Celestia Forums

The International Space Station will be correctly oriented relative to its orbit if you add the following parameters:

Obliquity 51.5684 # should be same as orbit inclination
RotationOffset 90 # to get the vertical axis right
LongOfRotationAxis 343.1518 # should be same as long. of ascending node

This puts it with the central module axis going fore-aft, with the US Destiny lab toward the front, and the main solar panel truss pointing transverse to the direction of motion. This is the current orientation of the station, and, I believe, the one planned going forward.

(The central portion of the main truss just got attached on the most recent Shuttle flight; most of it isn't there yet.)

Many artificial satellites should probably have similar tweaks added. I noticed that the Mars Odyssey and MGS orbiters seem somewhat haphazardly oriented.

And here's Mir:

Obliquity 51.6461 # should be same as orbit inclination
RotationOffset 180 # to get v-bar pointing vertically
LongOfRotationAxis 83.8459 # should be same as long. of ascending node

In this case I'm not sure about the front-rear orientation, but this at least gets it right way up (not counting the various times that it was tumbling out of control...)

Hmm... I tried doing the same thing for Skylab, but then looked at some old NASA photos and saw that Skylab seems to have been generally oriented so that its solar panels and solar telescope faced the sun, so the principles are different... it should have a period of one year, and the orientation might be tricky to figure out... and it's time for bed.

Well, I can't get Mars Odyssey oriented right, because the "north-south" axis of the model is parallel to the instrument orientation, so I can't get the instruments to turn to keep facing the ground (until I have a version of Celestia with precession working).

But I can at least get Mars Global Surveyor to point down:

Obliquity 93.0092 # should be same as orbit inclination
RotationOffset 290 # makes the business end point generally down
LongOfRotationAxis 28.452982 # should be same as long. of ascending node

Ok, NOW it's time for bed.

I'm adding an orientation parameter for Celestia 1.2.5. This will let you correctly orient spacecraft and other objects without having to tweak the model file. As it is now, the only way to rotate an object is to adjust its rotation axis.

--Chris

I've been trying (and failing) to get a ship to orbit earth while continually pointing in the direction of travel. Having read this I think it's the model, which stands on it's nose by default. But anyhow, is there a list anywhere of what effect the orbiting parameters have, so we can avoid the trial and error method?

Here's my best attempt at getting Shrox's Skylab model pointed in the right direction. I make no guarantees that this is historically accurate in every particular, but at least the Apollo telescope mount points at the sun, as close as I can manage:

Obliquity 23.45
RotationPeriod 8766.15265008 # sidereal year
RotationOffset 191
LongOfRotationAxis 0

The principle here is that the obliquity is the exact opposite of Earth's, so that its rotation axis (which happens to be down the axis of Skylab, in this case) is perpendicular to the ecliptic. The rotation period is the *sidereal* year-- the time it takes Earth to go around the sun as an absolute revolution in space, disregarding any particulars of its rotation, which are irrelevant here.

There's some unavoidable wobble because of the ellipticity of the earth's orbit. But the station stays more or less pointed at the sun for a period much longer than the historical life of Skylab!

Lostinspace wrote:I've been trying (and failing) to get a ship to orbit earth while continually pointing in the direction of travel. Having read this I think it's the model, which stands on it's nose by default. But anyhow, is there a list anywhere of what effect the orbiting parameters have, so we can avoid the trial and error method?

If the rotation axis of the model is fore-aft then you're out of luck for now. It needs to be transverse, and then you need to adjust it to be perpendicular to the plane of the orbit-- that is, perpendicular to the direction of motion and parallel to the ground.
By a combination of wild guesses and trial and error, I found that the way to do this is to set LongOfRotationAxis equal to the longitude of the ascending node (which doesn't make sense to me-- it seems like it should be 90 degrees off unless these terms don't mean what I think they mean, but never mind), and Obliquity equal to the orbit inclination. Once you do these things then the model will have a constant orientation with respect to the ground (if its orbit is more or less circular).

Then you tweak RotationOffset to whatever you need to compensate for the model's orientation. This only gives you one degree of freedom, of course, hence the need for a conveniently oriented rotation axis to begin with.

(By the way, my use of "v-bar" in the comments above is incorrect-- it should be "r-bar." I was misremembering my bizarre astronaut jargon gleaned from Shuttle/ISS crew interviews.)

Matt McIrvin wrote:If the rotation axis of the model is fore-aft then you're out of luck for now. It needs to be transverse, and then you need to adjust it to be perpendicular to the plane of the orbit-- that is, perpendicular to the direction of motion and parallel to the ground.
By a combination of wild guesses and trial and error, I found that the way to do this is to set LongOfRotationAxis equal to the longitude of the ascending node (which doesn't make sense to me-- it seems like it should be 90 degrees off unless these terms don't mean what I think they mean, but never mind), and Obliquity equal to the orbit inclination. Once you do these things then the model will have a constant orientation with respect to the ground (if its orbit is more or less circular).

You're correct that LongOfRotationAxis is 90 degrees off . . . I'm fixing this in 1.2.5, and was hoping that noone would notice in the mean time. Oops. I hope that there aren't too many add-ons making use of this yet.

--Chris

chris wrote:You're correct that LongOfRotationAxis is 90 degrees off . . . I'm fixing this in 1.2.5, and was hoping that noone would notice in the mean time. Oops. I hope that there aren't too many add-ons making use of this yet.

Heh! Well, it's easy enough to change the numbers I came up with.