How would I calculate a ships Angular Acceleration?

[Xty]

How would I calculate a ships maximum angular acceleration?

I was trying torque / the moi for each component of those vectors, but idk if I figured out torque wrong or something was wrong with the moi, but the roll acceleration came out to like 90000 something, the other ones were like 0.8 (the ship was in space and very simple with just a command pod, fuel thing, and engine).

When I checked out the moi the roll component was like 6E-05, whereas the other 2 were ~5.56. Does this seem reasonable? I wouldn't have expected the roll moi to be so ridiculously small compared to the other two?

Could anyone help me out?

Thanks

Edited January 4, 2015 by Xty

[sarbian]

MoI :

https://github.com/MuMech/MechJeb2/blob/master/MechJeb2/VesselState.cs#L538-L582

Torque from RCS :

https://github.com/MuMech/MechJeb2/blob/master/MechJeb2/VesselState.cs#L356-L374

And Reaction wheel

https://github.com/MuMech/MechJeb2/blob/master/MechJeb2/VesselState.cs#L419-L432

And other module can generate torque of course. And even more with mods.

With those you should get what you need.

As for the value it depends on the ship...

[Xty]

"// KSP's calculation of the vessel's moment of inertia is broken."

Whew, I knew that roll MoI of 90000 seemed ridiculous, thank you.

I'm a few version behind in KSP, I wonder if they have fixed it in newer versions.

Trying out that MoI function got me way closer, but in testing calculating angular acceleration I'm still getting numbers that are slightly off for roll. Pitch and yaw seem good, but for my very simple ship, my function calculates the angular acceleration should be ~7.6 now, but when I actually test it out by rotating for X time then dividing the rotational velocity by the time spent rotating it comes out at ~9.4 pretty consistently.

Edited January 5, 2015 by Xty

[sarbian]

Unity set a limit on a rigid body angular velocity, so you may hit it when doing your test. I don't remember if KSP change it but unity default is 7 and you can check the current default in Physics.maxAngularVelocity and the current value for a part in its rigidbody.maxAngularVelocity.

[Xty]

I did notice that there appears to be some kind of cap. I'm actually having the opposite problem though, the actual roll acceleration (~9.4) is higher than what I calculated through torque and the moi (~7.6), where as for pitch and yaw it's much closer (~.76 actual) (~.74 calculated through torque and moi).

[sarbian]

Do you use only reaction wheels on your test ship ?

[Xty]

It has a command pod, a fuel tank, an engine, and an rcs thruster that I re-purposed to be the part I have my module attached to (idk why I used an rcs thruster for it but i did, maybe it was the part for a tutorial). But the engine and rcs thruster aren't running while I'm testing (at least they shouldn't be, rcs is off and has no fuel, engine is off and also has no fuel).

[sarbian]

I don't know then. When MJ was written the value where ok, but something may have changed ?