Spaceplanes tumble, turn tail-first during re-entry

[AHHans]

55 minutes ago, Spricigo said:

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?

Yes, I have considered that. I know how to read the Pitch/Yaw/Roll Indicators in the bottom left of the UI.

If you don't believe me, why not try it out yourself? Build a spaceplane with Mk2 parts and a delta wing that is just stable, cheat it into orbit, deorbit it and then watch its behavior during reentry and when in flight at < Mach 2.

Edited November 24, 2020 by AHHans

[Spricigo]

Just now, AHHans said:

Yes. I know how to read the Pitch/Yaw/Roll Indicators in the bottom left of the UI. If you don't believe me, ...

It go without saying that is not quite so easy for me to read the indicator in your monitor. That's why I need to ask what exactly had you seen/considered to know. 

 

12 minutes ago, AHHans said:

why not try it out yourself? Build a spaceplane with Mk2 parts and a delta wing that is just stable, cheat it into orbit, deorbit it and then watch its behavior during reentry and when in flight at < Mach 2.

Mostly lack on interest in using mk2 spaceplanes. While the info may still turn to be relevant for me at some point.

[foamyesque]

I'm pretty sure stock aero models a drag shift, yeah. It definitely models a drag spike through the supersonic region. I've had spaceplanes that have been completely uncontrollable through 90% of their entry suddenly flatten out and fly smooth as silk once I drop out of high transsonic, then turn to lawndarts in the subsonic area.

If your plane is tumbling on re-entry you have a CoP/CoM inversion, there's no question there, but the cause can sometimes be tricky to diagnose. KSP stock aero does model stalls these days; if you turn on the aero overlay and generate enough angle, you will see the blue arrows from your wings disappear and red arrows spawn. If *some* of your surfaces are stalled out, and others not, your CoP can move around significantly, a situation that not only is very likely to occur in a tumble, it can make the tumble worse. In addition, stalled out control surfaces are dramatically less effective at actually steering your ship; you can pilot a nearly-unstable craft reasonably easily if you have good control authority, but if it jams up on you things will rapidly go downhill.

As a general preference -- and particularly when working with a delta design as yours is -- I tend to use canards over tail-mounted elevators. It makes it easier to get a CoP nicely centered (which in turn makes centering the CoM on it easier), and they behave better in high-angle flight. It is also generally a good idea for anything going supersonic to use all-moving control surfaces, as real supersonic planes tend to.

Edited November 27, 2020 by foamyesque

[Spricigo]

10 hours ago, foamyesque said:

As a general preference -- and particularly when working with a delta design as yours is -- I tend to use canards over tail-mounted elevators

It amuses me how often, when it comes to spaceplanes, someone will state a preference for X over Y because they are better in some aspect just for the next guy tell is the other way around in his experience.

...and I suppose now I, the "next guy", don't need to even tell what my preference is.

[foamyesque]

2 hours ago, Spricigo said:

It amuses me how often, when it comes to spaceplanes, someone will state a preference for X over Y because they are better in some aspect just for the next guy tell is the other way around in his experience.

...and I suppose now I, the "next guy", don't need to even tell what my preference is.

 

Heh. It's true, they are a subject of debate! On non-delta designs, where your CoP naturally tends towards the midbody, you can build a tail out and have it work fine, using a couple of all-moving pitch control surfaces, maybe some autostruts to make sure they actually move the ship instead of just the tail.  Deltas, though, tend to put the CoP far to the back, and if you want a CoM that doesn't shift, it usually means it's further forward. Canards are very helpful there.

 

Sometimes I even use both. The alterations in drag between the nose and tail can give you significant control authority even in a stall situation, allowing you to force the nose down to get out of the stall, and reducing the strain on any supplemental control systems (RCS, wheels).