Spaceplane Backflipping upon reaching high speeds

[Cronos988]

Hey everyone,

As is probably expected from new KSP players, I can handily get a rocket into orbit, but fail to reach orbit with my (self designed) spaceplanes. I am not using any stock crafts because I want to get a feel for which designs work and which don't by myself. Anyways, I designed a fairly standard airplane: MK1 and Mk 2 fuselage, 3 Turbojet engines (one center, two attached directly left and right of the fuselage) with a delta wing shape and canards. In order to get into space, I have 2 aerospike engines each with it's own tank (the tank that is the same size as the MK 1 jet fuselage plus one Oscar , situated in the middle of the delta wing. Tons of intakes (about 14.0 air intake max, though I never reach that).

After many tries, I can now get the plane to about 23 km altitude with a surface velocity above 900m/s. At this point, I have to essentially point the nose downward with all my might (or rather, the SAS does) to keep the airplane from just flying straight upwards and then start flipping. My angle will usually be at about 20° at this point (I go up almost vertically after liftoff, and after reaching 10KM point the nose downward to pick up speed). At this point, I usually am below 0.20 air intake, and as the intake drops below 0.05, I cut the engines to prevent uncontrolled flameout (that is, I shutdown all 3 engines via action group, took me a while to figure out that you could do that). Then start up the rocket engines.

And then it happens: My nose pulls up and no amount of counter-steering helps, leading the plane into a backflip and eventually uncontrollable spin. Sometimes I manage to get it back under control, but at that point I have burned too much fuel to get into orbit. My center of mass upon take-off is just slightly in front of the center of lift, and the center of thrust is in the middle of the plane, as near as I can tell. Basically, all planes I designed so far have this problem: They fly very well up to altitude, but at about their max altitude, they start backflipping. I tried going slower, but while that sometimes prevented backflipping, it also made me fall slightly. Then I tried going faster, which seems to keep the plane controllable for longer, but fails as soon as switch to rocket power.

Someone on the forum noted that this might be caused by the CoG drifting backwards as the tanks empty. That sounds reasonable, as the tanks empty from front to rear. I have no idea what to do to counteract that though. Any help is appreciated

Edited July 20, 2013 by Cronos988
Question answered.

[Skorpychan]

Seems to me that the backflip is because the control surfaces are no longer effective in pushing the nose back down.

Maybe fiddle with fuel lines to make sure the tanks drain more evenly, or mount the engines further forwards on pylons?

Jettison parts at the back (jet engines, etc.) to move the CoG?

[AnalogAddict]

Also, check your intakes, especially if you have only one radial intake on the top of your craft. Closing that intake should be enough of a drag reduction to prevent your craft from back flipping in thin atmosphere. I've had the best luck placing radial intakes in pairs (above and below the craft or on either side). Generally I try and avoid the radial intakes.

Might you post a screenshot or the craft file so I may give you a more informed answer?

EDIT* I see your craft does have 7 or so radial intakes on the top of the craft, I only see a matching pair of two on the bottom; however, that may be due to the angle of the picture. If you have assigned those intakes to hotkeys, remember that if you ever move one, the hotkey for its symmetrical sister is broken. If the previous statement is true of your craft, it would explain why you had a "rightward" spiral upon re-entering the atmosphere with it definitely explaining why your craft flips out at high altitudes (>20km).

Space planes are tricky business - especially with the pseudo in-atmo physics we have. I have long thought my planes were losing attitude due to to fuel loss and CoM shift, but drag from one intake on one side of your plane is enough to start you into a flat spin. Use of the "CoL touches the rear of CoM theory" has always been sufficient to prevent loss of control due to fuel flow in my experiments.

Edited July 19, 2013 by AnalogAddict
Adding more detail

[automcdonough]

i would suspect balance shift from fuel usage

[Cashen]

Also, check your intakes, especially if you have only one radial intake on the top of your craft. Closing that intake should be enough of a drag reduction to prevent your craft from back flipping in thin atmosphere. I've had the best luck placing radial intakes in pairs (above and below the craft or on either side). Generally I try and avoid the radial intakes.

Might you post a screenshot or the craft file so I may give you a more informed answer?

A good way to do this is to bind the air intakes and the jet engines to the same action group, so they toggle together at all times. Once you're not using the jets, you don't need the intake drag.

As for the OP, I'd look at balancing. Also, if your center of mass has shifted back past the center of lift, or if the center of lift isn't along the same axis as the center of mass, those are all possible issues.

[Cal'Mihe]

Since you have to continously hold the nose down, its clear the craft is unbalanced, however with both the gimbals on the turbojets and the control surfaces you're able to componsate. As you tell, when you switch the turbojets off and ignite the rockets, the crafts nose rises beyond your ability to hold it stable.

Two main reasons possible:

Your center of thrust has moved down below the center of mass, this would cause it to backflip.

Your center of lift has moved forwards of the center of mass, this would cause the nose to rise uncontrollably.

It might also be a combintation.

As its a sudden shift when you go from jets to rockets, I would look at what happens in that precise moment, check what is the fuel left in each tank at the moment when you would have turned on the rockets, try to estimate how much the CoM has moved, and then look at what happens when the rockets thrust replace the jets.

Also try running it on rockets at low altitude, switch back and forth, see if you can get it to replicate the behaviour at low altitude under more controlled circumstances.

[Cronos988]

Thanks everyone for the helpfull replies!

Here is a screenshot of the plane. Where do I find the craft files to attach?

http://postimg.org/image/d8cfo1hal/

The center of lift is now slightly behind and above the center of mass. The center of thrust is directly behind the center of mass, no deviation sideways or up/down.

By attaching some canards to the tailfin as additional control, placing a piece of empty fuselage behind the cockpit and managing fuel tanks a bit, I have successfully gotten the plane into a 70KM Orbit. Took almost all the rocket fuel. However, these "fixes" are only addressing the symptoms, not the problem itself: Upon re-entering the atmosphere, the craft became completely uncontrolable, flipping different directions and going into a rightward spiral that I could not stop. It is obviously unstable. The question is why?

Edited July 18, 2013 by Cronos988

[zerotwo]

As a test, take away a forward fuel tank, and replace it with some structural housings (the 0.4 weight ones), watch the center of mass move, after two or three tanks have drained (the image won't load for me for some reason, so I can't actually comment exactly)your CoM will be WAY off where you thought it should be.

Try locking the fuel in forward/furthest away tanks, fuel always trains from the furthest source first.

[Cronos988]

As a test, take away a forward fuel tank, and replace it with some structural housings (the 0.4 weight ones), watch the center of mass move, after two or three tanks have drained (the image won't load for me for some reason, so I can't actually comment exactly)your CoM will be WAY off where you thought it should be.

Try locking the fuel in forward/furthest away tanks, fuel always trains from the furthest source first.

I did try and replaced a fuel tank with empty fuselage (though that was only possible for the front tank), which obviously shifted the CoG, but not by a lot. It still overlapped with the CoL. The odd thing is, even if I control how fuel is being drained, the craft still is very unstable, and I do not know why.

Also, is anyone else having issues with that picture? It loads fine for me from different PCs.

[Skorpychan]

Part of the reason has to be drag from radial intakes. Ram intakes work far better, if you can figure out how to attach them all. Maybe stick tri-couplers on the ones in front of the aerospikes (two-side down to compensate for the flipping), and you look to be able to get away with bi-couplers on the ones ahead of the jets.

[AnalogAddict]

Part of the reason has to be drag from radial intakes. Ram intakes work far better, if you can figure out how to attach them all. Maybe stick tri-couplers on the ones in front of the aerospikes (two-side down to compensate for the flipping), and you look to be able to get away with bi-couplers on the ones ahead of the jets.

Yes, I am AnalogAddict and I approve this message.

[Cronos988]

Thanks again for the replies, that really helped. I put some RAM intakes (with engine nacelles) at the tip of the wing, and lost the middle engine. That fixed a lot of problems, as the plane now takes longer to build up speed, but is much more stable and has a better intake/engine ratio. I actually found out I could do with 2 engines because I took off a little too steep in one flight, and ripped of my middle engine. The plane continued to fly admirably, so i changed the design.

The backflipping is pretty much gone now, so I can handily fly the plane into orbit, without having to jettison anything! Landing is still a problem though, because with emptied rocket fuel tanks, the balance of the craft shifts massively, causing it to become uncontrollable upon re-entry. I guess I'll try transfering fuel around a bit to counteract that. For the time being though, I'll switch back to rockets and try getting a Kerbal to the moon (and back).