why do my aircraft wobble and fall out of the sky?

[kodemunkey]

I find this happens a lot, it doesn't matter if i have a propeller driven aircraft or a jet engined aircraft, when in flight with SAS mode on, the aircraft wobbles and falls out of the sky a few seconds later.

[Pecan]

NMI 10 char ;-)

IE: We need just a bit more to go on than that. Pictures, stats, etc. Without any of that the only guess would be 'too much drag ahead of CoM'.

[Tortoise]

Well first we should ask...Do you use FAR?

And...May we see a screenshot?

[kodemunkey]

I'll post a screenshot in the morning. I've recently started using FAR, is it overkill to have Mechjeb too?

[Pecan]

It's not overkill but very often MJ doesn't play nicely with FAR. Are you engaging 'just' SAS or MJ's version?

[kodemunkey]

Vanilla SAS, I can't understand what i should be doing with Mechjeb so i should probably get rid of it.

[Pecan]

Keep MJ, as you find out more the information displays it can give you are really useful and flexible. Unfortunately, that takes us back to having to see some pics/vids of you ships and more detail on when the trouble starts. My own guess - which is all it can be at the moment - and based on bitter personal experience :-( is that you have air-intakes ahead of the CoM and as you gain speed their drag makes the thing want to flip. In FAR getting too much stress off-prograde can even tear the thing apart, but in any case it'll stall and not fly.

First thing I'd suggest is seeing if they'll fly without FAR, just to check. At least then you'll know whether it is the more realistic aerodynamics that are causing the trouble.

Hey - pretty good 'asked to responses' time you've got here :-) Fingers-crossed that it leads to something helpful.

[Doh! To me - 'Aug 2012' - you're not exactly new to this, are you? Sorry if I'm teaching you to suck eggs.]

[kodemunkey]

I don't play very often, in fact i lurk more in the mods forum than i post.

[kodemunkey]

I prefer to build drones as i feel a bit..guilty about the amount of Kerbals that die when i try and build things.

Also, if you wish i can record the video again and save in a higher quality.

[Tex_NL]

Try throttling back a bit.

The FAR display clearly states High Dyn Pressure. You're pushing you're way trough the sound barrier. With FAR's reduced structural integrity your plane simply can't handle the load and breaks apart. That's all.

[Vaporo]

Well, you have those control surfaces towards the front there. They are your problem, I think. If they catch any air, they flip your craft over, especially above mach one. Remove these, or at least increase their maximum deflection angle. (removal is probably better.) The wobble is probable caused by a combination of the first thing and the fact that you are using such large control surfaces. In other words, too much control authority.

[Creature]

On a glimpse it looks like the SAS contributes to the wobble as it overcorrects back and forth. Also as noted you're flying too low and fast so even the small wobble is a lot of stress on the craft. You should be able to fly without SAS anyway if the craft is stable.

You can try going slow and get some more altitude before hitting max throttle but i agree that you should try losing the front canards.

[kodemunkey]

Ah, that used to happen to me before I installed FAR too, even with a propeller driven drone.

[Zac2Bac]

Needs moar struts.

[Kryxal]

Agreed, over-controlled, see what happens if you turn SAS off for a second or two. I also wonder how stable it is, a screenshot in the hangar with CoM and CoL turned on would help matters...

[Tex]

Oh yeah- the problem isn't falling out of the sky, it's falling apart! Just take it a bit easier around the curves, and until you get past Mach 1 you need to keep real smooth.

[LethalDose]

Try throttling back a bit.

The FAR display clearly states High Dyn Pressure. You're pushing you're way trough the sound barrier. With FAR's reduced structural integrity your plane simply can't handle the load and breaks apart. That's all.

Yep, what Tex_NL said. The low atmosphere is too thick for aircraft to go that fast and hold together. You can slow down or break apart. From experience, I try to keep my speed below 250 m/s under 5000 m altitude on Kerbin. You can push it to around 500 m/s at around 10 km.

Basically, anytime FAR is telling you High Dyn[amic] Pressure, you're exceeding the safe limits for the plane. Ignore that warning, and you're gonna have a bad day.

Edit: Wow, I'm surprised how many posters are chalking this up to a control surface issue. I use canards on the fore-section of most of my spaceplanes with great results. They provide great pitch authority and help balance out CoL. I just keep the speed under control in the low atmosphere. Obviously, everyone builds their vessels in KSP differently, so maybe it just works well for me.

I would say to the OP take the same vessel to > 8000 m altitude (keeping speed under 250 m/s until you got there) and then open up the throttle and see if you have the same problem with the wobble and the falling apart and sadness before FAR tells you "High dynamic pressure again".

If you still have the problem before you get the warning, yeah, I guess it's a control surface issue.

If you don't have the problem before you get the warning, I think it's your speed.

Edited June 12, 2014 by LethalDose

[Creature]

Edit: Wow, I'm surprised how many posters are chalking this up to a control surface issue. I use canards on the fore-section of most of my spaceplanes with great results. They provide great pitch authority and help balance out CoL. I just keep the speed under control in the low atmosphere. Obviously, everyone builds their vessels in KSP differently, so maybe it just works well for me.

Well there's two issues, one is the wobbliness which is a SAS and control surface issue, but of course the disassembly is caused by the aerodynamic force. I've used canards sometimes successfully too, but sometimes I get backflips because of them and have no idea how to predict it so as a rule of thumb I'd say try losing them on a problematic craft and see if it helps.

I might be wrong of course but from the video it looks like the SAS is overpredicting and would probably eventually push the plane to a spin anyway, at least in hypersonic speeds. I wouldn't fly a plane at mach 3 that wobbles like that and has half of wing surface area as control surfaces. But as a disclaimer I haven't had a plane break apart in-flight on me (mainly because I haven't flown planes much lately) so can't really say much about that but it's stated pretty clearly in the flight report that stuff breaks due to aerodynamic stress I'd imagine that the wobbliness doesn't help with that.

[LethalDose]

Well there's two issues, one is the wobbliness which is a SAS and control surface issue, but of course the disassembly is caused by the aerodynamic force.

Okay, fair enough, there are two issues: Wobbling, and the plane falling apart. At subsonic speeds with FAR, I also see some wobble with SAS. In my experience, it's completely independent of the control surfaces on the plane, meaning if I add or remove control surfaces, it doesn't get better. And when I get up to higher altitudes, it tends to go away.

I'm interested to see what happens if all he does is remove the canards. Given how he's launching, I'm guessing the plane is going to nose down and crash, probably before the end of the run way. Right now, those canards are providing lift on the front of the plane to balance lift from the wings at the back of the plane. This keeps the CoL near to the CoM. If the CoL is way behind the CoM (as would happen if he removed the canards), The CoM acts as a fulcrum, and the upward lift at the aft of the plane causes the the craft to rotate/pitch so the nose will point downward.

Which is a good segue to this:

I've used canards sometimes successfully too, but sometimes I get backflips because of them and have no idea how to predict it so as a rule of thumb I'd say try losing them on a problematic craft and see if it helps.

If you put the canards too far forward, the CoL also moves too far forwards, and the opposite problem happens: too much unbalanced lift at the front of the plane cause the nose to rotate/pitch upwards, leading to backflips/stalls/bad days. When you're building vehicles in the SPH, take a look at the CoL & CoM indicators. If your CoL is way in front of your CoM in those planes with canards doing back flips that you can't predict, I think this will help you predict them.

REGARDLESS! On to this:

I might be wrong of course but from the video it looks like the SAS is overpredicting and would probably eventually push the plane to a spin anyway, at least in hypersonic speeds. I wouldn't fly a plane at mach 3 that wobbles like that and has half of wing surface area as control surfaces. But as a disclaimer I haven't had a plane break apart in-flight on me (mainly because I haven't flown planes much lately) so can't really say much about that but it's stated pretty clearly in the flight report that stuff breaks due to aerodynamic stress I'd imagine that the wobbliness doesn't help with that.

I agree that in the video the SAS induced wobble screwed up the AoA, causing the aerodynamic failure, and therefore the wobble immeadiately precipitated the aero failure. However, I would argue that plane was going to fall apart in mid air regardless of the control surfaces. Even with no SAS or no canards or both, the high dynamic pressure is going cause instability, and then wobble which is going to screw up the AoA, even without the canards.

An aircraft that size going full throttle off the tarmac with two turbojets and hitting mach 1 before cresting 2500m is simply boned. And it looks like we agree on that, since your intitial response said throttle back and go higher.

Anyway, i don't think that plane in that video is falling apart due to a engineering flaw, I think it's operator error. IMO, the canards are fine, so long as the CoL and CoM are near to each other.

Oh, also, to clarify my perspective: I've had lots of planes break apart in flight in testing designs, but it's almost always been caused by too rapidly changing my area of attack due to maneuvers in thick atmosphere. I've never had that wobble cause it, but I've only come back to aircraft recently, so my experience is also somewhat limited.

Edited June 12, 2014 by LethalDose
Mispelling

[Creature]

Anyway, i don't think that plane in that video is falling apart due to a engineering flaw, I think it's operator error. IMO, the canards are fine, so long as the CoL and CoM are near to each other.

Yeah, I think we're on the same page. Mainly my point about the canards was that if you generally go very fast with so much control authority, it might be very hard to input any pitch controls without flipping the plane over and also at hypersonics there's some weird aerodynamic stuff going on with the taperings and such which I think has usually been my problem but like I said it's all magic land to me. But it's true that the engineering definitely isn't the issue here and the plane flies neatly until it breaks mach one so talking about the canards or control surfaces isn't actually relevant to the problem.

Now I got a sudden urge to build some planes and see how much they can handle D:

[Kryxal]

I wasn't saying anything against canards exactly, but canards AND elevons may be overdoing things ... I might be happier seeing the control surfaces at the rear go away.

[kodemunkey]

In my drone's current configuration i got it up to 6,300 feet and 920 mph before it started wobbling more than i could handle. I'll work on modifying the design tonight and post the results.

[DeMatt]

In my drone's current configuration i got it up to 6,300 feet and 920 mph before it started wobbling more than i could handle. I'll work on modifying the design tonight and post the results.

...It's worth noting that the altitude and speed readings in KSP (and in your video earlier) are in Metric, not Imperial, units.

So while 6300 metres is three times higher than you think it is, 920 metres per second is a LOT faster than 920 miles per hour.

[kodemunkey]

Well, i told FAR to put the speed into Miles Per Hour so i'd have a small chance of figuring out how fast i was moving.

[Hodo]

Ah, that used to happen to me before I installed FAR too, even with a propeller driven drone.

Your problem with the craft you posted the picture of is simple. You have not set your control surfaces jobs. Right now it looks like they are all doing everything which is WAY to much control for that small of a craft going that fast. So as it is adjusting its pitch it starts to oscilate which in turn causes it to over correct which only makes the oscilation worse. Till eventually you are going to fast and the airpressure is so high that you litterally rip the wings off the craft.

The fixs for this are simple.

1- Set your control surfaces jobs... The canards on the nose, set for roll, set your elevators on the wings for pitch, and the tail for yaw. You may want to reduce the max angle those controls can move to prevent over correction and over control.

2- struts..... B9 has some wonderful invisible struts that are great in keeping the craft strong. So you dont have to worry about the craft flexing apart under high dynamic pressure.

Lastly as others have said... you are going WAY to fast at to low of an altitude. Open up the flight data window in FAR and you will see you probably were pushing nearly 90kpa at that speed at that alitude. Which is beyond high, that is ******ed.