Jayjay

This is a fun java applet simulating small satellites orbiting around 2-dimensional earth/moon system.

http://www.colorado.edu/physics/2000...atellites.html

It's fairly easy to get a satellite in a stable orbit around the Earth ("stable" in this case meaning "doesn't crash in the planet or get catapulted away immediately"), but I haven't been able to do it with the moon. Is it possible even in theory?

Lobstrosity

Well, I can get one to orbit the moon three or four times before crashing. It should be possible to do much better--I think the limiting factor is the crudeness of the interface. Another limiting factor could be the accuracy of the numerical simulation itself. Is it just a first-order approximation?

NonHomogenized

for almost 10 minutes, I had one that would loop around the earth, then get pulled into a loop around the moon, and then back into a loop around the earth. Kinda like a planet orbiting moderately close binary stars. It was cool

No luck getting one orbiting the moon, though
But, yes, it is theoretically possible. I don't know how long the orbit would remain stable, though.

Calzaer

How strange.

After about half an hour of trying to make a stable earth orbit, I got frustrated and started clicking randomly. After about a hundred clicks, my agression was spent... and I had two satellites in a relatively stable orbit.

What's odd, though, is there are two other satellites in a HIGHLY erratic orbit, spending more time off the screen than on it, but they haven't gone rocketing off into space yet nor have the collided with the earth or moon. They just occasionally fly through the picture at seemingly random intervals.

Shake

I tried to get one to do something like that. None would stay that way for very long. I had some cool ones that would go around for a little while, and eventually get sucked in. There was one that started backing up, but once the oscillations started, it was doomed. I even had a couple that would leave the screen, be gone for minutes on end, and the reappear. None of them have seemed to last either, though.

Right now, I've got one going that's been up for a good 10 minutes, and it's still going strong! Heck, make that 15, now!

I quickly learned a few things:
1) try an elliptical orbit if you can
2) keep it small.

If the orbit is too large, perturbations from the moon will sap its energy, leaving it subject to the earth's gravitational pull. Too small, and you only get a few revolutions before it crashes. Much too big, and your satellite is gone forever, lost in space.

With the resolution on my screen here, the area shows up as being a bit smaller than a 5" square. My successful satellite has a slightly elliptical orbit which on my screen is never longer than 1" (yes, I actually put a ruler up to the screen). Interestingly, I just noticed that it's not always the same ellipse ... sometimes the orbit looks a little more circular.

Wow! This is so cool! I wish I could save it somehow, so I could show you all what it looks like. We're coming up on 20 minutes now, and it's still going!

Shake

Alright, I took a screen shot, and then added an ellipse using MS Paint, to give you an idea of how close my satellite is. She's still going! I don't want to close it!

Shake

D'oh! Forgot to attach the pic. Just going too fast, I guess.

trunks2k

I managed to get 4 in a stable orbit around the Earth at once for several hours. I then started randomly clicking around to see what interesting orbits I'd get. Which caused the destruction of two of the stable orbits. So I had two orbits that made it through the night. I just feel bad for Earth, the life forms on there must have been destroyed hundreds of times over because of me.

Rhaedas

From past experiences with other orbit simulations, one problem is that each calculation has to be done in steps. Ideally you want to minimize the position step increments to zero spaces inbetween. However the more calculations you do, the slower the program will run.

Where you notice the errors creep up is in high velocity areas, such as a close pass or in the case of trying to orbit a moving object, such as the moon in this case. The calculations for each step are not a correct summation of what the true interaction would be.

I've seen some programs that monitor those types of situations, and further divide those object movements into smaller portions to minimize the error.

And of course, there's math errors that add up, such as rounding.

In real orbits of course, the delta is zero, it's all in real time.

HeatherD

Death Incarnate:
So, where's your god now?

(Requires QuickTime®, 1.2MB size)