Shuttle wobbling

[Nacho Grosso]

I've got a problem with my shuttle, I can't understand why when I try to point it nose right to 45 degrees to set my desired apoapsis, it starts to wobble in its longitudinal axis and it turns completely uncontrollable. I tried to fix it adding more struts between the shuttle and the boosters as well as between the shuttle and the external tank, and adding more RCS thrusters and a SAS too, but it didn't work so much. Any ideas about how to fix it?

Edited May 2, 2016 by Nacho Grosso

[James Kerman]

It could be asymmetrical thrust that's being magnified by SAS.  If you could post an image or craft file, I'm sure the space-plane maestro's can help you out.

[CodeFantastic]

Could be the control surfaces on the shuttle, try disabling them.

[Gaarst]

Yes, pics and a description of your ascent profile will be helpful.

Space shuttles are inherently unstable so hard to make. Highly gimballed engines (Vectors for example) can cause some wobbling because SAS doesn't know how to keep a rocket pointing forwards.

[Snark]

Moving to Gameplay Questions.

Please post a picture of your shuttle.

A very common problem with shuttle designs is that people want to make it look just like the US space shuttle.  It's easy to do that with the parts in stock KSP (because they're designed specifically to make it "look like the shuttle") ... the problem, though, is that if you do that, you'll usually end up with an uncontrollable mess.

Best way to understand that is to look at the shuttle in the vehicle editor, with the CoM indicator turned on.  If you're not being careful, a NASA-shuttle-lookalike design in KSP will often have its CoM located way in the back of the ship, where the heavy parts (i.e. engines) are located.  This hits you with a double whammy.  On the one hand, it means that there's a huge, light, draggy fuselage sticking way out in front, making the ship very aerodynamically unstable.  On the other hand, the back end is where all your control surfaces are, too, and since they're too close to the CoM, they have very little control authority and therefore can't help compensate for the instability.

The solution will depend on your design, but it's often some combination of "move the CoM forwards" and/or "put some control surfaces at the front end, as far away from the CoM as possible."

Here's a similar thread that you may find helpful:

But it's hard to say for sure, without seeing a picture of your shuttle. Screenshot please?

[Gaarst]

^ As Snark said, mass balance is also a big issue when making STS-like space shuttles and when flying them back home.

In KSP, the mass balance of parts is far from what it's really like. Take the Vector: 4t heavy, 1MN of thrust, usually placed on 20-40t heavy shuttles for your average STS replica. Now consider the RS-25 (SSME): 3.5t heavy, thrust over 2MN, usually placed on a 70t shuttle (or a 3000t rocket but that's not the point here).

Placing 3 Vectors on your shuttle can account for a third to half its mass which is a lot, and will put your CoM way at the back.

[Variety_Pack]

I know this post is pretty much right at a year old but I'm gonna revive it anyways.

I'm having this same problem. I've built a shuttle and it works really well if it doesn't veer too far off course during ascent. The problem, as far as I can tell, is pretty much exclusively the SAS being waaaay too twitchy. If you stare at the little pippers indicating control inputs, you'll see that they snap wildly around. Sometimes the shuttle makes it to orbit with enough fuel, often the shuttle wiggles too far off course, occasionally it even throws itself too far in a direction and enters a sekrit backlip mode or a speshul flatspin ascent profile.

The best way I've found to counter this is to use my joystick, but I feel physically exhausted after 6 minutes of Jedi-quality midichlorian-tapping predictive flight correction. It's truly tiring but I can get to orbit just fine. Also, the gigantic shuttle tailfin is too big and causes a huge torque moment, but there isn't really a suitable replacement so I just disable the control surface or reduce it severely. The SAS wheels can be sufficient for yaw if you have enough of them, although I just rely heavily on Vernor RCS.

 

In short, if there's an improvement to be made in the game, it would definitely be the SAS control. It's too spastic and it doesn't predict trends like a human or a decent algorithm. Instead of damping wobbles it amplifies or exacerbates with too-strong of course corrections.