Problem with inverted control surfaces

[Shrike99]

KSP Version: v0.0.90.0.705(0.90.0.705) Windows 64-bit

Issue:For some reason the control surfaces on a recent space-plane design are behaving incorrectly, the issue is that elevators behind the Com are pitching the wrong way(ie the same way as the canards at the front). i have reattached the surfaces a few times however the issue persists. although it causes the plane to fly in a very entertaining manner it is not very helpful at into space.i am 95% sure this is a bug, though it may be a simple design flaw i overlooked

pics below may help explain my ramblings:

Images:

https://www.dropbox.com/s/o5g8bzccf3c09rd/screenshot0.png?dl=0

https://www.dropbox.com/s/bjsqyggcldgymzk/screenshot1.png?dl=0

these images show the issue, note that the rear control surfaces are lifting the back, despite my attempting to pitch up

https://www.dropbox.com/s/km4brbt0gwr890g/screenshot2.png?dl=0

this picture shows that the com/col is well ahead of the control surfaces affected

Craft file:

https://www.dropbox.com/s/84usjyczzvareht/ISSTO.craft?dl=0

if anyone tests this please note the ion engines have no power source atm, activate the main using action group 1 rather than staging key if you attempt to fly.

additionally, the issue only occurs once the com has moved back a small amount, meaning that when i first tested it the plane flew fine at first, however the plane linked and in the screenshots is one that has had its fuel layout tweaked to match the plane in later flight for the sake of saving time.

would be grateful for input as to whether this is just poor design, and also for someone else to have a look at the .craft and attempt to replicate,

thanks in advance

[sal_vager]

Known issue I'm afraid, because of the angle of the elevators the game thinks they are in front of the CoM, so makes them move the wrong way.

If you use less of an angle on the elevators you can prevent it from happening

[Supernovy]

Just out of interest, is the wing that the control surfaces are attached to attached to the plane ahead of the CoM?

[Shrike99]

Just out of interest, is the wing that the control surfaces are attached to attached to the plane ahead of the CoM?

Away from pc atm, will edit in a few hours if im wrong, but i do believe that the attachment point of the wings may be ahead, though the majority of the wing surfaces isnt

[Shrike99]

Oh i think i understand

basically this:

https://www.dropbox.com/s/enqunkwuru0syrj/got%20it.png?dl=0

if the control surface judges its position relative to the com like that i understand.

feel free to change this to solved or file it under known bugs or what ever it is organised people do:)

Edited December 20, 2014 by Shrike99

[Anthlon]

This is still happening to me even in 1.0, anybody solution for that? It is just as Shrike99 wrote, if the line is before CoM then it is inverted :-\

[WinkAllKerb'']

+1 annoying as it often "drastically" limit a lot of some crafts designs variations and possibilities by some totally counterintuitive elevator orientation results due to native unity part hard-binded axis orient.

edit ... ; ): tweakable invert could do the trick also on "flap" part anyway @ 1.04 invert result outside the sph are still a bit random :/

(probably main COM // flaps @ parent(s) part(s) root attach COM position to each other and unity axis)

(eventually check for parents parts "chain" // main body COM)

=> same craft with 1 grid offset for flaps parent part // main com

(*04bis.craft has inverted control results on runaway and the inverted tweakable from sph active or not has no effect once on runaway while *04.craft work properly)

https://drive.google.com/file/d/0B5JHmZea7s91Q21rN3ZYUG16OG8/view?usp=sharing

https://drive.google.com/file/d/0B5JHmZea7s91VVNxUVpHcDJYRFU/view?usp=sharing

COMs and attachement related to each other might be more annoying to solve for control surface directly attached on main body like delta-deluxe, tail fin and else :/

Edited July 29, 2015 by WinkAllKerb''

[WinkAllKerb'']

Oki so here a clue // to attachement and offset after a few more tests

part attached backward then offset front: work properly

part attached directly in the same position without offset: don't work properly

also notice that time to time & very rarely depending launch from sph or loading a plane from the runaway flap could be inverted not been able to reproduce that each time succefully yet. (might had something to do with memory usage may be, not sure)

this tend to confirm that the problem come from the main COM and flap parent part COM location to each other during (primary (?)) attachement and eventually the use of offset or not afterward.

Edited July 29, 2015 by WinkAllKerb''

[WinkAllKerb'']

also adding some wing in front of the *04.craft file:

while moving the craft COM to a better location do not change the fact flap continue to work properly wich tend to confirm the flaps parent(s) part(s) COM attachement matter related to main body COM.