trouble with atmospheric plane

[KerbMav]

Made my first real attempt at a plane - pure atmospheric for starters.

Why does every (as far as I tested other players crafts and now mine too) plane need an angle of 20 degrees to hold its height?

Javascript is disabled. View full album

Also I could pull my own hair that when I pull up 5 degrees, the plane's nose then drops 10-15 degrees to where it had been before - it is the same for pushing down to almost the same extend.

Another funny thing is the physics of the game - although I dont know if it is the same in reality? - that make me fly east for a few degrees while pointing straight north.

[NeMeSiS]

Atmospheric physics are far from finished.

Anyway, your problems mostly look like a lack of lift.

[espm400]

Made my first real attempt at a plane - pure atmospheric for starters.

Why does every (as far as I tested other players crafts and now mine too) plane need an angle of 20 degrees to hold its height?

~snip~

Also I could pull my own hair that when I pull up 5 degrees, the plane's nose then drops 10-15 degrees to where it had been before - it is the same for pushing down to almost the same extend.

Another funny thing is the physics of the game - although I dont know if it is the same in reality? - that make me fly east for a few degrees while pointing straight north.

If you want to get into atmospheric flight I would strongly recommend installing Ferram Aerospace Research (FAR).

Basically it does a complete overhaul on the drag models in the game, allowing for proper atmospheric flight.

[Tarmenius]

Also I could pull my own hair that when I pull up 5 degrees, the plane's nose then drops 10-15 degrees to where it had been before - it is the same for pushing down to almost the same extend.

This is thanks to the new SAS behavior. I've found that adding an Inline Reaction Wheel to relatively small craft can help make more crisp pitch adjustments and a faster "settling" of the SAS/

[numerobis]

Air pressure at 15km on Kerbin is quite low -- roughly equivalent to air pressure at 25km on Earth. You will either need a huge amount of lift -- think U2 -- or be using principally thrust (or near-orbital speed) to maintain altitude.

Edited August 5, 2013 by numerobis

[GDJ]

Part of it could be the CG/CL (centre of gravity /centre of lift) and where they are in relation to each other, also centre of thrust also comes into play in atmospheric flight.

In your picture, the centre of thrust is bang on, but the centre of lift is a bit too far back. I have found that if the centre of lift is within the Centre of gravity sphere the plane does smarten up quite a bit in regards to more level flight, but at a risk of going all snakey when you hit stall aerodynamically. Either reposition the main wings or add some weight to the rear end.

Also air pressure and density does have a strong influence, and the only way to beat that in this game is raw power and speed. Get it going really fast.

Now, as far as mods go, the Ferram Aerospace parts are fantastic. I use them almost religiously. They work very well.

[numerobis]

Oh, another solution to the angle of attack problem: tilt the wings in the SPH with shift-QWEASD. I typically have to hit every single one to figure which one will in this case tilt the wing 5 degrees, so that when the cockpit is going straight, the wing is at a positive angle of attack. You have the same angle of attack for the wings as if you had your nose pointing up, but your intakes can point in the direction of your flight path. Lift increases up to an angle of attack of a bit over 30 degrees, then it declines. If you point the wings up two or three notches, you can fly with your nose close to level, and not be in danger of stalling when you pull up.

[arq]

I think he's referring to Angle of Attack, not pitch control. Wings in KSP use fully-symmetric airfoils. Thus, they generate no lift at 0deg. As AoA increases, they create more lift. If you had a higher lift/weight ratio, your planes would fly at a lower AoA. Alternatively, you can use shift-wasdqe to pitch your surfaces upward slightly (5-10deg) on your ship, thus inducing a nonzero AoA and producing lift even when your navball reads 0deg (but upside-down flight will require an even higher AoA).

CoL/CoM position only affect what your plane does with no control inputs. Control surfaces dramatically move your CoL, so as long as they aren't too far apart you can compensate for any discrepancy. What I find to be much more important is Center of Drag. If your CoD is in front of your CoM, your plane will attempt to flip around and fly backwards. The further back your CoD, the easier it will be to recover from spins and tumbles. This is especially an issue for spaceplanes, which experience massive drag forces on re-entry. You want as many of your lifting surfaces as possible to be behind the CoM, with just enough in front to keep your CoL close to the CoM.

[Tiron]

Because you don't have enough lift and are having to use part of your engine thrust to keep you from going upward. That said, with the current aerodynamic model getting 'enough lift' to fly straight is basically impossible. Because it doesn't take account of things like angle of attack.

That's why they're mentioning FAR, but keep in mind that FAR messes up all kinds of other things as well. In particular I've heard that it screws up mechjeb's ascent module something horrible.