Mirrored control surfaces

[martinezfg11]

Hello,

Currently any control surface with "mirrorRefAxis" will cause the control surface to visually actuate in the opposite direction. Would it be difficult to make a plugin that fixes the issues caused by mirrored control surfaces? Would anyone be willing to tackle this?

Thanks!

[Boris-Barboris]

You mean, in editor? And which csurface are you describing, stock or AA's?

[martinezfg11]

1 minute ago, Boris-Barboris said:

You mean, in editor? And which csurface are you describing, stock or AA's?

I have a custom tail fin in the AoA Tech mod that is asymmetrical in nature. So I use mirrorRefAxis to make sure the orientation is proper. However, in the flight scene, when pitching up or down, the original part actuates correctly, but the movement of the mirrored control surface is also mirrored and therefore opposite of the original part. So pitching up will cause the two tailfins to move opposite of each other as opposed to moving in unison. It's purely visual as aerodynamically, they provide the proper lift, but it is annoying to see.

 

I'll post a video of it later.

[martinezfg11]

Here's a Video of the bug. Hopefully someone can help resolve this issue?

[Boris-Barboris]

@martinezfg11 What is mirrorRefAxis and how does it work, and what does this fin's config looks like? Do you introduce some cheeky unpopular transform parameters, or maybe craft file was edited manually?

[martinezfg11]

The craft file was not edited manually, it's been a long standing bug that used to even affect the B9 procedural wings, because it used to use mirrorRefAxis, I don't think they use it anymore because of this same issue. mirrorRefAxis can be found in the Squad Fixed Landing Gear, all it does it mirror the geometry on the defined axis, therefore meshes can be asymmetrical, yet properly mirrored when activating symmetry. 

See below for the wing configs

PART
{
	name = AoAero.TESTFin
	module = Part
	author = Martinezfg11
	rescaleFactor = 1
	mirrorRefAxis = 0, 0, -1
	node_attach = 0.5, 0.0, 0.0, 1.0, 0.0, 0.0, 1
	TechRequired = aerodynamicSystems

	entryCost = 3800
	cost = 600
	category = Aero
	subcategory = 0
	title = TESTfin 001
	manufacturer = AoA Technologies
	description = Control surface for testing the new KSP Symmetry.
	attachRules = 0,1,0,1,0
	mass = 0.125
	thermalMassModifier = 8.0
	// heatConductivity = 0.06 // half default
	emissiveConstant = 0.95
	dragModelType = none
	maximum_drag = 0.02
	minimum_drag = 0.02
	angularDrag = 2
	crashTolerance = 12
	maxTemp = 2400
	explosionPotential = 0.1
	fuelCrossFeed = True
	bulkheadProfiles = srf
	MODEL
	{
		model = AoATech/Parts/Aero/TestFin/model
	}
	MODULE
	{
		name = ModuleControlSurface
		useInternalDragModel = True
		dragCoeff = 0.6
		deflectionLiftCoeff = 0.61		// 2.18m^2
		ctrlSurfaceRange = 25
		ctrlSurfaceArea = 1
		actuatorSpeed = 20
		transformName = ctrlSurf
	}
}

@PART[AoAero.TESTFin]:FOR[FerramAerospaceResearch]
	{
		@module = Part
		@maximum_drag = 0
		@minimum_drag = 0
		@angularDrag = 0
		@dragCoeff = 0
		@deflectionLiftCoeff = 0
			!MODULE[ModuleLiftingSurface] {}
		MODULE
		{
			name = FARControllableSurface
			MAC = 0.97528
			MidChordSweep = 28.658
			b_2 = 1.80893
			TaperRatio = 0.38936
			maxdeflect = 20
					transformName = ctrlSurf
		}
	}

 

[martinezfg11]

So has anyone tried fixing this at some point? I think it was working in pre 1.0 I just don't know what change may have broken this. 

[Boris-Barboris]

I gave it a quick look, but didn't write anything. It looks like the way mirrorRefAxis is implemented is to set scale for some part to negative value for one of the axis, so control surface module will need to check for this fact and reverse visual rotation in this special case, but that's just a thought.