Green Vectors

[XLjedi]

This is dual engine setup with counter-rotating engines...

What are the green vectors doing?  Notice the lift vectors are all pointing in the correct direction.  The green vectors seem to cause a yaw in my craft at speed, and I'm not sure what they are telling me or why one of them seems to be pointing in the wrong direction.

 

[AHHans]

9 hours ago, XLjedi said:

What are the green vectors doing? 

They show the axis and direction of rotation.

With the new collective and in particular cyclic control that SQUAD introduced with 1.9 it is important to know in which axis and which orientation a rotor rotates so that the deflection will be in the right direction. Thus SQUAD added vectors to illustrate that. They don't show actual forces as the other vectors do.

Edited March 27, 2020 by AHHans

[mystifeid]

I'm gonna go out on a limb here and say that the green arrows show the amount of and direction of the torque of the engines. You've reversed the rotation of one of the engines so the green arrows should point in opposite directions. However you also reverse the deploy direction of the propeller blades on the engine with reversed rotation so that the thrust arrows (I'm assuming they're the pink ones) all point in the same direction. It looks like you might have some vibration judging by the splayed appearance of the green arrows - compare to :

Edited March 27, 2020 by mystifeid

[XLjedi]

The odd thing about them...  If I'm flying along in my craft and turn off SAS assist, the craft will roll/yaw in the direction of the forward facing vector.   

They green vectors should NOT have an impact on flight dynamics, but they sure seem to!

That seems like a bug to me.  As a result, you can't really fly a twin-engine without SAS on.  

@mystifeid  the "splayed" green vectors... or spread, is simply due to the fact that my propeller blades are larger than yours.  Question is, does your twin-engine craft roll/yaw to the right (based on your screenshot) when you are flying at max speed straight and level, and then turn off your SAS?  If it does...  this is an aerodynamics bug that needs to be fixed.

Also, as you're flying along straight & level with SAS turned on, you can turn on aeroforces overlay and watch your rudders compensating to keep the nose pointing straight.

 

5 hours ago, AHHans said:

They don't show actual forces as the other vectors do.

I'm suggesting they DO in fact, have an impact on roll/yaw... but they shouldn't!

Edited March 27, 2020 by XLjedi

[AHHans]

1 hour ago, XLjedi said:

Question is, does your twin-engine craft roll/yaw to the right (based on your screenshot) when you are flying at max speed straight and level, and then turn off your SAS?

Well, this plane (O.K. after the changes needed between 1.7.3 and 1.9.1) flies perfectly fine without SAS after setting the pitch trim. It does roll a tiny bit to the right, but I need to have it fly without any control input for a minute or more before I notice anything.

1 hour ago, XLjedi said:

I'm suggesting they DO in fact, have an impact on roll/yaw... but they shouldn't!

Well, if they would show actual forces, then the craft would yaw violently already on the runway. You can try this by setting the blade pitch to zero (= zero AoA on the blades at reset) and spining up the rotors: long green vectors but the plane doesn't move (well, except for the usual rolling down the runway).

[XLjedi]

1 minute ago, AHHans said:

Well, this plane (O.K. after the changes needed between 1.7.3 and 1.9.1) flies perfectly fine without SAS after setting the pitch trim. It does roll a tiny bit to the right, but I need to have it fly without any control input for a minute or more before I notice anything.

Well, if they would show actual forces, then the craft would yaw violently already on the runway. You can try this by setting the blade pitch to zero (= zero AoA on the blades at reset) and spining up the rotors: long green vectors but the plane doesn't move (well, except for the usual rolling down the runway).

The yaw/roll is a function of speed.  So you don't see it unless you are moving pretty good.

If you have the updated version of your sample craft, let me know and I'll check it out.

[XLjedi]

Hmmm... I just reversed the rotation on the engines and blade pitch, and it still rolls to the left.  So, I have to concede that it's probably not a function of those green vectors.  Which makes me happy.  

I just can't figure out why this craft gets squirrely and wants to roll left (if you turn off SAS) above 250m/s ?

https://kerbalx.com/XLjedi/P-39-Scorpion

I try to trim my planes out so they fly well regardless of SAS, so this one has been a bit of a puzzle.  It's like it hits some instability wall around 250-ish.  Does fine with SAS on of course.

Any ideas?

 

Edited March 27, 2020 by XLjedi

[mystifeid]

2 hours ago, XLjedi said:

Question is, does your twin-engine craft roll/yaw to the right

Nope, turns left but I can do things like move the CoM around a bit and then it has a right hand turn.

3 hours ago, XLjedi said:

the "splayed" green vectors... or spread, is simply due to the fact that my propeller blades are larger than yours

It might be that it has more to do with the size of the engines and not the size of the blades.

[AHHans]

1 hour ago, XLjedi said:

I just can't figure out why this craft gets squirrely and wants to roll left (if you turn off SAS) above 250m/s ?

Well, at those speeds drag becomes an important issue. And if one side of the plane has an additional ladder, and the other doesn't...

[XLjedi]

7 minutes ago, AHHans said:

Well, at those speeds drag becomes an important issue. And if one side of the plane has an additional ladder, and the other doesn't...

...then it makes no difference.  Ladders don't generate drag in the stock game.

I did of course test it without the ladder, same result.  I was thinkin maybe the slight weight differential had an impact though.  It didn't, so I left the single ladder on the design.  If I saw it behave differently without the ladder I would've added one to each side.

8 minutes ago, mystifeid said:

Nope, turns left but I can do things like move the CoM around a bit and then it has a right hand turn.

How fast does it go?  I don't run into issues until I get up around 250 m/s.

I think what might actually be influencing my design becoming a bit unstable at 250 is the fact that it is wider than it is long.  ...which is a bit out of the ordinary.  Meh, I think I'm fine with it.  I typically fly with SAS on anyway so no harm in that for me.

Edited March 27, 2020 by XLjedi

[AHHans]

14 minutes ago, XLjedi said:

I took the ladder off... it generates no drag anyway.  Same result.

Why do you think / say that a ladder doesn't generate drag? Just because it's "physicsless" doesn't mean that it doesn't generate drag. Its mass and drag is added to its parent part (which then will have the drag that the ladder generates added to it).

Anyhow, with the ladder added your plane veers off to the left already at 210 m/s or so. Without the ladder I can go to 240 m/s or so without the roll. Above 250 m/s it indeed rolls and yaws, usually to the left. But what I think is really the case is that it becomes aerodynamically unstable in yaw. If you get it nice and level with SAS on at high speeds (> 250 m/s) and then switch SAS off, then it will keep straight for some time (it may wobble in pitch). But as soon as something disturbs it (which may be the KSP "turbulence simulation") then it will settle at a few degrees of yaw. If you yaw to the other side, then you can get it to settle on those few degrees of yaw to that other side.

[XLjedi]

1 hour ago, AHHans said:

Why do you think / say that a ladder doesn't generate drag? Just because it's "physicsless" doesn't mean that it doesn't generate drag. Its mass and drag is added to its parent part (which then will have the drag that the ladder generates added to it).

Anyhow, with the ladder added your plane veers off to the left already at 210 m/s or so. Without the ladder I can go to 240 m/s or so without the roll. Above 250 m/s it indeed rolls and yaws, usually to the left. But what I think is really the case is that it becomes aerodynamically unstable in yaw. If you get it nice and level with SAS on at high speeds (> 250 m/s) and then switch SAS off, then it will keep straight for some time (it may wobble in pitch). But as soon as something disturbs it (which may be the KSP "turbulence simulation") then it will settle at a few degrees of yaw. If you yaw to the other side, then you can get it to settle on those few degrees of yaw to that other side.

Just going by the lack of drag vectors... but I s'pose lots of parts don't look like they generate drag and do.  Anyway, seems to be a moot point... doesn't solve the problem to add or takeaway a ladder.  It's a bit of a "toy" anyway, was just curious to me why this one was behaving the way it was.  I might check another plane that runs a bit faster to see if it behaves similarly.

That yaw instability might be due to the abnormal length/width ratio.  I'm not changing it, I rather like its kerbalness in that regard, but would be an interesting tidbit to stow away for future reference.

Edit:  Yeah, I got another one that will do 270 and it doesn't have the instability issues at all.  But it's also a more traditional build in the length/width sense and it's not fighting with a big pair of spinning airbrakes mounted to the engine for looks either.  ...even tried yawing it back and forth and it doesn't lose stability.  

Edited March 27, 2020 by XLjedi

[Redacted]

I made a video recently highlighting this problem as well.

Apparently with all of my Rotor Craft they yaw continuously despite having equal thrust via counter rotating propellers. Now as things stand, I believe there is something wrong with how rotors / propellers are scripted that's causing the rotor torque to induce Yaw. It doesn't matter which rotor / propeller direction is used, Yaw is induced.

[AHHans]

25 minutes ago, Redacted said:

I made a video recently highlighting this problem as well.

Please have a look at my first reply in this thread. The green vectors do not show "rotor thrust" or any other actual force. They just show the direction of rotation. (And that rotation is detected in the first place.)

[Redacted]

I get that part but the Vectors are creating a torque induced Yaw. Any craft I've created flown does this no matter the settings (Props / Rotor). Only way I've found to prevent this is to have all motors in the center-line of the craft and as close to each other as possible. (coaxial) Here is an example craft that I tried earlier and you can easily see that it is wanting to Yaw (right).

 

Edited April 2, 2020 by Redacted

[AHHans]

36 minutes ago, Redacted said:

I get that part but the Vectors are creating a torque induced Yaw.

I'm not saying that there isn't an effect that introduces yaw. I'm saying that it isn't caused by the green vectors.

[Redacted]

It is.... this it motor torque. I've seen the same effort on Quadcopters I've build in KSP. Otherwise whats your theory?

Side Note: I design, build and fly Quads, Bicopters and Tri's so I've a bit of experience in this area.

Edited April 1, 2020 by Redacted

[AHHans]

46 minutes ago, Redacted said:

Otherwise whats your theory?

That your design isn't actually 100% symmetrical. Maybe because the hinges bend in different directions.

[Redacted]

Even without the hinges its the same effect.

Furthermore I see the same effect on smaller, simpler craft soon as i add counter-rotating props / rotors.

[XLjedi]

For rotors try a coaxial design...  and see if the issue persists.

[Redacted]

@XLjedi In just saying "Try coaxial a design" is all but admitting there is a problem and no-one wants to address it. But yeah for the sake of argument, I've already "tried" this and yes it does solve the situation. However thats only because both axis of rotational torque are on the same plane. (mounted forward and aft of each other) Realistically though, props shouldn't be limited to only coaxial designs. Would be much better to see a fix then continue to act as there isn't a problem with Rotors.

[XLjedi]

3 hours ago, Redacted said:

@XLjedi In just saying "Try coaxial a design" is all but admitting there is a problem and no-one wants to address it. But yeah for the sake of argument, I've already "tried" this and yes it does solve the situation. However thats only because both axis of rotational torque are on the same plane. (mounted forward and aft of each other) Realistically though, props shouldn't be limited to only coaxial designs. Would be much better to see a fix then continue to act as there isn't a problem with Rotors.

Oh I think there's a problem with it; I'm just not sure exactly what the problem is at this point.  So far, I haven't really been able to see a significant performance hit in my craft (except the multiengine I posted to this thread, which I thought was more related to instability due to length/width ratio).

I do know exactly when it happened.  They introduced something goofy in the rotors/blades when they added the "cyclic" ability for helos.  No question.

I notice in particular, that at high speed the rotors now have forces on them that cause them to "tilt" in opposite directions depending on rotation direction.  For example take a look at the rotational plane on this helo when hovering vs. traveling at speed. 

Spoiler

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

This just looks goofy as hell as I fly the thing now... but doesn't seem to impact the flight performance beyond the visual stupidness.  Keep in mind, on this craft, I don't even have the cyclic controls enabled for the blades and it does this.

There is definitely something going on with the lift forces that is impacting the blades (in a mirror fashion) that is dependent on their rotation direction.  I also suspect that similar forces are being applied to the props, you just don't notice the tilt because the rotational plane is smaller and maybe not as obvious.  So far, it hasn't drastically impacted the performance of this craft, but it certainly looks goofy!  The rotors did NOT behave/look this way in version 1.9.0.  So has to be something to do with the cyclic changes they made.   

The error that was introduced (whatever it was) is more pronounced when the craft are traveling at speed through the atmosphere.  If you are just hovering around at 20 mps no one is going to notice.  It may not have been so much an issue if the stupid BG parts were not so "bendy".  The tilt forces on the rotor blades seem to be causing the engine rotor parts to bend.

If we can all agree this is a bug, I'd be willing to post or upvote a report on the @SQUAD bugtracker.

 

Edited April 2, 2020 by XLjedi

[Redacted]

@XLjedi Sure I can agree its bugged.

I would mention that I was not using the Cyclic Controls though. All of the P-Y-R controls were disabled so I'm pretty convinced that this is a torque issue. Maybe the torque is causing the Cyclic Controls to misbehave as they attempt to counteract the forces. Should be easy to see as you throttle up / down the Rotor. Behavior should get worse the higher the RPM's and Prop load are pushed.

 

Edit: Would be interesting to see exactly where the Center of Torque (CT) is for the Rotors. If the CT is not in the middle of the Rotor (too far forward or backwards) then this would explain why.

Edited April 2, 2020 by Redacted

[XLjedi]

For comparison, here's the same exact craft in v1.9.0 prior to the introduction of cyclic controls:

Spoiler

Sadly, after looking at one of my flight test videos (screenshot above) it did take a performance hit.  As seen here, I was getting the craft up to 179-180 m/s in level flight.  ...and notice the rotor planes are fine!  In v1.9.1 the same craft struggles to get above 160 m/s.  So I took a 20 m/s performance hit on this craft as a result of what they did with the cyclic.

[XLjedi]

16 minutes ago, Redacted said:

Should be easy to see as you throttle up / down the Rotor. Behavior should get worse the higher the RPM's and Prop load.

I don't see that though...  seems to get worse or more pronounced dependent only on airspeed.  I can hover at full RPM load and there's no visible impact.