Twin Boar Yaw Gimbal Reversed

[Srpadget]

The title pretty much sums up my issue; I was wondering if this is a known bug, or if something's gone screwy in my install (and if the latter, how to fix it)?

Since 1.0 rebalanced jets, I have used SSTO vertical-launch rockets rather than spaceplanes.  Which to me generally means liberal use of the Twin Boar as a core booster once it's unlocked...and dealing with its quirks.  And by "quirks", I mean heating (of course), and my long-standing observation that Twin-Boar-based boosters seem to need unusually large fins/aero control surfaces.  The former is well-discussed here, but I don't think I've ever seen anything about the latter (maybe this should have been my first indication that something was wrong). 

I have noticed a tendency for my TB-based boosters to yaw west right around the beginning of my gravity turn, before eventually pulling east as intended.  I always assumed that was because my TB vehicles tend to have fins at angles other than 0/180 degrees (3- or 6-fold symmetry, or 4-fold in an x config rather than a + due to the TB's engine cowlings), thus no aero control surfaces suitable for pure yaw control.

Seems I was wrong.

With 1.0.5 and its rebalancing of heating, my fins have been blowing up during reentry.  Not a big deal, as by then they're not only unnecessary but a nuisance (big draggy things with lots of lift capability at what is now the FRONT end of my retrograde-oriented vehicle), but it was aesthetically displeasing, and didn't do any favors for the vehicle's economics.  So I decided to see if I could do without the fins -- I've seen many assertions here that larger vehicles actually have LESS need for fins, after all.  That experiment was an unmitigated disaster:  they tipped rapidly to the west and never recovered...unless, oddly, I commanded a yaw to the west.  In which case I got a prompt eastward gravity turn...up until it turned into a flip-and-tumble around Mach 1.  So I sat some TBs on the launchpad, arranged the camera to get a good view of the nozzles, and just hit the WSAD keys to see what happened.  W and S behaved as anticipated, but A and D resulted in counterintuitive nozzle motion.  As a check, I did the same thing with a Terrier: that nozzle moves as expected, so I haven't accidentally reversed the controls. (Also, the yaw monitor in the lower left deflects as expected, so again it appears the issue is with the TB's reversed response to a correct command.)
 

Is this a known bug?  Or is something corrupted in my install somewhere?  And how do I fix it?  Locking gimbals prevents the undesired gimbal behavior, but reaction wheels by themselves really don't provide enough control authority around max-Q (flip-and-tumble is NOT my friend, no matter how much Jeb and Val enjoy that roller-coaster feeling!), and the whole point of this exercise was to get rid of fins.  The most satisfactory solution is to fix the TB's nozzle response, but I'm at a loss as to even where to begin.

=================

EDIT: I should probably add, this game is mostly-stock.  Beautification mods EVE and Texture Replacer (but the only non-stock textures I use are suits so I can distinguish professions easily, no part textures), KAC, and Claw's Stock Bug Fix/StockPlus.  The latter is the only one that I would think could affect craft behavior, and I saw the "yaw west until speed is high enough for aero surfaces to 'bite'" behavior long before I ever installed that one.

Edited February 14, 2016 by Srpadget

[JTpopcorn]

This might need to go to support/staff, because you seem to have found a bug...

[Srpadget]

Well dang.  I was really hoping that someone would tell me something in the config got corrupted, and provide a quick-and-easy way to fix it.  ::grumble::

So how do I move this to a different sub-forum?

[RexKramer]

There is a known bug with the Twin Boar, which results in the engine having no roll authority despite the nozzles gimbaling in the correct direction.  However, the result you describe is different than the reported bug.

If possible, could you please post a screenshot of the vessel you are seeing this with?  

[NathanKell]

@RexKramer that was fixed in 1.0.5.

[Foxster]

OP...

I just tried putting this craft into orbit with just a TB...

Worked fine without fins. There was a slight wobble when I powered off when I had the Ap I wanted but that was easily recovered from. 

Want to try this craft yourself and see if it behaves oddly?

Oh, regarding watching the gimbal movement on the ground: Don't forget it will depend on the orientation of the capsule or probe core.   

 

Edited February 15, 2016 by Foxster

[RexKramer]

6 hours ago, NathanKell said:

@RexKramer that was fixed in 1.0.5.

Take a look at #5977 (and corresponding linked bug), which is actually new in 1.0.5.

I wasn't able to duplicate the results the OP had here, but the lack of roll authority in the twin boar (single centerline installation) is present in 1028.  The engine gimbals properly, even with opposite gimbaling with roll input, but no roll moment occurs (make sure you disable reaction wheels to test...) 

#5977 has a better explanation of the cause of the roll issue.

Also, howdy NathanKell!

Edited February 15, 2016 by RexKramer

[NathanKell]

Blargh, I *know* I fixed exactly that. Must not have gotten to the release version or something. Ah well...

EDIT: Yep, it's not in git. Apologies.

[blowfish]

I was able to replicate pitch/yaw flipping under specific circumstances - basically the thrust transforms must be below the vessel's CoM, but the part's CoM must be above the vessel's.  Probably because the gimbal module uses the part's root transform rather than the thrust transform  to determine whether to flip pitch and yaw.  Thanks @Shadowmage for suggesting reproduction steps (if indirectly).

Edited February 16, 2016 by blowfish

[RexKramer]

4 minutes ago, blowfish said:

I was able to replicate pitch/yaw flipping under specific circumstances - basically the thrust transforms must be below the vessel's CoM, but the part's CoM must be above the vessel's.  Probably because the gimbal module uses the part's CoM rather than the thrust transform  to determine whether to flip pitch and yaw.  Thanks @Shadowmage for suggesting reproduction steps (if indirectly).

That makes sense, and I see how it could be accomplished since the Twin Boar is such a long engine.  Used to be a problem with engines mounted above the COM would not gimbal the correct way (ex. Sky crane type landers..), that was fixed a while ago thankfully.

To eliminate the problem described here, the game would need to use the engine's thrust point(s) to determine if it is above or below the vessel's COM, not the engine's COM as it appears is occurring now.  

Thanks for the reproduction tip blowfish.

[Srpadget]

6 hours ago, blowfish said:

I was able to replicate pitch/yaw flipping under specific circumstances - basically the thrust transforms must be below the vessel's CoM, but the part's CoM must be above the vessel's.  Probably because the gimbal module uses the part's CoM rather than the thrust transform  to determine whether to flip pitch and yaw.  Thanks @Shadowmage for suggesting reproduction steps (if indirectly).

That would certainly explain what I'm seeing.  I have always had more issues with wobble from tall rockets than with drag from short squat rockets, so my SSTOs in particular tend to be squat.  it would not surprise me at all if the vessel CoM is below the Twin Boar CoM.  This would also explain why it does not seem to be a problem in the upper atmosphere (I'd been puzzling over that, as fins don't do much good up there and thus maintaining prograde becomes all about gyros and vectored thrust.  But if my problem is as blowfish describes, then I can expect thrust vectoring to work correctly later in the mission as fuel is burned and CoM moves forward.

[blowfish]

Err, slight correction to what I said earlier, it's the part's root transform, not the CoM.

[Srpadget]

Screenshot of the vessel in question, with CoM when fully-fueled on the pad.  And yes, the vehicle CoM is WAY LOW on the Twin Boar, but climbs as the outboard tanks drain.

[NathanKell]

@blowfish which is the CoM unless CoMOffset is used, so you're correct either way.

 

And the reason for my confusion: I did fix it, but apparently for 1.1 not 1.0.5. It blurs... :]