Help with overcomensating SAS

[Cranium]

I cannot for the life of me figure out why this thing doesn't want to stay in control. I have a suspicion it's something with my actual game, and that I'll likely have to reinstall the game, but meh...

Upon liftoff, it tends to start overcorrecting on everything, and it won't center itself. The body is solid, so it can't be floppiness, and the CoL and CoM is well positioned. I'm told on IRC that I need to move around the flaps, but honestly, it hasn't done anything to mitigate the problem. If anyone could help me make this thing fly straight, I'd appreciate it.

http://puu.sh/ceLfO/6bdf3db5b8.craft

As a sidenote, I've tried this with and without FAR, to no avail. I've also tried throttling down, and of course disabling gimbal on the turbojets. Throttling down only exasperates the issue. Since the gimbals are so close to the CoM/CoL, it doesn't really affect the craft at all to disable/enable them.

Another sidenote: It's almost acting as crafts used to before SAS was "fixed"

[Geschosskopf]

I don't think this is totally an SAS issue, I think it's more of a control issue.

First off, how are your control surfaces set up? Have you limited them to specific axes or all they all still on default for all axes? It usually helps a lot to have each surface only act in 1 direction. If you leave them on all axes, most of them are poorly placed to act on at least 2 axes but by trying to anyway, they screw up the plane's stability.

Second, on a plane that size, it's usually not a good idea to have the rudder up above the fuselage like that. Better to put rudders on the trailing edge of the wing, so they're even with the centerline. This prevents the rudders from causing roll even if the rudders themselves are set only to work in yaw.

And finally, there's torque. Torque is wonderful for helping SAS keep planes stable but the SP+ cockpits don't have very much of it and neither does the Mk2 probe core except in pitch. I usually find it necessary to add a small cargo bay full of inline reaction wheel units and the RTGs to power them (IOW, it's a service module, not a working cargo bay). Looks like here you've only got the cockpit's torque.

[Taki117]

Just loaded up your craft and found your CoL and CoM are not "Well positioned" which might be the problem.

Also, you do have all your control surfaces set to their defaults. (With the wings Yaw is active) I will continue to update with my findings as I fly your aircraft.

After Adjusting the control surfaces I cannot reproduce the problem, however, due to the high takeoff speed excessive pitch up on takeoff (Which is normal for spaceplanes) results in catastrophic failure of the forwards wings.

Aerodynamic failure occurs approximately 1 minute into the flight, suspect it is caused by CoM shifting Backwards.

Edited October 16, 2014 by Taki117

[Cranium]

Well I do have the rudder set to do only yaw, elevators set to do only pitch, and ailerons to only control roll(only the outer ones, mind you. That's a lot of roll authority if I set both of them to roll). For the most part, I don't have any issues with over rolling, since it has a slight dihedral design allowing it to stay fairly level in flight. The problem is more to where if I do exert some control over it, say pitching up or rolling one way or the other, it won't center itself anymore, and will constantly bounce back and forth along that control axis. If I pitch, it does a dolphin act and starts going up and down constantly, without any more input. If I roll, it does the wing-waggle dance for a while.

The slope of the place where the rudder is helps it stay straight. Since there's very little yaw control, I don't have much of an issue with the tail causing too much roll. Even if it does, there's more than enough control on the ailerons to conpensate.

I don't really know if just adding tons of almost completely dead weight to it just to make it more controllable would help in this instance. I don't know of any planes that use reaction wheels in favor of changing aerodynamics to stabilize.

- - - Updated - - -

Just loaded up your craft and found your CoL and CoM are not "Well positioned" which might be the problem.

http://i.imgur.com/61II8cF.png

Also, you do have all your control surfaces set to their defaults. (With the wings Yaw is active) I will continue to update with my findings as I fly your aircraft.

After Adjusting the control surfaces I cannot reproduce the problem, however, due to the high takeoff speed excessive pitch up on takeoff (Which is normal for spaceplanes) results in catastrophic failure of the forwards wings.

That's certainly odd... It's definitely not like that when I load it in the SPH...

[Taki117]

EUREKA! I have located the issue!

It's the aerospike engine.

I decided to rebuild your aircraft from the ground up. The issue arises when The aerospike engine is used as demonstrated in these pictures.

Without the Aerospike:

Notice how the CoL is in front of the CoM? That's how it should be.

Add the Aerospike and the CoL moves well behind the CoM.

Replacing the Aerospike with the LV-909 has fixed the issue.

Edited October 16, 2014 by Taki117

[Cranium]

What the flip???

That makes absolutely no sense whatsoever!

I'll try with your suggestion, and see if I can get this girl flying.

[Cranium]

Ah wait, I noticed the back part was made into two pieces. I used the longer adapter piece for the entire tail.

[Alshain]

It's the radial intakes, your creating a lot of drag ahead of the center of mass, that isn't a good idea, it wants to pull the nose around the wrong way. Also your CoL is a little closer to the CoM than you probably want.

When using FAR, 1 shock cone per engine is all you need, your going to be able to get up to the maximum speed those engines can do on just that and that will significantly reduce the drag if you remove the others.

Remember, FAR puts further limits on the engine for realism. When you get up to around 1500m/s your going to lose the engines no matter how much air you have left. With 1 shock cone per engine you can get to around 24km and get to that speed there, pull up before it cuts out and probably make 35 - 40km apoptosis.

tl/dr: remove the intakes except the cones, pull all of the wings back engines and all so the CoL has some space between it and the CoM, similar to Taki's first image would be good (maybe a little closer). Consider moving the rear wheels to the engine fuselage for easier landing.

Edited October 16, 2014 by Alshain

[Taki117]

It's the radial intakes, your creating a lot of drag ahead of the center of mass, that isn't a good idea, it wants to pull the nose around the wrong way. Also your CoL is a little closer to the CoM than you probably want.

When using FAR, 1 shock cone per engine is all you need, your going to be able to get up to the maximum speed those engines can do on just that and that will significantly reduce the drag if you remove the others.

Remember, FAR puts further limits on the engine for realism. When you get up to around 1500m/s your going to lose the engines no matter how much air you have left. With 1 shock cone per engine you can get to around 24km and get to that speed there, pull up before it cuts out and probably make 35 - 40km apoptosis.

I thought that too so I removed the intakes for one flight and the same flip happened Hence why I started to rebuild.

@Cranium The point remains the same. The Lifting Body CoL moves well outside the actual body, This is hidden by the wings because they are actual lifting surfaces.

[Alshain]

I thought that too so I removed the intakes for one flight and the same flip happened Hence why I started to rebuild.

@Cranium The point remains the same. The Lifting Body CoL moves well outside the actual body, This is hidden by the wings because they are actual lifting surfaces.

You are also using an old version of FAR though. You can tell by the little arrow that extends out of the CoL, ferram4 removed that because it didn't make sense. You loaded up his plane and yet your first screenshot is different than his. That concerns me that something changed. EDIT: Except you don't have that in the later screenshots so I'm a bit confused.

OP: I also just noticed your CoL is high. It's because your mounting the tail plane high. The only reason to have a high tail plane is if you have a low mounted wing... like this one. In the end it cancels out and the CoL is directly in line with the CoT and CoM. If your wings are already in line with the CoM, the tail plane should be as well.

Edited October 16, 2014 by Alshain

[Taki117]

You are also using an old version of FAR though. You can tell by the little arrow that extends out of the CoL, ferram4 removed that because it didn't make sense. You loaded up his plane and yet your first screenshot is different than his. That concerns me that something changed. EDIT: Except you don't have that in the later screenshots so I'm a bit confused.

I also just noticed your CoL is high. It's because your mounting the tail plane high. The only reason to have a high tail plane is if you have a low mounted wing... like this one. In the end it cancels out and the CoL is directly in line with the CoT and CoM. If your wings are already in line with the CoM, the tail plane should be as well.

The later screenshots don't have lifting surfaces, however I did just update FAR so I will run my battery of tests again.

Alshain is right on both counts, it's the radial air intakes, which I missed to to a recently outdated version of FAR. (Apparently he updated yesterday and I missed it) Also though, I would take a look at your Control surface setup (The outermost ones on your main wing seem to be disabled)

Edited October 16, 2014 by Taki117

[Cranium]

Yeah, I disabled those outermost ones on purpose, due to the large amount of control authority I would have for rolls if I had them both enabled. They're only there at this point for aesthetics. I'm not planning on removing them though.

[Taki117]

Yeah, I disabled those outermost ones on purpose, due to the large amount of control authority I would have for rolls if I had them both enabled. They're only there at this point for aesthetics. I'm not planning on removing them though.

They work wonders for pitch.

[Alshain]

Ideally the roll should be on the wing tips the inner ones would be best served as spoilers which will help you descend without picking up too much speed. You can also try them as flaps which in theory are supposed to reduce your minimum safe speed for landing but I've never been able to get that to quite work for me in FAR.

[Cranium]

So I updated FAR, and tried moving around the control surfaces. Every craft I create has this problem. It's KSP. I'll reinstall now, and see if this fixes the problem.