maneuverability, stabilization

[Mr. "I need help guy"]

what are some tips on keeping a space plane stable and easy to maneuver?

[HebaruSan]

In atmosphere, keep the center of lift just behind the center of mass, and put control surfaces far from the center of mass, in a direction reflecting their function. For pitch or yaw, they should be at the front or back. For roll, they should be to the left and right of the center.

In space, have reaction wheels and a power source.

[Starman4308]

Those are actually diametrically opposed goals.

By placing center of drag behind center of mass, your vehicle becomes dynamically stable: drag effectively "pulls" your CoD behind your CoM, so any deviation from going straight forwards tends to get damped out.

By placing it in front of center of mass, it becomes instable: drag is effectively pushing back on your CoM, and it becomes difficult to manage as aerodynamics wants to pull your ship off the forward vector.

If you want it to stay maneuverable, I'd put CoD a short distance behind CoM (stuff can change in flight; I lost one large spaceplane because I didn't recognize that, with fuel tanks drained, CoM went too far back).

[Snark]

8 hours ago, Starman4308 said:

Those are actually diametrically opposed goals.

Depends what one means by "maneuverable".  If you mean "can spin on a dime like a fighter jet", then yeah, excessive stability could get in the way of that.

But I suspect what the OP meant was basically "easy to steer without completely losing control, and easy to recover if things don't go well, and is simple to fly without it yawing every time you try to roll and vice versa".  So in that case it's pretty synonymous with stability.

To the OP:  You want your CoM to be in the front of the plane, not in the back.  You want to have pitch/yaw surfaces as far behind the CoM as you can manage.  It's okay to have some pitch surfaces up in the front (e.g. canards), but depending on how you fly, they have the potential to be less "safe" than surfaces that are behind the CoM.  For your yaw surfaces (e.g. tail fin), make sure you've disabled the actuators for everything except yaw-- i.e. you don't want your tail fin to be trying to help you roll the plane, because then hilarity will ensue.

You can get a bit of roll stability (i.e. a plane that "wants" to fly right-side-up) by mounting your wings above the centerline of the fuselage (i.e. so that your CoM is below the plane of the wings), and also by giving the wings a slight upward dihedral (i.e. the wings are angled slightly above the perfectly horizontal).

 

[Vanamonde]

This question about how to play the game has been moved to Gameplay Questions. 

[foamyesque]

Something that can be very poisonous in KSP, I've found, but that doesn't tend to be discussed often, is inertial coupling; masses above or below the CoM induce an instability into the system, and need wingspan in order to compensate. This is a problem real-life aircraft designers only started running into when they were trying to make high speed, very streamlined aircraft... and guess what KSP's spaceplanes want to be. The upshot is that through conservation of angular momentum, attempting a roll will produce a massive yaw moment, sufficient to overpower even very large tails -- large tails can in fact make the problem worse, by further unbalancing the plane.

Edited January 29, 2018 by foamyesque

[Guest]

Stable: shape it like a lawn dart. Pointy, heavy stuff in the front, wings in the back.

Manoeuvrable: shape it like the Red Baron's Fokker: lots of wings near the centre, big tail surfaces. Just make sure the centre of mass is still just ahead of the centre of lift.

Note that the CoM shifts as you burn fuel. This is a problem especially with bigger planes. You can pump any fuel that you have left back and forth or control how it shifts with fuel flow priority (hint: lowest priority for forward tanks). However, it's important to check both the empty and the full CoM or you risk a nasty surprise on your return trip.

Hint: place everything you can with "snap to node" (alt key down on Windows). That way you'll avoid surprising aero problems when you thought you snapped something onto a node but accidentally only surface attached it.

[foamyesque]

CoM shift, incidentally, can be majorly reduced -- or even eliminated entirely -- if you can arrange matters such that you have an equal amount of variable-mass torque, relative to the CoM, on either side of your CoM. In practice the simplest way to achieve this is to put equal amounts of fuel equal distances in front of and behind it, with fixed masses (engines, control pods, gear and so on) arranged so that they keep the CoM in the middle of the fuel tankage. This, except on the very smallest crafts where a pod is the payload and can counterbalance a rear-mount engine, usually means you want your engines as close to the center of the plane as possible, as they are by far the heaviest and densest pieces of kit on most planes, and rapidly overwhelm any nose counterbalancing unless the nose is on a very long lever arm relative to the engines.

Edited January 29, 2018 by foamyesque

[Guest]

1 hour ago, foamyesque said:

CoM shift, incidentally, can be majorly reduced -- or even eliminated entirely -- if you can arrange matters such that you have an equal amount of variable-mass torque, relative to the CoM, on either side of your CoM. In practice the simplest way to achieve this is to put equal amounts of fuel equal distances in front of and behind it, with fixed masses (engines, control pods, gear and so on) arranged so that they keep the CoM in the middle of the fuel tankage. This, except on the very smallest crafts where a pod is the payload and can counterbalance a rear-mountengine, usually means you want your engines as close to the center of the plane as possible, as they are by far the heaviest and densest pieces of kit on most planes, and rapidly overwhelm any nose counterbalancing unless the nose is on a very long lever arm relative to the engines.

Yep.

In practice, bigger planes work better if the engines are on relatively short pods on the wings. This way, you can move them forward and back until you get the CoM staying more or less stable.

I've found that as I scale them up, my designs evolve naturally from payload in the nose/engine(s) in the back/wing in the middle, to engines and (delta) wing in the back with a long neck and then the payload, back to a similar pattern as where I started except now the engines are in pods on the wing.

[RizzoTheRat]

Re. the comments about CoM in front of CoL, an aspect of real world aircraft that's often overlooked on here is that in a conventional design the horizontal tailplane produces downforce not lift.  If the aircrafts pitch angle increases the wing generates more lift, but the tail produces less downforce so the aircraft is pitched back down again. Front mounted Canards tend to reduce stability as they're producing lift so are a lot trickier to design.

[Guest]

Listen to @RizzoTheRat, he knows what he's talking about.

As to canards, I found them highly troublesome to tune before I started using Atmosphere Autopilot, which uses them intelligently instead of just applying control inputs directly.

Nowadays I use them in "medium-weight" designs -- SSTOs which lift maybe 20-40 tons or so, and still have the engines mostly toward the back. These usually experience a certain amount of CoM shift, which means it's too far forward on launch. Canards help them unstick from the runway. Without AA you do have to be careful both with tuning and with flying, as there is a real risk of doing an unintended backflip immediately after take-off. But tails are definitely easier. I have sometimes angled them down a little for more stability.

[GoSlash27]

My advice:

- You don't really want maneuverability in a space plane. You want it to be extremely stable, like a sled.
- Don't make it look "fast" by clustering the engines at the back, and using canard delta. Make it look boring by keeping the CoM and CoL near the longitudinal center and a conventional tail.
- It's important to keep the pressure center behind the CoM as well as the CoL. Otherwise you wind up with a plane that wants to fly backwards at high speed. The pressure center isn't shown, so you have to guesstimate it.
- Turn down the deflection angles on your control surfaces so they don't oscillate.
- Make sure your control surfaces only respond to their intended inputs. Ailerons only respond to roll, elevators only respond to pitch. I personally leave my rudders turned off, but if you use them, they should respond only to yaw.
Best,
-Slashy

[foamyesque]

I actually find myself often turning my control deflections up, for purposes of keeping my nose where I want it during re-entry. In general, spaceplanes will try to bring their nose prograde, which is a great way to wind up overheating your craft coming back from space; I prefer to fly with a 45 degree or better angle of attack. There are several ways to do this, and doing it purely aerodynamically is somewhat challenging. Higher control deflections (especially on control surfaces that are not all-moving), use of all-moving control surfaces for larger CoP shifts for the same control deflections, a flap system to move the CoP forward, etc all help. I tend to prefer canards over conventional tails, since they put more mass towards the nose of the plane, usually require less control surface for the same control authority, are less vulnerable to colliding with terrain, are more likely to be lined up vertically with the CoM (reducing inertial coupling), and have no chance of interfering with or being destroyed by engine exhaust.

And while lots of people will say a fixed vertical stabilizer is enough, in my experience that's only really true of small craft, where pod torque is enough. Induced yaw is murder on your flight dynamics and a rudder is the most powerful tool you've got to control it. If you need ailerons for your roll control, you should be putting a moving rudder on. You don't necessarily need a large rudder -- a number of my small-to-mid weight planes use a single small elevon part, with no fixed component at all, as their rudder -- but it's a good idea to have something.

[Zeiss Ikon]

11 hours ago, RizzoTheRat said:

Re. the comments about CoM in front of CoL, an aspect of real world aircraft that's often overlooked on here is that in a conventional design the horizontal tailplane produces downforce not lift.  If the aircrafts pitch angle increases the wing generates more lift, but the tail produces less downforce so the aircraft is pitched back down again. Front mounted Canards tend to reduce stability as they're producing lift so are a lot trickier to design.

This is a common misconception.

If you look at the designs for a lot of free-flight model aircraft, they're set up so the tailplane produces lift, and all canards are (it's common for a canard craft to have the COM foward of the leading edge of the wing, where with conventional layouts it's usually 30-50% of the way from the leading to trailing edge, not counting sweep).  The critical point, for both canards and conventional layouts, is that the forward wing must fly at a higher coefficient of lift than the rear.  This produces a tendency to pitch up as speed increases, or pitch down as speed decreases; this gives what's known as longitudinal stability.

Now, the easiest way to do this is with decalage -- a little difference in incidence between the front and rear wings (whether that's canard pitched up relative to the wing, or tailplane pitched down).  However, with high performance free flight (which spend most of their time aloft gliding, so L/D is very important, as well as minimum sink) it has often been the case that different airfoils would be used, so that with zero decalage the tailplane would develop less lift, relative to speed, than the wing -- not no lift, certainly not negative lift (because that requires the wing to generate more lift, which increases drag), just less lift.  This way, the nose will rise and fall as needed with small changes in speed, but the tailplane is helping hold the tail up, rather than pushing it down.
 

If you're mainly flying straight and level, a canard is actually easier to make stable than a conventional -- put your COM just ahead of the leading edge of the wing, then adjust the canard incidence to get a stable glide and use pitch trim to set level flight speed.  Unlike a conventional, the COM location isn't very critical at all, and the canard will always be in a lifting mode (unlike conventional with decalage).  For high-G maneuvering, a three-surface is a very good choice, if the surfaces are sized correctly relative to each other -- the COM can be exactly on the COL, and with both canard and tailplane contributing, the craft can maneuver like, well, a jet fighter.  Stability, however, is likely to depend on a computer doing the actual flying, while the pilot just tells the computer where to go.

[RizzoTheRat]

7 hours ago, Zeiss Ikon said:

However, with high performance free flight (which spend most of their time aloft gliding, so L/D is very important, as well as minimum sink) it has often been the case that different airfoils would be used, so that with zero decalage the tailplane would develop less lift, relative to speed, than the wing -- not no lift, certainly not negative lift (because that requires the wing to generate more lift, which increases drag), just less lift.  This way, the nose will rise and fall as needed with small changes in speed, but the tailplane is helping hold the tail up, rather than pushing it down.

Interesting, I'd have thought models would need more passive stability than manned aircraft as it's harder to tell what they're doing. 

More advanced CFD models and active stabilisation have probably changed things a bit and allowed the margins to be reduced a lot, but in order to have stable flight the CoL and CoM must be in the same position horizontally (can be separated vertically obviously), otherwise the resultant torque will pitch the aircraft up or down.  Passive/static stability means having the tailplane set up such that an increase in angle of attack causes the tail to pitch the nose back down, and a decrease causes it to pitch it back up.  The usually way to do this is to have the tailplane set to provide downforce , which counters the CoM being ahead of the wings CoL.  Increasing angle of attack decreases the downforce and forces the nose down, decreasing AoA increases the downforce and pushes it back up.  As an added bonus it helps ensure the wing stalls before the tail.

You can achieve the same effect by having the tail set at a lower AoA than the wing but still producing lift, but there's less margin for error, I'm guessing low speeds and fixed weights in models means this is doable.  In order to do that of course means the CoM is behind the wings CoL.

All this is then complicated by moving the CoM around which is not just fuel, but different weight passengers too. Less of an issue with light aircraft, which have a nice heavy engine up front and just set a weight limit for the pilot/passengers, but gliders often carry ballast as it is possible to be under weight, eg with a single pilot in a 2 seater.  Presumably you need the ballast models pretty carefully, but the CoM then isn't going to shift.

As a further complication the CoL moves with speed and angle of attack.  Not much of an issue for a high aspect ratio low speed aircraft like a glider, but becomes a big issue when you go supersonic.  Concorde pumped it's fuel around to move it's CoM as it's CoL moved back something like 6 feet at cruise from it's position at takeoff speed 

 

This explains the torque bit reasonable well, the diagram's a bit shonky though

https://www.grc.nasa.gov/www/k-12/airplane/trim.html

 

Of course a lot of military aircraft are designed to be inherently unstable as it's lower drag and higher manoeuvrability, but I wouldn't fancy flying on of them on the WASD keys

[foamyesque]

6 minutes ago, RizzoTheRat said:

Of course a lot of military aircraft are designed to be inherently unstable as it's lower drag and higher manoeuvrability, but I wouldn't fancy flying on of them on the WASD keys

I have a dynamically unstable multi-hundred ton plane. Flying it on the keyboard is... an experience.

[Zeiss Ikon]

On 1/30/2018 at 3:41 AM, RizzoTheRat said:

Interesting, I'd have thought models would need more passive stability than manned aircraft as it's harder to tell what they're doing.

Free flight models need to be stable "hands off" at all times -- yet those are the ones most likely to have been originally designed with lifting tail surfaces.  Yes, they can be tricky to trim -- COM location is critical -- but it's done for performance reasons.

Those performance considerations generally don't apply to our spaceplane designs (or even our atmospheric-only aircraft); we're better off with longitudinal stability, with a pitch-up moment from the surfaces offsetting a pitch-down moment from the COM (slightly) ahead of the static COL.