Spaceplanes tumble, turn tail-first during re-entry

[maddog59]

I'm sure I'm missing something in my construction or maybe re-entry technique: 

My spaceplanes have a nasty habit of getting to about 40-45km from the surface, then suddenly they start to tumble, usually ending up facing backward (maybe they're wanting to get back into space?   ). I've used SAS along with manual efforts to try to maintain the direction and heading, to no avail. I'm usually able to get back control by around 3-5km, but I'm getting really tired of the drama, as well as cleaning up the crew cabins. 

I've tested each of my sp's from the runway and they fly decently enough; the CoG and CoL's are overlapping both with and without fuel (I'm usually entering with a fair amount of fuel on-board). The RCS clusters I've installed seem to be ineffective in either maintaining or recovering during the re-entry. 

Any and all suggestions are welcome. Help me, Obi Kerbal Gods; you're my only hope!

[Caerfinon]

What pitch are you coming in at? I try to hold mine between 4 to 6 degrees. If I pitch too high then the ship wants to flip as the atmosphere gets thicker. If I pitch too low I don't bleed off speed fast enough.  

Mods that I find helpful;

• AtmosphereAutopilot (Fly-By-Wire) - in Standard Fly-By-Wire mode with Moderation and Coordinated Turns set on.
• Kramax Autopilot Continued - because I can fly, but embarrassingly, I don't so much land as explode 

 

[HebaruSan]

29 minutes ago, maddog59 said:

Any and all suggestions are welcome.

Pumping some of the remaining fuel towards the nose might help.

[maddog59]

41 minutes ago, Caerfinon said:

What pitch are you coming in at? I try to hold mine between 4 to 6 degrees.

That's pretty shallow, but I suppose I can give it a try... I've seen YouTube's of people coming in at 45-90 degrees, and I've tried manually between ~10-20; and once I tried letting it follow the SAS set to prograde, without any difference. But maybe there's a sweet spot. 

I'll look at the mods you suggested as well; thanks!

32 minutes ago, HebaruSan said:

Pumping some of the remaining fuel towards the nose might help.

Nice idea, but my planes are small, with just a single tank,  so there's nowhere for me to put them. 

On a related note, is there a way to see where the CoG and CoL currently are, while in flight? That would really be handy...

Edited November 22, 2020 by maddog59

[Vanamonde]

In flight, you can rotate the camera around your craft. It defaults to centering on the center of mass, so the point that doesn't move as you rotate is the current COM. I know of no tricks to find the COL in flight. There's the aero forces display, but it's hard to interpret. 

How about a picture of your craft? This often helps with diagnoses.  

Oh, and your question has been moved to the gameplay questions sub, since you're seeking help rather than offering it. 

[Caerfinon]

1 hour ago, maddog59 said:

But maybe there's a sweet spot

The spot is different for each plane depending on design. Try and fail until you don't... this is the way (or at least the Kerbal way).

[maddog59]

1 hour ago, Vanamonde said:

In flight, you can rotate the camera around your craft. It defaults to centering on the center of mass, so the point that doesn't move as you rotate is the current COM. I know of no tricks to find the COL in flight. There's the aero forces display, but it's hard to interpret. 

How about a picture of your craft? This often helps with diagnoses.  

Oh, and your question has been moved to the gameplay questions sub, since you're seeking help rather than offering it. 

And thanks for moving the question... I misunderstood the title of the Tutorials section...

Edited November 22, 2020 by maddog59

[jimmymcgoochie]

Spaceplanes tumbling and ending up tail-first is usually because the centre of mass moves behind the centre of drag, which is aerodynamically unstable- in everything from badminton shuttlecocks to styrofoam gliders to archery arrows, the heavy bit needs to be at the front and the draggy bit at the back to keep it pointing the right way. Heavy rocket and jet engines on the tail is a common reason for spaceplanes being unstable during re-entry.

Try- shifting your centre of mass further forwards by changing the fuel drain priority of the fuel tanks to drain any nose-mounted tanks last or not at all; using the inline Mk2 cockpit with a Mk2-1.25m tank then an NCS tank set as the lowest fuel priority; adding some RCS thrust to the top of the nose which will push the nose down during re-entry; adding some sort of drag at the tail by deploying all the elevons/tail fins with an action group (I usually use abort/backspace); increasing the control authority on your front canards so they don't end up adding drag at the nose when it's at a high angle of attack; and/or sticking to a fairly shallow pitch which will make it harder for the plane to start pivoting.

[FleshJeb]

@maddog59 I can’t figure out the problem from the picture, but this is a very helpful resource on spaceplane design and general aerodynamics: 

 

[AHHans]

The center of pressure (where the sum of all aerodynamic forces apply) shifts with speed and air density! At low speeds (< Mach 2) and high density it is dominated by lift, which is the center of lift that is shown in the editor in the game. But at high speeds (> Mach 3) and low air densities the aerodynamic forces are dominated by drag. In a delta-wing design - like yours - this means that the center of pressure moves forward in the plane and can get in front of the center of mass.

So, yes, your plane became aerodynamically unstable, and the solution is to move the center of mass further forward.

[vv3k70r]

17 hours ago, maddog59 said:

My spaceplanes have a nasty habit of getting to about 40-45km from the surface, then suddenly they start to tumble, usually ending up facing backward (maybe they're wanting to get back into space?

Atmosphere is being weak at this point and so Your stability depended on wings turn to be COM dependent. How is Your COM balance for this?

Close intakes! Block the front fuel tanks from being used.

You can try to work around (by wasting fuel) by pointing prograde before it happen and then roll around prograde so it will try to turn over always in Your prograde direction. But working with design could help to avoid this madness.

Canards helps, but they would create more drag ASL thus wasting fuel.

Another way to work around is to get high up quite straight and then waste fuel to follow the horizon on orbital altitude.

Back to hangar - back to desing?

17 hours ago, maddog59 said:

The RCS clusters I've installed seem to be ineffective in either maintaining or recovering during the re-entry. 

Force is to great to counteract with so small amount of newtons.

16 hours ago, maddog59 said:

On a related note, is there a way to see where the CoG and CoL currently are, while in flight? That would really be handy...

In SPH You can free some fuel from tanks to see what happen to COM in such a case.

 

14 hours ago, Vanamonde said:

There's the aero forces display, but it's hard to interpret. 

Issue exist because these forces dont do much on this altitude. So there is not much to read from F12.

 

Edited November 23, 2020 by vv3k70r

[king of nowhere]

when it comes to spaceplanes, this forum tend to be quite single-minded with center of mass and lift and drag. sure, it is solid advice, but sometimes you do have important reasons to not shift them around, and there are other solutions.

i am surprised nobody is mentioning extending the landing gear. i don't see it in the image, but I assume you do have retractable wheels for landing. it greatly increases the drag, acting as airbrakes, and they have excellent thermal resistance. now you can afford to enter atmosphere without any angle, and still you will brake  enough. even better, if you follow a standard setup with two wheels in the back and one in the front, it will increase drag in the back and help stabilize your plane. I have a plane where my center of mass is in the back and needs to be in the back (it's a seaplane, it takes off from water, and it needs a CoM in the back to be able to point the tip upwards while floating), and that's what i do. keep prograde during reenetry, with landing gear extended and cargo bay opened for drag.

I see all Mk2 parts, they are sub-par for most performances but the one good thing they have is thermal resistance. there should be no way you burn in atmosphere. i tried to make a mk2 plane only once, and i lost control of it during reentry, but even though i spent the whole reentry flapping around without control, i still slowed down without burning, and i was able to recover control in the low atmosphere. there should be no way you get incinerated. my space plane is not Mk2, but it can reenter prograde just fine.

[vv3k70r]

37 minutes ago, king of nowhere said:

now you can afford to enter atmosphere without any angle

I suspect the problem is with getting out, not in the atmosphere.

[king of nowhere]

On 11/22/2020 at 5:00 PM, maddog59 said:

I'm sure I'm missing something in my construction or maybe re-entry technique: 

My spaceplanes have a nasty habit of getting to about 40-45km from the surface, then suddenly they start to tumble, usually ending up facing backward (maybe they're wanting to get back into space?   ).

 

8 minutes ago, vv3k70r said:

I suspect the problem is with getting out, not in the atmosphere.

The bolded part strongly implies they are getting back from space

[JadeOfMaar]

1 hour ago, king of nowhere said:

when it comes to spaceplanes, this forum tend to be quite single-minded with center of mass and lift and drag. sure, it is solid advice, but sometimes you do have important reasons to not shift them around, and there are other solutions.

i am surprised nobody is mentioning extending the landing gear. i don't see it in the image, but I assume you do have retractable wheels for landing. it greatly increases the drag, acting as airbrakes

I can understand the claim about the forum being single-minded, but the idea of using your airplane wheels as airbrakes (while very novel and interesting) is simply something that no one would reasonably or sensibly consider.

8 hours ago, vv3k70r said:

Close intakes!

That has no effect on aerodynamics. There is no visible animation associated with it (like deploying chutes or inflating a shield). If it does have an effect, the effect is so small that it's not worth your time to bother with. Closing intakes only matters to mods that give it a strategic value, like mods for exotic jet engines or mods that make it so letting reentry plasma in will destroy the engine.

8 hours ago, AHHans said:

The center of pressure (where the sum of all aerodynamic forces apply) shifts with speed and air density! At low speeds (< Mach 2) and high density it is dominated by lift, which is the center of lift that is shown in the editor in the game. But at high speeds (> Mach 3) and low air densities the aerodynamic forces are dominated by drag. In a delta-wing design - like yours - this means that the center of pressure moves forward in the plane and can get in front of the center of mass.

So, yes, your plane became aerodynamically unstable, and the solution is to move the center of mass further forward.

Oooh. I like this. Have you confirmed this in-game or is this info based solely on research?

On 11/22/2020 at 12:00 PM, maddog59 said:

I'm sure I'm missing something in my construction or maybe re-entry technique: 

My spaceplanes have a nasty habit of getting to about 40-45km from the surface, then suddenly they start to tumble, usually ending up facing backward (maybe they're wanting to get back into space?   ). I've used SAS along with manual efforts to try to maintain the direction and heading, to no avail. I'm usually able to get back control by around 3-5km, but I'm getting really tired of the drama, as well as cleaning up the crew cabins. 

I've tested each of my sp's from the runway and they fly decently enough; the CoG and CoL's are overlapping both with and without fuel (I'm usually entering with a fair amount of fuel on-board). The RCS clusters I've installed seem to be ineffective in either maintaining or recovering during the re-entry. 

Any and all suggestions are welcome. Help me, Obi Kerbal Gods; you're my only hope!

Having the CoM and CoL overlap is fine for stunt planes (as they need to be able to turn very easily at any time) but not for spaceplanes which need to be able to hold their attitude most of the time. That overlap is for purposeful aero instability which you don't want. Bring the CoM forward, and where possible, pull back the CoP ("Pressure" -> Drag). Parts that add to the CoP's position are adapters, cockpits, surface attached objects that have plenty forward/prograde surface area, and stack parts with protruding bits like the mk1 Divertless Supersonic Intake (a Mk1 tank with intake scoop).

The more your CoL is behind the CoM, the more your plane can resist flipping over, but bring it too far back and you can't pitch at all --> "Lawn Dart Syndrome." Consider increasing your control surface area too. The more you can do aero deflection (for pitching) in thin atmo, the better.

[AHHans]

1 minute ago, JadeOfMaar said:

Oooh. I like this. Have you confirmed this in-game or is this info based solely on research?

I have confirmed this solely on research in-game.

If you mean "Have you had space planes that flew fine at low speeds flip out when trying to accelerate to orbital velocity?" or "Did your space-planes that you trimmed to be just stable during reentry become annoying nose-heavy for the low-level flight to the KSP?" then the answer to both is: yes. Another thing I learned is that the Mk2 parts are particular prone to this problem.

P.S. I trim my plane by moving the remaining fuel around, and my definition of "just stable" is that when at 90 deg to the airstream (= belly first) it just so tends to pitch the nose forward and the the aft. And, yes, I regularly pump fuel back aft after transitioning to "normal" flight.

[king of nowhere]

16 minutes ago, JadeOfMaar said:

I can understand the claim about the forum being single-minded, but the idea of using your airplane wheels as airbrakes (while very novel and interesting) is simply something that no one would reasonably or sensibly consider.

 

....

are we playing the same game?

I don't know, perhaps you play solely by making realistic replicas of real vehicles and flying them realistically. But that's not the case for most here.

Can't you hear the cries for moar boosters and moar struts? Have you seen the huge asparagus stack, the kerbalized vehicles, the early landers made with inadequate parts and kept together with tape? the kerbonauts heroically sent flying from kerbin to jool in an external seat, those parachuting out of their vehicles during a reentry, those doing orbital manuevers on their jetpacks alone; the rovers with roll bars, the engine cones used as landing legs, the thermal shields used as wings? Not to mention the kraken drives and the magic wings (i.e. kraken aerodynamics)

If we made a ranking of all the inane and outlandish ideas people have tried in this game and have made to successfully work, "using your wheels as airbrakes" wouldn't even break the top100.

furthermore, using airbrakes is not an outlandish idea, but the only parts that the game provides that are supposed to be used as airbrakes have low temperature tolerance and so are useless in a reentry. can you really blame people for using other parts to achieve the same effect? And if it was parts that they had to put on their plane already, causing no extra weight, that's just a plus

 

[JadeOfMaar]

18 minutes ago, AHHans said:

I have confirmed this solely on research in-game.

Wise guy.

@king of nowhere ok you got me. I didn't easily remember all that. It really does suck a lot that the only actual airbrake is only for fighter planes and stunt planes to use.

[vv3k70r]

51 minutes ago, JadeOfMaar said:

That has no effect on aerodynamics.

Try^^

51 minutes ago, JadeOfMaar said:

There is no visible animation associated with it (like deploying chutes or inflating a shield).

Airspeed taken by intake result in drag directly.

51 minutes ago, JadeOfMaar said:

Closing intakes only matters to mods that give it a strategic value, like mods for exotic jet engines or mods that make it so letting reentry plasma in will destroy the engine.

I'm playing plain game without mods and I see diference with different airspeed in climbing and descending. Closing and opening intakes affects aerodynamic.

51 minutes ago, JadeOfMaar said:

Bring the CoM forward, and where possible, pull back the CoP ("Pressure" -> Drag).

On high altitude with going to orbital speed COM is more important then drag - press F12 to see.

 

25 minutes ago, king of nowhere said:

the kerbonauts heroically sent flying from kerbin to jool in an external seat

And Kerbals are proud of that - keep a deep breath and take a seat, next station Jool.

25 minutes ago, king of nowhere said:

furthermore, using airbrakes is not an outlandish idea, but the only parts that the game provides that are supposed to be used as airbrakes have low temperature tolerance and so are useless in a reentry

I'm using cheapest winglets - they burn exactly on the edge of thermosphere when I decoplue boosters with them. They do the job.

1 hour ago, king of nowhere said:

The bolded part strongly implies they are getting back from space

In such a case I can only blame the pilot.  I had same trouble on the beggining untill I get what are the ways to descent something with wings. It is not so easy to pass from fall to flight.

Easiest is to prolonge high altitude by lowering apoapsis inside the atmosphere after passing periapsis on its edge.

 

 

 

Edited November 23, 2020 by vv3k70r

[king of nowhere]

24 minutes ago, vv3k70r said:

I'm using cheapest winglets - they burn exactly on the edge of thermosphere when I decoplue boosters with them. They do the job.

 

See? using something creatively for a purpose for which it wasn't originally designed! that's the spirit! that's progress!

after all, humankind would have gotten nowhere if it had stuck with what it was supposed to do. silicon, for example, is supposed to be a rock. it reacts with oxygen, with which it has a really high affinity, then it is a rock. sometimes it gets fluorinated, because it's got a high affinity for fluorine too. That's what silicon is supposed to do. And we're supposed to walk on it. Sometimes pick it up and use it to hit stuff, but only if we feel particularly fancyful.

and we went and turned it into the core of a pseudo-thinking machine. that we're using to build and fly unlikely contraptions. I don't think anyone would have ever sensibly considered such a thing before it happened

[sturmhauke]

1 hour ago, king of nowhere said:

If we made a ranking of all the inane and outlandish ideas people have tried in this game and have made to successfully work, "using your wheels as airbrakes" wouldn't even break the top100.

I don't have a problem with using crazy tricks, but landing gear as airbrakes just seems especially ridiculous somehow.

[sturmhauke]

1 hour ago, AHHans said:

I have confirmed this solely on research in-game.

If you mean "Have you had space planes that flew fine at low speeds flip out when trying to accelerate to orbital velocity?" or "Did your space-planes that you trimmed to be just stable during reentry become annoying nose-heavy for the low-level flight to the KSP?" then the answer to both is: yes. Another thing I learned is that the Mk2 parts are particular prone to this problem.

P.S. I trim my plane by moving the remaining fuel around, and my definition of "just stable" is that when at 90 deg to the airstream (= belly first) it just so tends to pitch the nose forward and the the aft. And, yes, I regularly pump fuel back aft after transitioning to "normal" flight.

I'm not sure that's evidence of the CoP moving around at different speeds, so much as just the CoM moving around as you burn fuel. FAR models the behavior you're talking about, but as far as I'm aware stock aero doesn't. As for the CoM movement, I use the EEX mod so I can see wet, dry, and average CoM in the editor before I ever launch. I try to design my planes and spaceplanes so that it doesn't move much over time - especially useful for VTOL craft like this shuttle I'm working on. Yes, it uses a lot of part mods, but the principle applies regardless. The VTOL engines are at the fore and aft of the outboard nacelles.

[AHHans]

52 minutes ago, sturmhauke said:

I'm not sure that's evidence of the CoP moving around at different speeds, so much as just the CoM moving around as you burn fuel. FAR models the behavior you're talking about, but as far as I'm aware stock aero doesn't.

Well, during the "just stable during reentry" to "nose-heavy during flight" case I didn't burn any fuel in between. And to show this behavior stock aero only needs to implement different dependencies of the drag (well, whatever the red arrows in the aero overlay represent) and lift (blue arrows) on the airspeed and/or air density. (I believe the biggest effect is that the drag increases faster with airspeed than the lift does.)

52 minutes ago, sturmhauke said:

I try to design my planes and spaceplanes so that it doesn't move much over time

Same here! Except that I don't use a mod, but manually empty the tanks in the SPH to see how the CoM moves.

Edited November 23, 2020 by AHHans

[vv3k70r]

2 hours ago, king of nowhere said:

then it is a rock. sometimes it gets fluorinated, because it's got a high affinity for fluorine too.

There is in fact glass-steel in production. It is a very spring like material. It dosent deform as easy as mettal alloys. Its base on silicon instead of coal. I thing You can buy it already as a "foil" on ebay.

2 hours ago, king of nowhere said:

See? using something creatively for a purpose for which it wasn't originally designed! that's the spirit! that's progress!

Its my trade.

 

 

 

[Spricigo]

13 hours ago, king of nowhere said:

when it comes to spaceplanes, this forum tend to be quite single-minded 

Do you think? I mean: we barely agree on "if draggy bit are ahead of heavy bits, craft tend to flip". 

Alas, we can see in the first two reply to that very thread that Caerfinon and HebaruSan considered different possible reason why that is happening with OP's craft. Is also interesting that what they suggested was not to change CoM/CoD relationship but rather preventing the change. Actually you suggested, just using the landing gears variable drag to achieve it.

16 hours ago, king of nowhere said:

center of mass and lift and drag. sure, it is solid advice, but sometimes you do have important reasons to not shift them around,

And  it's only fair to expect the OP to point out if such reasons exist. 

15 hours ago, king of nowhere said:

i am surprised nobody is mentioning extending the landing gear

In other hand, I'm not surprised at all. OP mention it is a systematic problem (in his own words: is an habit of his spaceplanes), thus IMHO a more general solution is likely to be a better first approach. Is not really necessary to think outside of the box while we still have unexplored option inside.

Also keep in mind that not only the way each one of us play the game but also the way each one of us offer advice is a matter of preference to some extent.

 

15 hours ago, AHHans said:

If you mean "Have you had space planes that flew fine at low speeds flip out when trying to accelerate to orbital velocity?" or "Did your space-planes that you trimmed to be just stable during reentry become annoying nose-heavy for the low-level flight to the KSP?" then the answer to both is: yes.

Maybe you had spaceplanes slightly unstable with reaction wheels just strong enough to maintain attitude at low speed/ high altitudes. Had you considered that hypothesis?