Somethings going wrong with the physics

[maceyneil]

I have a ship with a swept back flower petal design powered by a single engine 900000miles from earth where there is no wind buffeting barrelling out of control all of a sudden.

The ship design is balanced using the auto snap to feature and technically it is just a spear head of mass thrusted in a single direction against gravity and as i tried to align it even with gravity.

I mean WTF i can understand the problems of the \'in atmosphere chair lifting\' designs causing this sort of problem, but there seems to be no rhyme or reason a balanced single directional thrust vehicle with 3 advanced SAS units and a vectored L2 thruster should be rolling like a marble in space.

[togfox]

Sounds like you\'ve been smacked by the Space Kraken. You can search these forums for that one.

[The Cobbler]

Please get your post into a healthy condition first, I cant even understand whats going on.

[softweir]

There are two possible explanations.

One is that some of the many components are overlapping. This means the PhysX engine is struggling to deal with parts of the ship pressing against each other, and it usually fails, badly, leading to 'magic' forces that can cause a ship to pitch, roll or yaw.

Alternatively, if you are as far from Kerbin as you say you are then you may be encountering the Space Kraken. This is a bug that happens when you leave Kerbin\'s Sphere of Influence and enter interplanetary space. When this happens, the game stops representing your velocity as being relative to Kerbin, but instead calculates your velocity relative to Kerbol - which it does by adding your velocity to Kerbin\'s velocity as it orbits Kerbol. The combined velocity is so great that PhysX loses precision and large errors can occur. Under some circumstances this can rip ships apart, but it can also create vibration and turning forces.

The collision problem has been fixed as of Release 0.16, and it is hoped the Space Kraken bug will be fixed in a not-to-distant future release.

[E. F. Kranz]

Overlapping parts. Gotta realize, it\'s not the visible parts themselves, but the collision meshes which are 1. colliding, and 2. invisible.

EDIT:

And now, wait. Why are you using three ASAS units?

The ASAS is the brain. The SAS is the brawn.

(yall feel free to steal that)

[Vanamonde]

There\'s some confusion in my mind still, but isn\'t that a matter of control systems? Advanced SAS must work through control devices and the module exerts no force itself, correct? In that case, the single gimballed engine would be the only thing trying to control the mass of the ship against any slight wobbles or incipient rotations. Shouldn\'t RCS and/or regular SAS help? Also, I\'m told multiple ASAS are redundant while multiple SAS are cumulative, right?

[ManaaniWanderer]

The ASAS is the brain. The SAS is the brawn.

Siging it.

Meet the kraken, it enjoys longwalks on the beach, rain showers and totally messing up your rockets.

[Endeavour]

Advanced SAS must work through control devices and the module exerts no force itself, correct? In that case, the single gimballed engine would be the only thing trying to control the mass of the ship against any slight wobbles or incipient rotations. Shouldn\'t RCS and/or regular SAS help? Also, I\'m told multiple ASAS are redundant while multiple SAS are cumulative, right?

That\'s right.

[samstarman5]

Feel free to share with your friends who also play KSP that you only need one ASAS per craft. Any others would just conflict with each other or be useless mass that would eat up precious delta-v. I have been getting by with just one SAS to supplement in my 3-4 stage rockets that are Mun and Minmus mission capable. They are pretty effective in their own right, as well.

Bear in mind that these are only temporary. It shouldn\'t be too much longer before those little kerbals we keep in the rockets start assuming the duties of the SAS, at least. I wouldn\'t be surprised if the ASAS gets converted to an autopilot(like a stock Mechjeb) for unmanned missions and satellites. In fact, I fully support it.

[Zephram Kerman]

There are two possible explanations.

One is that some of the many components are overlapping. This means the PhysX engine is struggling to deal with parts of the ship pressing against each other, and it usually fails, badly, leading to 'magic' forces that can cause a ship to pitch, roll or yaw.

One of my early failed attempts to launch a cart accidentally demonstrated this problem. As the vessel reached the middle atmosphere, something about the air speed caused the wheels of the cart to gradually extend farther and farther from the cart. Eventually, one or more of them would overlap with a rocket part. These overlaps caused quickly increasing turn forces which would lead to loss of control. The only solution was to either slow down, or get out of the atmosphere. Then the wheels would return to their normal positions, and normal physics would resume.

Be aware that parts can flex and twist during maneuvers. So if the parts twist a certain way into an overlapping position, that might explain the intermittent nature of your sudden loss of control.

[rkman]

In that case, the single gimballed engine would be the only thing trying to control the mass of the ship against any slight wobbles or incipient rotations.

That, and the command pod (which has a small build-in SAS unit)

Vessels up to \'medium\' size can be made stable with only a command pod and (one) ASAS. (your mileage may vary depending on your definition of \'medium\').

Shouldn\'t RCS and/or regular SAS help? Also, I\'m told multiple ASAS are redundant while multiple SAS are cumulative, right?

Correct.

Although a bug causes SAS units to be less effective then they should be. (And mechjeb ignores SAS units all together).