Airplane control

[Mastermemorex]

I love the rocket physics, but airplane are still unrealistic. Real airplanes should recover the roll and pitching moments, with a decent design of the center of mass and lift, and use the engine thrust to increase or reduce the altitude. Right now it is more like an arcade game to struggle with the controls while the airplane spins uncontrollable.

Edited April 10, 2016 by Mastermemorex

[Rocket In My Pocket]

Are you sure you are building your airplanes properly?

When built correctly they fly beautifully. (I fly in real life, there isn't anything glaringly wrong with KSP's flight simulation.)

Did you isolate the control surfaces? Ie. Aileron's control only roll, Tail rudder control's only yaw, Elevator's control only pitch? By default all control surfaces contribute authority to all 3 axis of movement. Make sure you use caps lock to place the game in "soft input" mode as well. Best of luck!

[Nich]

9 hours ago, Rocket In My Pocket said:

Are you sure you are building your airplanes properly?

When built correctly they fly beautifully. (I fly in real life, there isn't anything glaringly wrong with KSP's flight simulation.)

Did you isolate the control surfaces? Ie. Aileron's control only roll, Tail rudder control's only yaw, Elevator's control only pitch? By default all control surfaces contribute authority to all 3 axis of movement. Make sure you use caps lock to place the game in "soft input" mode as well. Best of luck!

Are you using FAR?  Planes in stock are not even close to real.

 

OP dihedral and COL COM both work for roll stability.  Swept wings do not help with roll stability.  Negative tail fins create pitch stability my guess is you don't know how to make a decent plane.

 

Here is my Eve plane that flys really well without SAS and minimal trimming.  It could use a little dihedral and removal of the drill as it causes COM to be off center.  I compensated with a little built in trimming.  If you have ever flown a real plane there is a good deal of trimming to do depending if 1 or 2 people are flying you will have to adjust roll and yaw and every time you change speeds you have to readjust the pitch trim.  Luckily all of your fuel is at COM in real life so COM doesn't move as you burn fuel.

https://drive.google.com/file/d/0B0iPRE-stt2CcWJZUXd2S1A3T00/view?usp=sharing

[Pds314]

Planes don't use increasing thrust to pitch up. They use elevators. Well-designed stock planes do indeed fly well. In fact, they fly unrealistically well because stock assumes that even inefficient stubby delta wings get the same lift coefficient as an ideal wing of infinite span at subsonic speeds. I.E. 0.109 CoL/degreeAoA in all cases.

Edited April 10, 2016 by Pds314

[Pds314]

15 hours ago, Mastermemorex said:

I love the rocket physics, but airplane are still unrealistic. Real airplanes should recover the roll and pitching moments, with a decent design of the center of mass and lift, and use the engine thrust to increase or reduce the altitude. Right now it is more like an arcade game to struggle with the controls while the airplane spins uncontrollable.

Center of pressure, mass, thrust are not the only things to consider. For example, an aircraft with a strong dihedral will be stable in roll even if the wings are entirely below CoM.

Also, increasing thrust to increase altitude doesn't work in all aircraft.

It might help if you would post some pictures of your airplanes using Imgur or another similar service, so we can understand what you think is wrong with the aerodynamics.

Also, you may be interested to know that you can adjust trim and make airplanes behave... roughly... how you described, using alt+QWEASD.

[Nich]

1 hour ago, Pds314 said:

Planes don't use increasing thrust to pitch up. They use elevators. Well-designed stock planes do indeed fly well. In fact, they fly unrealistically well because stock assumes that even inefficient stubby delta wings get the same lift coefficient as an ideal wing of infinite span at subsonic speeds. I.E. 0.109 CoL/degreeAoA in all cases.

Actually you do use thrust to control your descent rate while landing and planes that are flying faster then they are supposed to have a tendency to climb and require trim to keep them from pulling up.  You need to use a different trim settings for pattern work, climbing, cruising and descent.  I have only flown a cherokee and a bonanza and neither plane recovered like your expecting.  If you were diving they would self correct but it was VERY VERY slow about 5 degrees a minute.  Roll was actually very neutral maybe 1 degree every 5 minutes and just shifting your body left or right could cause the plane to bank assuming your perfectly trimmed. 

 

The reason you use thrust to control climb is because once you pass L/D max pulling back on the stick actually makes you come up shorter and makes your decent rate increase (other then flaring)

[Pds314]

1 minute ago, Nich said:

Actually you do use thrust to control your descent rate while landing and planes that are flying faster then they are supposed to have a tendency to climb and require trim to keep them from pulling up.  You need to use a different trim settings for pattern work, climbing, cruising and descent.  I have only flown a cherokee and a bonanza and neither plane recovered like your expecting.  If you were diving they would self correct but it was VERY VERY slow about 5 degrees a minute.  Roll was actually very neutral maybe 1 degree every 5 minutes and just shifting your body left or right could cause the plane to bank assuming your perfectly trimmed. 

 

The reason you use thrust to control climb is because once you pass L/D max pulling back on the stick actually makes you come up shorter and makes your decent rate increase (other then flaring)

Well yes, there are of course situations in which thrust is used to reduce descent rate. I HIGHLY doubt the plane's the OP is having trouble with are Bonanzas or similar. They probably have extremely high TWR, engines inline with CoM, jet engines, and a relatively high wing loading. More like a fighter jet.

Also, keep in mind that Cherokees and Bonanzas have only mild dihedral and no or minimal tail dihedral. A properly-designed plane can indeed require no corrections, but it's so slow at correcting itself as to be useless at it. You absolutely need the controls. Thrust-only flying is considered extremely difficult, with few, if any, landings ever occuring.

[Rocket In My Pocket]

11 hours ago, Nich said:

Are you using FAR?  Planes in stock are not even close to real.

 

21 hours ago, Rocket In My Pocket said:

there isn't anything glaringly wrong with KSP's flight simulation.

I said it was close enough for government work, not that it was perfect.

If you build a plane that looks like a real plane in KSP, it should fly fine.

The OP clearly isn't suffering from any of the edge cases your talking about, just basic design principles.

[Nich]

Well I guess to be fair I have never played a flight sim that didn't feel video gamey except for the trainers at Northwest airlines but these were millions of dollars.  Landed a 747 with a 45 knot crosswind   Even the Sim we had at Gulfstream was not very good as I landed a G650 equivalent on a 2500 ft runway.  Usually you want 5000 or more.

Edited April 11, 2016 by Nich

[KerbonautInTraining]

KSP simply isn't precise enough for the type of flying you're talking about.

I know what you mean; flying with power, trim, and two fingers on the yoke.

But in KSP things happen much quicker than in real life (launch, rendezvous, docking, etc.) and flying around is no exception. The way planes in KSP are built and flown is very similar to RC planes in real life. 

Edited April 11, 2016 by KerbonautInTraining

[Kerbart]

I think it's all about managing expectations. Does it mimic reality? Assuming your reality is literally based on a world 600km in diameter and populated (or so they say) with bug-eyed little green men, then yes. If you expect a fairly realistic rendition of reality on earth, no. But then there's x-flight and Prepar3D for that (or better for those that have access to professional simulator systems).

The fact that you can slap together something in 20 minutes that looks like an airplane and kind of behaves like an airplane is very impressive to me.

On 4/10/2016 at 8:57 AM, Rocket In My Pocket said:

Did you isolate the control surfaces? Ie. Aileron's control only roll, Tail rudder control's only yaw, Elevator's control only pitch? By default all control surfaces contribute authority to all 3 axis of movement. Make sure you use caps lock to place the game in "soft input" mode as well. Best of luck!

That, more than anything else, makes the biggest difference.