Reducing Orbital "tumble"

[XonE32]

hey all,

Figured this was the better place to pose this question as it has to do with the physics of mass in space/orbiter.

I've recently spent hours of fun building a new station using Greg Burch's great Building Building Blocks 4.0/4.1 add-on. All told the station consists of of over 80 building blocks (vessels) including 2 of his older Airdock2's. I'm not at home at the mo so I can't give specifics on it's total mass or even throw up a snazzy pic, but I've noticed she does like to "tumble" around in orbit.

I've adjusted certain settings in the .ini like rcs thrust etc.. to make it so it's pretty easy to killrot when another ship wants to dock, but I doubt these settings are realistic.

My question has to do with reducing the "tumble" naturally based on orbit altitude and eccentricity.

In short, would something tumble less or more in Earth orbit if it was in higher orbit or lower orbit. I've tried Alts from 250km to 10000 and she still likes to tumble way more than most other vessels I see (e.g. ISS).

I would have assumed: the higher the mass of an object and the lower the orbit alt, the less the tumble, but I just don't know.

In fact I don't even know if or how much the ISS tumbles around today or if it does how they compensate for it.

Any thoughts would be appreciated and thanks in advance.

regards

XonE32

 

[Andy44]

Assuming you are discussing attitude? If so, try turning off gravity gradient, and make sure your entire orbit is above the atmosphere, which in Orbiter cuts off completely at 200 km. Also, Enjo made a plugin that fixes a bug in Orbiter called "Absolute KillRot" or something like that. I don't know if we still need that anymore, but it's worth a search.

If you have gravity gradient on, the best you can do is try to find an attitude or design your station such that it works with gravity gradient rather than against it. For instance, in real life, the space shuttle orbiter flies with its tail pointed at earth center, and you can do this in Orbiter, too, but you will never get it perfect; you'll always need to hit the thrusters once in a while to correct it.

 

[Missioncmdr]

Do you have gravity-gradient torque on?

Edit: Damn. Andy beat me!

 

[JamesG]

The ISS has flywheels on board that when spun up (remember that basic high school physics lession with the bicycle wheel?) provides stability and damps drift.

 

[XonE32]

heh, yeah I have all "realism" settings turned on, turnin' em off is a no go for as I like my sims as real as possible... unless of course the gradient torque option is bugged or something.

Yes Andy I am referring to attitude. Can't remember for sure , but I think the current orbit right now is about 430km-600km. Thanks for reminding me about the killrot plugin, I had it before I think, but i've been away from orbiter for a looong time, so I'll have to go go lookin' for it.

As it stands now, the changes I've made to the airdock2 .ini work fine with the extra mass added from all of the building blocks I've added, but they (the .ini settings) are definitely a far cry from the default settings. (I seem to recall changing rcs thrust setting to something unrealistic like 700000 lol to compensate for all the extra mass). Hence, why wanted to know if there was a "natural" way to reduce the tumble using orbital alt.

With the default rcs settings on the Airdock and all this extra mass a killrot takes forever after only a few orbits lol.

I will look into a redesign (i.e. adding certain modules elsewhere to even out the mass and make it more symmetrical). Maybe that'll help as you suggest.


-----Posted Added-----

The ISS has flywheels on board that when spun up (remember that basic high school physics lession with the bicycle wheel?) provides stability and damps drift.

Cool, thanks James. I had no idea.

My knowledge of the ISS falls short of even "limited".


EDIT: There is a preliminary (ie old) pic in this thread http://orbiter-forum.com/showpost.php?p=52589&postcount=6 ,but I've since added and changed some stuff. Perhaps it's center of balance is off and is causing it to tumble more than most vessels?

 

[Sherrif of Nottingham]

I remeber the lesson\playback with the SRB's one spinning and the other is still, the Non-Spinning one "tumbles" around while the spinner is staying at the same attitude. I think it has something to do with Gradien Torquing. I usually set my stations around a Dragonfly that tries to maintain Prograde, that way my station stays the right way in orbit.

 

[joeybigO]

Basically I think how this was fixed, IIRC, was with the quaternon matrix that dampened out oscillations (check spelling on that).
Basically firing thrusters every once in a great moon, to dampen, otherwise it kicks out, exp. at lower altitudes, with no atmosphere.

 

[tblaxland]

The ISS has flywheels on board that when spun up (remember that basic high school physics lession with the bicycle wheel?) provides stability and damps drift.

Technically they are called control moment gyroscopes (CMGs). Also, the station nominally flies in what is known as TEA (torque equilibrium attitude). This close to +XVV -ZLV (Harmony forward, port truss left, Z1 truss up) but is skewed a little so that the effects of gravity gradiant and atmospheric drag cancel each other out. TEA leaves the CMGs with little to do most of the time.

As it stands now, the changes I've made to the airdock2 .ini work fine with the extra mass added from all of the building blocks I've added, but they (the .ini settings) are definitely a far cry from the default settings. (I seem to recall changing rcs thrust setting to something unrealistic like 700000 lol to compensate for all the extra mass). Hence, why wanted to know if there was a "natural" way to reduce the tumble using orbital alt.

With the default rcs settings on the Airdock and all this extra mass a killrot takes forever after only a few orbits lol.

I will look into a redesign (i.e. adding certain modules elsewhere to even out the mass and make it more symmetrical). Maybe that'll help as you suggest.

Try flying your station in an attitude that has the axis with the highest moment of inertia perpendicular to the radius vector from the Earth. This will help stop the rotation getting started in the first place. From your picture this will probably mean having one solar panel on the Earth side and one on the space side (assuming you have the PMIs set right in your config file).

I'm not sure if Spacecraft3 supports gravity gradient damping. If it does, you may want to increase these numbers. If your station is off the nominal attitude mentioned above it will start to rotate due to the gravity gradient. Without gravity gradient damping it will continue to tumble (a bit like a friction-free pendulum). In real life, gravity gradient damping is caused by spacecraft not being true rigid bodies (they have liquids on them, for example).


-----Posted Added-----

Basically I think how this was fixed, IIRC, was with the quaternon matrix that dampened out oscillations (check spelling on that).
Basically firing thrusters every once in a great moon, to dampen, otherwise it kicks out, exp. at lower altitudes, with no atmosphere.

Nits:

1. quaternon -> quaternion
2. dampened -> damped
3. Strictly speaking, quaternions are not matrices although they can be represented as such. Orbiter uses rotation matrices instead of quaternions.

 

[joeybigO]

I knew it needed help on the spelling. Some days it's not there. I took a shot and lost on that one.

Chris Kinestrick who wrote the original Land MFD for the final 3 programs for orbiter had the same problem, I do not recall how he fixed it.

 

[Andy44]

Also, if you're going to try to work with gravity gradient stabilization, make sure your orbit is nearly circular (ecc =~ 0). This will ensure that your rotation rate will be as close to constant as possible, otherwise gravity gradient forces will cause you to start rocking back and forth.

 

[tblaxland]

I knew it needed help on the spelling. Some days it's not there. I took a shot and lost on that one.

Chris Kinestrick who wrote the original Land MFD for the final 3 programs for orbiter had the same problem, I do not recall how he fixed it.

I didn't know Chris was involved in Land MFD, I thought that was LazyD's work.

Normally, gravity gradient torque is not an issue under powered flight since it is much smaller than any torque's applied by engines/thrusters. Are sure you are not getting confused with torque-free precession? This can be seen as "tumbling" once a rigid body is in motion. This affects one addon I know Chris was involved in - AttitudeMFD. As far as a solution goes, AttitudeMFD didn't - and still doesn't - take torque-free precession into account although doing so would make it more efficient in terms of thruster fuel used. I do not think that solving for the polhode is possible in closed form the general case, although it is for specific cases where at least two of the principle moments of inertia are equal.

EDIT: I should have looked harder, thanks to Google Books, Analytical Mechanics seems to have a solution to the Poinsot Construction (the general case mentioned above). I think I just found my bed time reading for tonight
http://books.google.com/books?id=1J...hanics&sig=ACfU3U3dv98w21Jy7n_2x_pKFoo8EMF8dA

 

[XonE32]

Finally back at my sys. Thanks for all the replies, as there's some great info here for me to play with to see if I can stabalize the "rockin and a rollin" naturally.

I remeber the lessonplayback with the SRB's one spinning and the other is still, the Non-Spinning one "tumbles" around while the spinner is staying at the same attitude. I think it has something to do with Gradien Torquing. I usually set my stations around a Dragonfly that tries to maintain Prograde, that way my station stays the right way in orbit.

Yes I've watched that .scn as well. The Dragonfly would be a great idea because it would mean one main vessel controlling the station as opposed to 2 Airdock2's fighting for control lol, but I was under the impression/assumption (never having used the Dragonfly myself) that it wouldn't have the power to correct (i.e Killrot) this much mass.

Rough est on total mass is 40000kg per Airdock2 plus 300000kg fuel for each plus God knows how much kg for each of the 87 or so vessels I've got attached. It would take quite a while to compare the station's .scn file with the vessels ini's and .cfg to come up with a total lol.

Also, if you're going to try to work with gravity gradient stabilization, make sure your orbit is nearly circular (ecc =~ 0). This will ensure that your rotation rate will be as close to constant as possible, otherwise gravity gradient forces will cause you to start rocking back and forth.

I've got ecc to zero Andy, but within a few days of orbits, it basically expands to a range from 0 to .0009. Not too bad I guess.

Try flying your station in an attitude that has the axis with the highest moment of inertia perpendicular to the radius vector from the Earth. This will help stop the rotation getting started in the first place. From your picture this will probably mean having one solar panel on the Earth side and one on the space side (assuming you have the PMIs set right in your config file).

I'm not sure if Spacecraft3 supports gravity gradient damping. If it does, you may want to increase these numbers. If your station is off the nominal attitude mentioned above it will start to rotate due to the gravity gradient. Without gravity gradient damping it will continue to tumble (a bit like a friction-free pendulum). In real life, gravity gradient damping is caused by spacecraft not being true rigid bodies (they have liquids on them, for example).

I will fiddle with orientation, and see if that helps too. Thanks for that TB.

 

[XonE32]

Sorry for bringin' this up again, but I didn't want to start a new thread on essentially the same thing. I left this prob alone after I tinkered with the rcs settings of the Airdocks (i.e cheating, or at least making the settings somewhat unrealistic). But I'm back at it trying to use the default rcs settings. But I just wanted claricification on this quote from Tblax:

(assuming you have the PMIs set right in your config file).

I haven't made any vessels for Orbiter so very few of the settings in the .cfg's make sense although some do. The PMI settings are ones that I don't understand (they are currently set to the default for the Airdock2 PMI=(58.2,62.1,7.7). All I know about PMI is what it says in the Orbiter .PDF which hasn't helped elucidate and explanantion for me. It basically tells me what the acronym stands for and that's about it lol.

mass-normalised principal moments of inertia
(PMI)

How do I know if the above default settings are correct or need to be changed and if they need to be changed how should I change em. In other words can someone splain to me PMI.

Any input or thought would be most welcome.

Regards

XonE 32


Are the settings for PMI=(58.2,62.1,7.7) x, y, z?

 

[Epsilon]

As far as I can gather, PMI is an object's resistance to rotation.

http://en.wikipedia.org/wiki/Principal_moment_of_inertia

It gave me some insight. Low PMI would mean it's easier to make something rotate, and easier to stop it rotating. Hopefully someone who actually knows what they're talking about can chime in. <_<

Edit: (oops. Easier to stop, not harder.)

 

[Andy44]

Yes, the settings are for x, y, and z. A low moment of inertia along the z axis means that it's easier (meaning less inertia and thus less resistance to angular acceleration) to rotate the object about the z axis.

If you take a long, thin object, like say, the spaceship Discovery from 2001: A Space Odyssey, and you define the long fore and aft axis as the z axis, then you would say that the z PMI of Discovery is small relative to the x and y PMIs, which will be relatively large. Discovery, therefore, in the presence of a large gravity gradient in a place like LEO, will have a tendency to want to point its nose or tail directly at the center of the Earth.

 

[XonE32]

Yes, the settings are for x, y, and z. A low moment of inertia along the z axis means that it's easier (meaning less inertia and thus less resistance to angular acceleration) to rotate the object about the z axis.

M'kay, graspin it now Andy. Thanks for that and everyone else that's replied. I've been away from Orbiter for too long and totally forgot about the visual helpers for force and axis.

So according to the the PMI (for the Airdock2), the highest moment of inertia is the y axis? And thus according to Tblax (who sounds like he knows a whole lot more than me), I should rotate my station so that the y axis is "perpendicular to the radius vector from the Earth". I'll post a piccy and maybe someone will comment if they think it's correct or not.

Thanks again all.

Regards

XonE 32

 

[Andy44]

The axis with the lowest PMI is the one that will tend to rotate parallel to the R-vector (between the center of Earth and the vehicle), so unless you want to waste fuel, that's the axis you should point at the earth for a gravity gradient stabilization scheme. That is what the space shuttle orbiter does by pointing it's tail at the Earth when it's just cruising along.

I don't think the ISS does this, though, because it uses gyros to store angular momentum, which from gravity gradient should average to zero over an orbit, someone correct me if I'm wrong. In Orbiter, as far as I know, there is no way to model gyroscopes yet, at least not without a way-cool custom dll.

 

[XonE32]

Oooh,k. so I guess the below pics are wrong. I will try with the Z axis next as it's has the lowest PMI according to the .cfg. 1st pic is the station with no indicators so ya can see what it looks like without all the axis & force bars. Maybe the fact that it's a compilation of vessels using SBB's and it has a rotation HAB is complicating the issue of stabilization?

http://i5.photobucket.com/albums/y189/xone32/Orbiter/Station1-1.jpg

Is below what is meant by orienting perpindicular to the y-axis?

NOTE: Ignore the difference in force Newtons between the airdocks. There was a temporary screw-up in the .ini with regard to propellant.

-----Posted Added-----

The axis with the lowest PMI is the one that will tend to rotate parallel to the R-vector (between the center of Earth and the vehicle), so unless you want to waste fuel, that's the axis you should point at the earth for a gravity gradient stabilization scheme. That is what the space shuttle orbiter does by pointing it's tail at the Earth when it's just cruising along.

Well that worked a bit better I think Andy. Me sitting there counting in 5 second increments and then measuring angular momentum.. pretty sad lol. So I have my Z axis pointing at the earth which is fine and I also tinkered with the station design to make it more symmetrical, but I just have one question:

If I load up one of the old default scenarios like "Quickstart" in the checklists folder, then switch my view to the ISS to watch how it behaves, then turn on the visual helpers for axis and force... The yellow GN bar matches perfectly with a grey F bar. I assume that's the R vector you're reffering to. My questin is why can't I see that grey F (Force?) bar when I load up my scenario? The only grey bars I see when I turn all the visual helpers on are the X,Y,Z axis bars. None labelled F or Force?

I'll scootch around the orbiter manual, maybe I'm missing a module?


Regards

XonE 32

 

[Andy44]

I almost never use the visual helper stuff, so I can't help you there...

 

[XonE32]

Yeah the visual helper stuff doesn't really work in this case. I can point the station along the proper axis as you suggested and it does reduce the amount of angular motion, but reducing it to almost nill and keeping it there over long periods would probably mean a slightly finessed orientation and figuring that out would take some serious math. And no matter how finessed, your gonna need to killrot sometimes no matter what as you suggested.

At least I did find out that the reason the force bar dissapearing was because it was a "multi-vessel" station. Undock a piece or look at a single ship and you see it fine. Pg. 93 of the Orbiter .PDF mentions it can get broken, just not how so I tested it with a single piece from the station and it showed up right away.

Regards

XonE 32