spaceplane 'bounces' when pitched up

[Stratoroc]

Certain spaceplane designs have an annoying behavior where upon using the pitch control the nose rises rapidly then falls back down, then rises again and falls again. This occurs most often with designs that have canards.

What causes that and how do I avoid it?

I'm using FAR if that matters.

[LethalDose]

Certain spaceplane designs have an annoying behavior where upon using the pitch control the nose rises rapidly then falls back down, then rises again and falls again. This occurs most often with designs that have canards.

What causes that and how do I avoid it?

I'm using FAR if that matters.

First, yes, using FAR matters, TYTYTY for stating it up front.

Second, pics are super helpful, because there may be a few things at work.

I would guess that it's a CoM/CoL balance issue combined with a lot of pitch authority (which is good), but it's hard to say without seeing the vessel. That vessel may just need a gentle hand on the stick, or a little more SAS to keep it better under control.

[Wanderfound]

Google up regex's PID controller mod. You can eliminate pitch bounce by cutting the default pitch values to about 1/3rd.

And, yes, FAR makes it more obvious, and excessive pitch authority will aggravate the issue.

[Stratoroc]

This is the plane in question. The control surface deflection is 9 degrees at the canards and 16 at the elevators. The elevators are the innermost control surfaces

edit:

So a little experimentation shows that a large ratio of control surface to wing area causes the control surfaces to stall. I swapped the wings for larger ones and the thing flies a lot better.

Edited October 24, 2014 by Stratoroc
science

[nikumai]

I'm by no means an expert, but when right clicked during flight do your canards stall at all? If so, the loss of lift means your nose will dip until those canards get air flowing properly over them.

[Franklin]

Can you pull your CoL back more? A tight CoM/CoL like that is going to make your control surfaces work that much harder to maintain higher inclinations (which may also explain why you find the 'bounce' happens more with designs that use canards).

Spreading your CoM/CoL out further will make maintaining pitch easier on the [weak] control surfaces.

edit: Fitting it with bigger wings may've indirectly did just this.

Edited October 24, 2014 by Franklin

[sal_vager]

Stock SAS over corrects in FAR, so yeah Regex's PID control can help, but it's more than just that.

Flight surfaces behind the CoM will naturally try to return to a position in-line with the airflow, but flight surfaces in front of the CoM will catch more air as they are deflected and will produce more lift, pushing you further round until they stall.

You can see this with any rocket that has winglets above the CoM, for example, getting to space becomes a lot more interesting.

Canards are great if you want a fighter jet and high manoeuvrability, modern fly-by-wire systems can keep them on track, and you have bags of lift to turn with when you need it.

But they aren't so great if you want stability, you can try smaller canards, move them nearer the CoM to reduce the moment arm, or omit them entirely, your elevons look to be far enough rearwards to still be quite effective.

[Franklin]

Gonna sound dumb, but take a pen and balance it on your pointer finger. With your other hand, press down on the "tail" of pen so the "nose" of the pen pitches up. Shift the balancing finger (CoM) further and closer from the point you're pressing down (CoL). It should take noticeably more downward force to maintain the upward pitch when the balancing finger is closer to the pushing finger.

That's the change in work your control surfaces need to deal with in your designs. It's reversed in this exercise obviously, cause you're balancing against gravity instead of lift, but it's easier (for me at least, I'm literally retarded) to see the CoM/CoL balancing game as a reverse cantilever.

edit: meant to back-to-back my posts but sal got in the way. now i look obsessive. >

[Alshain]

As others have said here, your Col is almost on top of your CoM, that makes it maneuverable not stable, which is desirable on atmospheric craft but SSTO's usually go 1 direction most of their flight and do not need maneuverability.

Also it looks like you are mounting the wings low, but not dihedral and you don't have a high tail plane, another aspect of fighter jets for maneuverability. Maybe not, it's hard to tell from that picture but if your CoL is below the CoM, either bring your wings up, make them dihedral, or add a T- tail or cruciform tail plane to bring the CoL back up to the CoM.

A decent reference, though I've never heard of "negative dihedral". I always have called that anhedral.

Edited October 24, 2014 by Alshain

[Tex_NL]

Especially when using canards with FAR/NEAR stall is usually the culprit when it comes to porpoising.

When you pitch up the nose rises. When you pitch up harder the canards will stall and loose lift; your nose will drop. You stop pulling up, the canards regain lift and your nose rises again.

I've had this issue several times and the easiest and as far as I know only solution is to reduce the max deflection angle on the canards. Doing so will sacrifice some pitch authority but they won't stall that quickly any more.

[Hodo]

As others have said here, your Col is almost on top of your CoM, that makes it maneuverable not stable, which is desirable on atmospheric craft but SSTO's usually go 1 direction most of their flight and do not need maneuverability.

I hate to say this but this would not be that big of a deal seein as FAR also emulates the mach affects, where you CoL will shift back towards the rear of the wing at supersonic and hypersonic speeds. So that craft would be quite stable at the speeds needed to achieve orbital velocities. His position is actually pretty much perfect for that style of SSTO. If he were to move his CoL further back away from the CoM he would run into a serious mach "tuck" affect where the nose would become extremely heavy at supersonic speeds, and quite possibly impossible to get any pitch up authority once the canards stall out do to the high speeds and high angle of attack required to bring the nose up at those speeds.

[Franklin]

He needs to be able to reach the applicable mach for that to be a consideration in design, though. And although I can't test his design, I don't think mach tuck under common circumstances is as severe as you're proposing.

edit: And I don't think anyone here's suggesting he crank his CoL far back. Just more.

Edited October 24, 2014 by Franklin

[Hodo]

He needs to be able to reach the applicable mach for that to be a consideration in design, though. And although I can't test his design, I don't think mach tuck under common circumstances is as severe as you're proposing.

With the current design it wouldn't be much of a factor, but if he were to move his CoL further back it would become quite a problem. Ideally you want the CoL just kissing the back of the CoM, this is what I have found to be the perfect position for any of my SSTO space planes, and I have built plenty of them probably far more than anyone else on these forums. All of them use FAR+DRE and have since I started playing KSP. I found that if you have the CoL to far back, it just makes the pitch up control that much more difficult at higher speeds. If it is to far forward you will find the craft impossible to control at any speed.

Like this craft, that I built for a challenge and used as a test craft for most of .24.

Javascript is disabled. View full album

Note its CoL possition is right behind the CoM. Also take note that at the control information on the bottom left of the screen, see how little actual pitch effort is being exerted on the craft to maintain its ascent angle.

That is one of my smallest SSTO space planes and cheapest, I built it for a test craft and used it for pretty much anything I didn't feel like using my bigger more expensive SSTOs for. I have no pictures of anything I have created this version yet. But I havent had a chance to really play much.

[Franklin]

I appreciate that your design flies, but it doesn't address what I'm suggesting: that mach tuck under FAR isn't as severe as you're suggesting, nor enough for him to not want to budge his CoL to address his stalling, which your concern doesn't approach.

I mean nevermind the numerous counter examples of other designs with a longer CoL/CoM gap reaching orbit perfectly stable under FAR.

[Geschosskopf]

A MAJOR cause of spaceplane bouncing when applying controls is the gimbal of the engines. It's a very good idea to lock all jet engine gimbals .

[Empiro]

I usually handle this by almost exclusively using trim settings to pitch up and down when taking a space plane into orbit. It lets you have much finer control and you don't have to keep tapping keys.