I have a(nother) aerodynamics problem...

[SlabGizor117]

I built another plane from the last time I asked about aerodynamics, and this is having another bad loss of control.  The aero overlay (F12) didn't show anything, but here's some Col/CoM pics:

I also just confirmed that it's an aerodynamic problem, instead of overcompensation by the engine gimbal, by disabling the vectoring.  No effect.  

Edited February 2, 2016 by SlabGizor117

[Temstar]

Unstable in which axis? Roll, pitch, yaw or some combination?

Anyway in the first picture, CoM is too close to CoL, move it forward or move CoL back by moving the main wings back. Also, V-tails are always problematic, especially V-tails made from little wing and control surfaces instead of an all moving tail. You're asking two tails (or even worse, two small control surfaces on two tails) to do the job of three with a rather unsophisticated fly-by-wire control that is KSP SAS, so avoid v-tail at all cost.

Lastly are your horizontal stabilizers made from four control surfaces, two of which is back to front relative to their normal orientation? That's just asking for trouble. Never do inverted control surfaces unless you really know what you're doing.

[MoonMan22]

Just like Temstar said,

 

Move CoM forward, and remove the wing parts u have running vertically on the side.

 

good luck!

[SlabGizor117]

Ok, I'll try.  The bigger control surfaces towards the front of the elevators are disabled, though.  Just an aesthetic alternative to other elevator options.

EDIT: unstable in pitch, by the way.  As soon as I pitch up enough to get in the air, it flips back to retrograde.

EDIT 2:  Also, are you saying that a V-tail is trying to handle pitch, yaw, and roll when it shouldn't have to? Because I have all control surfaces only enabled for their purpose(ailerons only roll, rudders only yaw, elevators only pitch).

EDIT 3....   Ok so after a little troubleshooting, it seems that the plane isn't so much easy to lose control of, as it is always wanting to pull up.  I'm changing one questionable thing at a time for process of elimination, I'll see how it goes.

EDIT 4:  Wow, lots of edits... I played around a bit with the CoL/CoM, and it really helped... I couldn't believe that it was that uncontrollable even when the CoL was still behind the CoM..  I'll keep playing around with it and get back to you guys.

Edited February 2, 2016 by SlabGizor117

[RizzoTheRat]

Have you got any pitch on the wing and tailplane to start with?  If you angle the wings up wards a few degrees it will help keep the fuselage's angle of attack down, which reduces drag on the front end which can cause a flip as the angle of attack increases.

In real life you put the centre of lift of the wing behind the centre of mass, and angle the tailplane downwards so it's balancing the lift by pulling the tail down,  That way when you increase pitch the wing gains lift but the tail loses downforce so you return to a stable position.  If your design has both wing and tailplane generating lift it's generally less stable.

[Plusck]

Try rotating the whole thing (root part) slightly to give the craft a 10-15° angle of attack. What do the CoL/CoM markers show?

I'm trying to figure out how you built that plane, and why the CoL marker appears where it is. I'm having a hard time. There seems to be a massive amount of clipping and plenty of invisible parts (like air intakes).

 

However, it is certainly that clipped-in elevon that is causing the problem. By rotating its AoA like that, you are making the SPH give a completely false interpretation of where the CoL actually is.

In fact, the 'true' CoL of your plane is certainly somewhere towards the rear of your wing. It is also almost certainly in front of your CoM.

When you try to take off, the clipped-in elevon will give significant lift and push the nose of your plane down, while the wing has a very low AoA and very little lift. As soon as you try bringing up the nose, however, the rear control surfaces will cancel out lift at the rear, and your wing suddenly starts producing massive amounts of lift, far ahead of your CoM. Therefore it is completely unstable and flips.

[SlabGizor117]

Hmm.. So I still have more CoL/CoM testing to do, but it seems to be a steady automatic pitch-up(as said before) instead of a loss of control.  As soon as i turn off SAS, I lose control.  Should I just quick and dirty fix it with a CoL moved back more?

[GoSlash27]

SlabGizor,

 You have a lot of fuselage ahead of the CoM and CoL. This is basically a lot of lightweight structure that generates huge amounts of drag when not aligned with the airflow. See @Val's tutorial on the subject here:

You have a center of pressure (where the drag acts) ahead of your center of mass. That's why it's flipping.

The quick and dirty remedy for this is to have more tailplane area in the back.

 The proper remedy is to have the mass distributed more evenly along the aircraft's length instead of concentrated in the tail.

Best,
-Slashy

 

[Snark]

You really need to move the CoM forward.   It's back at the back of the plane.  Anything with a CoM that far back is going to just kill your stability-- no matter what games you play with the CoL, you've got that big fuselage sticking out in front and it's gonna just kill you.  And body lift/drag from the fuselage won't show up on your CoL indicator.

The other reason why a CoM way at the back kills you is that it's too close to all your control surfaces.  A control surface's effectiveness depends on having a good long lever arm to work with, which means being far from the CoM.  Look at where your CoM is, and then look at all your control surfaces for pitch.  They're really close together, which means very little lever arm to work with and those control surfaces can't do much.

So one thing you could do, for example, if you have trouble moving your CoM up to the front of the plane:  you could stick some canards on the front.  This is kinda counterintuitive (it'll move your CoL forward, for example) ... but because the canards are so far from your CoM, they'll have a lot of control authority, and the canard deflection may be enough to restore stability.  It's worth a try, at any rate.

But the #1 culprit here is CoM.  Move it way farther forward from where it is now.