't want to use reaction wheels to paper that over, you have to do it aerodynamically. For forward flight at a reasonable speed a tail with a stabiliser ought to do it; if you want it also to work for hover, you'll have to add a tail rotor and make the blades on that control yaw.

If it makes you feel any better, helicopter pilots tell me that's how it works with real helicopters too, they require continuous control input to stay going where you want them to go.

(As a consequence I only build and fly helicopters in KSP for fun. If I want to make a VTOL craft for career use that I intend to fly long distances, I make it a tilt-rotor, those are nice and stable in level flight, and also a lot faster.)

 

[AHHans]

3 hours ago, xendelaar said:

Is this minimum rotation velocity (with sas ) a common thing?

Well, it is not unheard of. The simplest case is if you don't have enough control authority (reaction wheels or control surfaces) to keep the craft stable.

On planes and rockets SAS is fairly good at keeping your craft pointed in one direction, if you have constant torque so that once it finds a "trim setting" it doesn't need to change that setting much. It will keep pointed somewhat offset to its "reference position" though. And any time you give a control input it resets its "reference position" and it will then settle on a new offset to that position. Having more reaction wheels means that it a) doesn't need as much offset to find the correct "trim setting" and b) finds the "stable" setting sooner.

But my version of a helicopter also keeps drifting around. My guess is that this is caused by two issues: one is that all control axes are coupled. Control input in pitch will also affect yaw and roll, etc. The other one is that the torque and thus the needed control input keeps changing as your airspeed changes in direction or speed.

[xendelaar]

2 hours ago, Brikoleur said:

I think it's a thing: the torque transfer between the rotors isn't 100% perfect because there is also some friction between the servo and the helicopter body, so some of the torque will bleed into it. IME using two identical rotors stacked on each other with the lower one unmotorised works better.

If you don't want to use reaction wheels to paper that over, you have to do it aerodynamically. For forward flight at a reasonable speed a tail with a stabiliser ought to do it; if you want it also to work for hover, you'll have to add a tail rotor and make the blades on that control yaw.

If it makes you feel any better, helicopter pilots tell me that's how it works with real helicopters too, they require continuous control input to stay going where you want them to go.

(As a consequence I only build and fly helicopters in KSP for fun. If I want to make a VTOL craft for career use that I intend to fly long distances, I make it a tilt-rotor, those are nice and stable in level flight, and also a lot faster.)

 

Thanks for the tip! it does make sense that there are some friction losses which I have to take into account.

I don't mind using reaction wheels btw. in fact, the originol craft has 2 small (not tiny) reaction wheels.  it's just that.. the whole craft weighs approximately 1,6 tonnes... when I add 5 large 0,2 tonne reaction wheels, the craft still rotates a little!
if i just tap "Q" or "E" the craft spins directly and accordingly, so there is plenty of torque available to stabilize the craft, but apparently SAS doesn't work properly. I tried mechjeb2 autopilot to stabilize the craft but that made things even worse haha

I'm planning to circumnavigate Duna and without an extreme stable craft, it will be impossible to complete this mission. My vessels' top speed is currently 60 m/s. it would take me almost 10 real-life hours to fly across the planet.  I need a really stable craft so I can go AFK when I feel like it. Else I'll go crazy!    

45 minutes ago, AHHans said:

Well, it is not unheard of. The simplest case is if you don't have enough control authority (reaction wheels or control surfaces) to keep the craft stable.

But my version of a helicopter also keeps drifting around. My guess is that this is caused by two issues: one is that all control axes are coupled. Control input in pitch will also affect yaw and roll, etc. The other one is that the torque and thus the needed control input keeps changing as your airspeed changes in direction or speed.

Thanks for the reply.
the thing is, I added 60% of weight (5x large (0.2 t) reaction wheels  on a 1,6 t craft) in the form of reaction wheels to the craft and it still spun around with SAS on. I assumed that this would be enough to stop any craft from spinning.. (the craft is completely (Radial) symmetrical btw.) I could add even more reaction wheels I guess, but i doubt that will solve the problem. 

the tail rotor idea is brilliant btw! If all other thing fail, I could alway try and look into that!  I sounds like a neat puzzle/experiment! 

1 hour ago, AHHans said:

On planes and rockets SAS is fairly good at keeping your craft pointed in one direction, if you have constant torque so that once it finds a "trim setting" it doesn't need to change that setting much. It will keep pointed somewhat offset to its "reference position" though. And any time you give a control input it resets its "reference position" and it will then settle on a new offset to that position. Having more reaction wheels means that it a) doesn't need as much offset to find the correct "trim setting" and b) finds the "stable" setting sooner.

Does trimming actually work when you have SAS on? It thought SAS removed the trim when you activate it. did the developers change that? 

[AHHans]

1 hour ago, xendelaar said:

I'm planning to circumnavigate Duna and without an extreme stable craft, it will be impossible to complete this mission.

Start on Eve. The atmosphere there is so thick that solar powered propeller planes work really well there.
(Yes, I could stay AFK for quite a while while my Eve plane was in transit.)

1 hour ago, xendelaar said:

I could add even more reaction wheels I guess, but i doubt that will solve the problem. 

Yes, I don't think that will help.

I think SAS is also more effective in keeping pitch and yaw constant, and it doesn't care that much about roll. And for a rocket that's usually fine, there you don't care much about roll...

1 hour ago, xendelaar said:

the tail rotor idea is brilliant btw!

Ehhh... It's not really my idea. But thanks anyway.

That craft only works due to the features of cyclic and collective control that SQUAD introduced with 1.9.

1 hour ago, xendelaar said:

Does trimming actually work when you have SAS on?

What I mean with "trim setting" is the amount of control input that SAS generates to keep the craft steady. Not the manual trim that you can set with <R-Shift>-WASDQE (or <ALT>-WASDQE on Windows).

[XLjedi]

I don't see the rotation problem in coaxial designs...  So those are all I tend to focus on.  And unlike many, I seem to have no problem with reaction wheels for yaw control if the flight control is more stable.  I tend to be more interested in viable mission craft that do what I want, rather than worrying over RL concepts that can't really be executed with what we have (due to both mechanical and atmospheric inconsistencies with the real world.)

Spoiler

https://kerbalx.com/XLjedi/AH-11-Dragonfly

https://kerbalx.com/XLjedi/MH-45-Cyclone-III

 

Edited March 4, 2020 by XLjedi

[xendelaar]

1 hour ago, AHHans said:

Start on Eve. The atmosphere there is so thick that solar powered propeller planes work really well there.
(Yes, I could stay AFK for quite a while while my Eve plane was in transit.)

funny that you mention it, but I already did this on Eve.. it also took me 10 hours (i think). If you're interested: check this link to see the video. (please neglect the crappy and choppy quality) 

 

1 hour ago, AHHans said:

 

What I mean with "trim setting" is the amount of control input that SAS generates to keep the craft steady.

okay.. so I don't know what you mean by trim settings. I will look it up! Thanks  

58 minutes ago, XLjedi said:

I don't see the rotation problem in coaxial designs...  So those are all I tend to focus on.  And unlike many, I seem to have no problem with reaction wheels for yaw control if the flight control is more stable.  I tend to be more interested in viable mission craft that do what I want, rather than worrying over RL concepts that can't really be executed with what we have (due to both mechanical and atmospheric inconsistencies with the real world.)

  Reveal hidden contents

https://kerbalx.com/XLjedi/AH-11-Dragonfly

https://kerbalx.com/XLjedi/MH-45-Cyclone-III

 

thanks! so i'm guessing i'm  the one to blame. thanks for the links to your craft! i will check them out later if the "ol' ball and chain" let me

[XLjedi]

1 hour ago, xendelaar said:

thanks! so i'm guessing i'm  the one to blame. thanks for the links to your craft! i will check them out later if the "ol' ball and chain" let me

It didn't sound like you were working on a coaxial design.  I was just pointing out the design option for you to try.

[AHHans]

4 hours ago, xendelaar said:

funny that you mention it, but I already did this on Eve.. it also took me 10 hours (i think). If you're interested: check this link to see the video. (please neglect the crappy and choppy quality) 

Quite a nice video. I also don't mind the hard cuts. But - in particular at the beginning, when you were going to Eve - the fast cuts meant that I couldn't follow what was happening. Later, during the actual circumnavigation that was O.K. because I knew that e.g. the short cuts to the map screen weren't important.

4 hours ago, xendelaar said:

okay.. so I don't know what you mean by trim settings.

Maybe I shouldn't have called them "trim settings". What I meant was just how SAS steers the craft. E.g. on a typical - slightly nose heavy - plane you need to apply some pitch up to keep the plane in level flight. If you use SAS to stabilize the plane, then SAS will apply some pitch-up control to keep the nose from pitching down. That's what I meant, essentially you use SAS to "trim" the plane to fly level.

[xendelaar]

12 minutes ago, AHHans said:

Quite a nice video. I also don't mind the hard cuts. But - in particular at the beginning, when you were going to Eve - the fast cuts meant that I couldn't follow what was happening. Later, during the actual circumnavigation that was O.K. because I knew that e.g. the short cuts to the map screen weren't important.

Maybe I shouldn't have called them "trim settings". What I meant was just how SAS steers the craft. E.g. on a typical - slightly nose heavy - plane you need to apply some pitch up to keep the plane in level flight. If you use SAS to stabilize the plane, then SAS will apply some pitch-up control to keep the nose from pitching down. That's what I meant, essentially you use SAS to "trim" the plane to fly level.

i appreciate the feedback on the video. it was actually just a Hotkey-script that took a snapshot every 30 seconds. I didn't want to record 11 hours of data to my pc for just a single mission.

 

Anywho! I found the solution, thanks to you, @Brikoleur and @AHHans! stacking two motors on top does give some friction losses, which in turn result in a spinning vessel! I countered it by making a Chinook-like helicopter!

[Rpatto92]

@xendelaar Just recently posted a guide. We'll probably never have a helicopters that stays stable over a long flight without active control but it is possible to compensate for a lot of the asymmetric lift you encounter when building a helicopter, with an articulated system.

 

here's a link: