Yaw stability and re-entry

[xrayfishx]

Like many players of KSP, I have trouble creating good spaceplanes. By now, I have managed to improve my building skills to the point where I can make very decent planes, but I'm still having problems with the "space" part, mainly the re-entry. All my spaceplanes and shuttles seem to fly very well in low atmosphere and they even survive suborbital tests and slow re-entries, but whenever I try to get down a bit faster so I won't have to wait half an orbit to land, the spacecraft tend to yaw to one side until they flip. The pitch and roll are stable, but yaw just hates me. Also, without RSS I got several stable spaceplanes so I fear the problem must be with hypersonic speeds.

Here's my latest shuttle, which is capable of gliding above 65k until the speed is down to around 4km/s and then land, however when I tried a bit more aggressive re-entry, it encountered yaw instability below 70k, around maybe 50 it flipped and broke.

Note that I'm using Realism Overhaul and to match that, all the B9 S2 parts have been scaled up by a factor of 1.6 and the masses and amounts of fuel have been increased accordingly (unless I'm an idiot and it should be something else instead of multiplying mass and fuel capacity by 1.6³).

Also, I couldn't find anything on that by searching the forums, but if anyone can point me to a good tutorial that would explain what I did wrong, I would really appreciate that.

Edited July 12, 2014 by xrayfishx

[GoSlash27]

I've been experimenting with high altitude/ speed ion gliders lately and have picked up a few things. It should translate to your problem...

Drag from wings is much higher than parasitic drag from parts at high velocities. They tend to act as parachutes. Your low center of lift is probably to blame. First thing to go is typically yaw...

If you can, try raising the wings and canards to the CoG height. See if that improves matters.

Good luck!

-Slashy

[LethalDose]

Well, at 65 km, you're not gliding... You're basically still just orbiting. sorry, missed the realism overhaul note.

Looking at your vessel, and reading your post, I have the following advice/questions/notes:

• Where is your CoL vs CoM when your tanks are emptied to the point where you're attempting reentry?
  
• In the picture, your CoM is way above your CoL. This leads to instability, and would explain your shuttle flipping over before breaking up. If you rotate the wing upward (a "dihedral wing"), it may help.
  
• Have you 'tweaked your control surfaces to only act on particular inputs? e.g. your Rudder should only respond to Yaw. On my spaceplanes, I usually have Rudders act on yaw only, Elevators (aft) act on pitch only and canards (fore) act on roll and pitch.
  
• What's your Angle of Attack (AoA) on re-entry?

I think thats it.

Edit: OH forgot, is there a bigger rudder you could use? If you look at the actual shuttle, it's rudder was huge. Also moving your wings higher up on the fuselage instead of flush with the bottom would lift the CoL too.

Edited July 12, 2014 by LethalDose

[Redwizard000]

You have bad center of mass and center of lift. This will cause your space plane to wildly spin out of control on reentry. This gets worse if your fuel tanks aren't properly balanced. Your CoM and CoL will change as your fuel tanks are emptied. I had this problem on my very first real attempt at a spaceplane. The plane flew just fine until it burned about half it's fuel and then it spun around like a top and tumbled through the air.

Check out this guide on the forums: http://forum.kerbalspaceprogram.com/threads/52080-Basic-Aircraft-Design-Explained-Simply-With-Pictures

Also check out the RCSBuildAid plugin. It has features that help with balancing on space planes. You can resize the blue and yellow spheres and it will show your dry mass. http://forum.kerbalspaceprogram.com/threads/35996-0-23-x-RCS-Build-Aid-v0-4-6-ARM-patch-fixes (not to mention it is just a handy tool to have in general...)

[xrayfishx]

Your low center of lift is probably to blame.

In the picture, your CoM is way above your CoL. This leads to instability, and would explain your shuttle flipping over before breaking up.

You have bad center of mass and center of lift.

That shuttle is largely based on the real shuttle, which had its wings very low. If that causes problems, I can't help but wonder how the real shuttle stayed stable. I'll try raising the wings.

Where is your CoL vs CoM when your tanks are emptied to the point where you're attempting reentry?

This gets worse if your fuel tanks aren't properly balanced. Your CoM and CoL will change as your fuel tanks are emptied. I had this problem on my very first real attempt at a spaceplane. The plane flew just fine until it burned about half it's fuel and then it spun around like a top and tumbled through the air.

I have intentionally put the fuel tank as close to CoM as I could, so that it would remain stable regardless of the amount of fuel. This has mostly worked, the glide tests were fine regardless of how much fuel I had, but I still had to empty the tanks (except for monoprop) for re-entry.

Have you 'tweaked your control surfaces to only act on particular inputs? e.g. your Rudder should only respond to Yaw. On my spaceplanes, I usually have Rudders act on yaw only, Elevators (aft) act on pitch only and canards (fore) act on roll and pitch.

Rudder is set to yaw only, elevators have pitch and roll and the canards have pitch only.

What's your Angle of Attack (AoA) on re-entry?

I try to keep it around 40 degrees, like the real shuttle.

is there a bigger rudder you could use? If you look at the actual shuttle, it's rudder was huge.

I think my Rudder/Orbiter ratio is about the same as the real shuttle, if not better. However this isn't like that with most other craft so I will try a bigger one.

I'll see if all that works and post the results.

[xrayfishx]

It helped the yaw stability, but now it turned out that to maintain the 40 degree angle, I would need to hold the nose down, which means that even though I think my shuttle is capable of it, the slightest mistake will mean it still flips. Any ideas how I could make something that can fly stable, but would need the nose to be pulled up instead of down to maintain an angle during re-entry?

This game really needs a wind tunnel.

[Starwaster]

Re: The real shuttle's center of mass: Its center of mass was further downwards, close to its wings. Keep in mind, that area was empty cargo hold, the bottom was heavier structural components, hydraulics, machinery, etc. Top is lightweight cargo bay doors. But the B9 parts you're using have their CoM at the geometric center of the mesh by default so naturally its CoM would be above CoL.

To make that more realistic, it would have to have a comOffset (or is it CoMOffset? it is case sensitive...) of maybe 0, -0.125, 0 (just eyeballing / guesstimating)

Also.. airbrakes on the rudder. I did that to the KSO25 to help keep its nose up.

[xrayfishx]

Yeah, I didn't even think to use the airbrakes that my shuttle already has. But I could very well remove them now, because moving the wings backwards while keeping the canard in the same place did the trick. Best shuttle I've ever built.

[GoSlash27]

That shuttle is largely based on the real shuttle, which had its wings very low. If that causes problems, I can't help but wonder how the real shuttle stayed stable.

Limitations in the KSP aerodynamics is the short answer. IRL the center of pressure (where the drag comes from) is never at the same point as the center of gravity, whereas in KSP they're always at the same place. This makes "weathervaning" impossible in KSP, which is a large part of the reason why the real shuttle is dynamically stable on reentry.

For wings, the center of lift and center of pressure move around and the center of lift is always ahead of the center of pressure IRL. In KSP, they're all in the same place. And as I mentioned earlier, the drag physics for wings are wonky in KSP.

Put all that together, and you have a reasonable shuttle IRL that's dynamically unstable in KSP.

Any ideas how I could make something that can fly stable, but would need the nose to be pulled up instead of down to maintain an angle during re-entry?

First thing I do is balance beam the fuselage so that the center of mass is exactly the same as the center of mass for my fuel tanks. I hook a rig up to it exactly at the center of the fuel tank until it balances on knife edge. This assures that the center of gravity won't move as the tanks drain.

For high speed flight, I set the wings so that the center of lift is level with the center of the fuel tank and maybe throw in some dihedral on the outer wing panels to make it dynamically stable in roll.

Placing the center of lift on the same axis as the center of the fuselage's mass eliminates the dynamic drag instability of KSP. Then it's a matter of taste for how far to put the lift behind the CoG. The further back, the more dynamically stable it will be in stall recovery, but the less efficient your aircraft will be since it has to create drag to counteract the nose-down tendency.

For what I'm doing, I place it exactly at the center of mass for maximum efficiency and minimum control input. In your case, you want it back a bit.

Best,

-Slashy

Edited July 12, 2014 by GoSlash27

[Nothalogh]

This game really needs a wind tunnel.

I would pay for that

[Wanderfound]

Forget about the extreme AoA on reentry; at that angle, your vertical stabiliser is virtually useless. The body of the ship is shielding it from the airstream. The shuttle used its OMS to maintain stability while it was doing that. You can recreate the OMS stability with Vernors, but you'd be better off just using a less aggressive reentry.

My normal reentry profile generally uses a roughly 5° AoA until I get it back to level flight. If you're having trouble slowing, use S-turns and airbrakes. Airbrakes in FAR are easy to create; just set pairs of control surfaces as maxed-out spoilers in opposite directions.

As for general FAR design tutorials...see the first few posts in the Kerbodyne thread and my build videos; both are linked below.