.dll Question Centrifugal Damping

[CaptnDave]

Hi! I'm not sure if that's what it's called, but is there some way to accurately simulate the rotational damping caused by a rotating centrifuge?
I'm trying to do this for counter-rotating centrifuges, so forces acting at 90 degrees is not important, only the factor of rotational damping along axis which are not the axis about which the centrifuge spins.

Thanks!

 

[tblaxland]

It sounds like you are talking about gravity gradient damping to me. Orbiter does simulate it and, IIRC, there are some coefficients that you need need to set in your dll. Check the API docs.

 

[Hielor]

It sounds like you are talking about gravity gradient damping to me. Orbiter does simulate it and, IIRC, there are some coefficients that you need need to set in your dll. Check the API docs.

Hmmm, no, I think he's talking about the effect you get when you attempt to turn a spinning wheel--the wheel resists the spinning motion.

As for doing this in Orbiter, I'm not sure, but you could simply apply a rotational acceleration opposite to any outside rotational accelerations. I'm not sure of the maths involved, and it would require a custom dll for the vessel.

 

[Zatnikitelman]

I'm not sure tblaxland, I think he's talking about resistance to changes in motion of a gyroscopic element. Orbiter is capable of simulating rotational stability to a degree. I believe it's one of the sample scenarios showing the features of Orbiter '06 where one SRB is rotating about its long axis and the other is initially stationary. That said, that's only being applied to one whole vessel at a time as opposed to parts of a vessel like I think you're talking about. So other than faking it by possibly incorporating seperate vessels with rotating docking ports (rot vectors change over a given time) and having Orbiter do what it normally does to docked vessels, I can't see an obvious answer.

(if I messed up any terms, I apologize, it's late here )

 

[tblaxland]

Hmmm, no, I think he's talking about the effect you get when you attempt to turn a spinning wheel--the wheel resists the spinning motion.

You may be right. Certainly such a system requires a torque to change the direction of the angular momentum vector, so perhaps that is what the OP is referring to (although no "damping" is involved). Also, any real spacecraft is non-rigid so any angular momentum it has will tend to couple into the axis with the highest moment of inertia, eg, the unstable Apollo CSM "barbecue roll" (perhaps that is the "damping" the OP is referring to?).

As for doing this in Orbiter, I'm not sure, but you could simply apply a rotational acceleration opposite to any outside rotational accelerations. I'm not sure of the maths involved, and it would require a custom dll for the vessel.

Orbiter simply integrates Euler's equations at each time step, so the effects of resistance to changing the angular momentum vector and torque free precession are modelled quite well in my observations. That said, it does assume that the vessels are rigid bodies, so any flexible body effects would need to be modelled by manipulating the angular velocity vector directly.

 

[CaptnDave]

I'm not sure tblaxland, I think he's talking about resistance to changes in motion of a gyroscopic element.

Yes, that's it. What I am looking for is some way to have centrifuges' resistance to outside rotational forces (such as gravity gradient torque, or RCS burns). The basic setup of my ship is over at http://www.orbiter-forum.com/showthread.php?t=8279. The two things with three arms are the centrifuges (aka gyroscopic elements).

I am writing a custom .dll for this ship, so any fancy code that has to happen will. Does anyone know the math for this, or what fraction of the force is opposed?

I see two options for this (although Grade 11 Physics doesn't delve much into this ) : either have AddForce add a counteracting force to the rotation based on some as-yet-unknown-by-me formula, or could this be done by adding some also-unknown amount to the PMIs?

Thanks for your quick replies!

 

[tblaxland]

I am writing a custom .dll for this ship, so any fancy code that has to happen will. Does anyone know the math for this, or what fraction of the force is opposed?

Try this thought experiment:

1. Set up a vessel that is a rigid planar disk such that the normal to the disk surface is in the direction {1,0,0} in local vessel coordinates.
2. Set the principle moments of inertia such that Ix > Iy = Iz.
3. Spin the disk with some angular velocity about the x-axis.
4. Apply a forces with attack points of {0,1,0} & {0,-1,0} and directions of {1,0,0} & {-1,0,0} respectively.This creates a torque about the z-axis.
5. The disc will now precess about the y-axis and the amount will be inversly proportional to the initial angular velocity about the x-axis.

Orbiter will model the above behaviour correctly.

Since your gyros are counter-rotating, any torque applied to the vessel in the pitch and yaw axes will result in the gyros producing a torque that cancels each other out and the ship will rotate as if the gyros are not even there. Similarly for the roll axis, since any friction between the gyros and the remainder of the ship will cancel out. Witness the totally normal behaviour of a helicopter with counter-rotating baldes as an example.

 

[DarkWanderer]

'm trying to do this for counter-rotating centrifuges, so forces acting at 90 degrees is not important

tblaxland is right here, contra-rotating gyroscopes of equal moment will cancel effects of each other. So you may not need any math at all.

 

[CaptnDave]

Yes, I've found that is correct. Thanks, guys!