Some questions about planes.

[Chrysoprace]

Hello! Latecomer to KSP here, didn't have a computer worth mentioning for the last several years, but more or less everybody I asked for game recommendations mentioned KSP and I am loving it. 

Anyway, I just started trying to mess about with some planes and have a few questions I can't figure out. 

This is the first plane I made - I just followed a YouTube tutorial because I had no idea at all where to even begin.  https://www.youtube.com/watch?v=xHMIapk41iA

This is the second successful plane I designed. Needed something with more range and a booster to get up above 18km until I get better engines later. 

Which brings me to my questions. 

1) The first plane is very unstable. It is flyable, but twitches around and wants to nose down a bit at all times (I suspect this may be due to the air scoop on the bottom hidden in the picture). If I turn on SAS it freaks out. The second plane is steady as a rock, I can set a heading, turn on SAS, turn up the time to x4, and more or less stop paying attention until I am approaching my destination. Anybody know why?

2) The first plane wants to drift to the side on the runway, even when stopped with no throttle, even with brakes applied. A couple of other designs I tried were even worse about this. I got into the habit of throttling up immediately before it could get out of hand. This second plane does not, although it will drift forward slowly without brakes. Why?

3) When I made the first plane and took it out to mess about with it I discovered that the controls were backward to what I was used to. I use an old PS3 controller with a dead battery to play KSP. There was a series of failures to take off as I exploded into the runway because pulling back on the joystick was making me nose down (took a few tries to figure out that this was the problem and not something with the design). So I inverted the joystick in the options. Problem solved, the plane now flies more or less as I am used to from other simulators. Star Fox and X-Wing are simulators, right?  Then when I took the second plane out I was having the exact same problem. The stick is still inverted, but this plane behaves in reverse of the first. What gives?

Thanks in advance!

[chrisb2244]

1) I'm not certain, but I'd say you have not enough pitch authority for the first plane. There's no tail, and so only the ailerons on the wings give you pitch control, and they're not far in front of/behind (I.e. front or back away from) the CoM. 

Although the second plane has a similar problem, it looks like you have the Swivel engine there, so you can use the gimballing of the engine to add pitch control.

2) Might be your wheels are not completely square to the ground. Try using the rotate and offset gizmos in "absolute" mode to correct. Having a plane too heavy for the wheels also does this, but your planes seem ok.

3) No idea, sorry. 

[bewing]

Generally, you want the little blue CoL ball indicator to be a bit behind the yellow CoM indicator ball. So that they are not quite touching. When they touch, your plane will have pitch problems. When you have them right on top of each other, generally you will have severe pitch problems and your plane will be unflyable, as you found. Why your second plane works is a mystery.

Beyond that, control surfaces attached to swept wings can be confusing ... both to the program and to the player. The general point is that a control surface that is in front of the CoM has to operate inverted to one that is behind the CoM. (Think about how a canard has to operate compared to a rear stabilizer.) So the program tries to guess whether each control surface is ahead or behind the CoM, and inverts their operation if they are in front. Which brings you to swept wings -- the program basically can't figure these out automatically. So it always gets them wrong. If the program gets it wrong, then you can use the control surface's context menu, and set the "authority" to -100 to partially fix the problem. Or just mentally use the controls inverted, as you have done.

However: those Type A swept wings (on the first plane) are a very bad deal. They are expensive, extra heavy, and very draggy, and give you very little lift in return -- compared to all the other wings in the game. Also, the game does not actually model the wing sweep in the aero portion of the game; swept wings have no drag bonus. So if you really want to use a swept wing for aesthetic purposes, go ahead. But if you are having trouble making your plane fly, then switching to straight wings might be a smart choice. You will always know if your control surface is ahead of or behind your CoM.

 

[Rath]

Nope, you want the col just inside the com usually.  The advice above is true for before the old aero, and usually sorta works.

 

1: Yep, not enough pitch authority.  Just take a pair of the elevons, turn off roll and shove them on the back.

2:It looks like you might have a quadricycle setup on the first plane.  Don't do that.  Change your plane to have a single gear in the front and two on the back.

 

The control inversion may be because of a bug in the way that the game recognizes the way that control surfaces should act.  if the root part is in front of the COM, but the pitch control is behind, it will think that its in front, not behind.  What you could do is add a tail and disable pitch on those control surfaces.

[Magzimum]

For me the really basics are:

1. Blue bal (CoL) behind yellow ball (CoM).
2. At least 1 horizontal control surface in front of yellow (and blue), and one behind yellow (and blue).
3. Tail fin on the back.

My favorite designs (because they are stable) usually have large wings in the back (helps to push that blue ball back), and winglets or canards near the front. Because the canards are often quite far from the yellow/blue balls, they give excellent control. And it's not science fiction either. The Typhoon uses it for example:

 

[Chrysoprace]

Thanks for the input everybody. Very helpful. 

I will mess around with the designs a bit and see if I can figure out why the second plane is so incredibly stable when it apparently should not be. Perhaps the gimbaling on the rocket engine is working as a tail?

[Plusck]

For the first plane, the angle of the elevons (their axis of movement passes through the centre of the plane in front of the CoM) is probably causing problems. If they were rotated to be perpendicular to the direction of travel, they'd look weird but they'd probably work a lot better.

I think this is why the second plane works fine. Even though the elevons are right at the edges of the wings, they're far enough behind CoM (or at least, their axis of rotation intersects the centre of the plane far enough behind CoM) to work well.

For question 2), I'm pretty sure it's because [edit - misread your question - thought you were talking about drifting strongly to the side when rolling along for takeoff. Chakat Firepaw's answer is more like it. Still, you are probably also experiencing unstable rolling because... ] most of the mass of the plane is sitting above that single, steering wheel at the front. If you moved all of the wheels forward a fair bit (in any event, your rear wheels are too far back on both planes, they should be closer to the CoM to allow the plane to rotate on takeoff) it should roll better.

3) is the result of the first paragraph above. Your controls were fine to start with - you shouldn't have inverted the axis - because in fact the game is interpreting your first design the wrong way round since the axis of the elevons is (as far as KSP is concerned) the opposite to what you think it is supposed to be. Again, rotate the elevons so that a line along their length intersects the plane behind CoM, and all should be well.

Edited November 4, 2016 by Plusck

[Chakat Firepaw]

To answer the 2nd part of #2:  Planes tend to roll to the middle of the runway because the runway is flat, and thus lower at the middle.  This seemingly strange statement comes about because Kerbin is round, and thus the ends of a flat surface tangent to it are farther from the centre of the planet.