binary star

[Pierebean]

this topic is only for the 1.4 nex version.

What if we add a third mass in a binary system, for example a planet?

Will the soft include the interation of both stars or just one star? Will the trajectory of the planet be a simple conic or not?

here an animation of the kepler's laws with three masses, we can notice that there are a lot of trajectories possible:

http://www.physique.usherbrooke.ca/~dse ... pler3C.htm

[selden]

Celestia draws objects where you tell it to draw them. The Mass of an object has no effect: Celestia does not implement gravity at all. It is up to the designer of a system to make sure that the orbits are correct in their STC and SSC catalog files.

[Pierebean]

So, the disigners are currently including those trajectories. and by the way: perhaps it will be more simple to take account of the mass. therefore the trajectories will be automaticly ploted thanks to the Kepler laws ect....

[maxim]

The whole gravity issue (already mentioned several times) is simply a question of rendering speed - that's all.

If someone can provide some formula that does orbits at least as exacly as they are now AND allows for 20-40 fps on a usual computer system - I'm sure they woul be included ASAP.

maxim

[Evil Dr Ganymede]

The whole gravity issue (already mentioned several times) is simply a question of rendering speed - that's all.

If someone can provide some formula that does orbits at least as exacly as they are now AND allows for 20-40 fps on a usual computer system - I'm sure they woul be included ASAP.

Do the boxed programs like Starry Night use gravity in their orbit calculations? Or are those solely based on ephemerides?

[selden]

Starry Night uses several ephemerides, although I haven't found a reference saying which ones they use. See http://www.starrynight.com/support/Know ... d=111&c=34

[Bob Hegwood]

Starry Night uses several ephemerides, although I haven't found a reference saying which ones they use.

Sorry to interject, but could one of you more knowledgeable types
explain - in plain English please - what are "ephemerides?"

Thanks, Ol' Brain-Dead

[selden]

Bob,

An ephemeris is a table or relatively simple formula that provides pre-calculated location information for bodies in space. Someone already has done all the calculations to account for all the complicated gravitational interactions. You just need to know the precise date and time in order for an ephemeris to provide the location of a body.

Some ephemerides provide the altitude and azimuth angles, i.e. the elevation and geographical direction that you should look toward to see a planet. Others provide the 3D location of the planet in space.

Celestia, like many commmercial planetarium programs, uses a set of ephemeris formulas that's called VSOP-87.

[Bob Hegwood]

Thanks very much for that explanation, Selden.

You are the MAN when it comes to explaining things for the
Brain-Dead, and I really appreciate it.

Take care of yourself, and your health...

Thanks again, Bob

[Pierebean]

if the gravity cannot provide a accurate trajectory for most of the space bodies.
Perhaps will be usefull to space bodies outside the solar system where the accuracy is not capital.I mean put the gravity as an option that we can toggle or not, just like the speed light travel delay.Thus, in binary system for instance there won't be any wrong trajectories.

these are only suggestions. i hope i don't make the forum redundant.

[t00fri]

if the gravity cannot provide a accurate trajectory for most of the space bodies.
Perhaps will be usefull to space bodies outside the solar system where the accuracy is not capital.I mean put the gravity as an option that we can toggle or not, just like the speed light travel delay.Thus, in binary system for instance there won't be any wrong trajectories.

these are only suggestions. i hope i don't make the forum redundant.

We normally do not implement physical phenomena into Celestia that can only be realized incorrectly. As a design goal, it would not be very convincing to /partially/ incorporate gravity effects on the observer, depending on where he/she happens to be...

I did not understand your remark about binary systems. I am putting in a considerable effort to get the binary trajectories RIGHT. This is not easy, but possible. They can easily come out WRONG

My light-travel delay feature is an /exact/ implementation.

Bye Fridger

[Pierebean]

There was a misunderstanding:
i'm rephrasing:
the trajectories of two stars orbiting around each other seems perfecty RIGHT (two ellipses resulting of one "barycentrick" other ellipse).

but if you add severals other masses (even one) it became quickly quiet chaotic and very hard to plot (i don't this that the differential equation which discribes the motion is integrable exactly).

so i did propose to include a gravity laws in oder to plot automaticely the trajectories of the added masses.

and of course, to avoid any hard computing and wrong accuracy those gravity laws would occur only outside the solar system for instance.
the gravity laws are as exact as the light travel delay, i guess.

sorry again to the misunderstandind, i noticed that maths and physics are hard to translate with accuracy.

[selden]

Celestia never will implement gravity. The reasons have been discussed ad-nauseam.

Orbits must be defined analytically.

The current versions of Celestia support three types of orbital definitions:

1. internally using ephemerides (currently VSOP 87)
2. externally using xyz trajectories
3. externally using Keplerian orbital parameters

If you can propose (and better yet, provide the code for) other methods of defining orbits which have time as the only independant variable, please do so.

[Michael Kilderry]

Is there any program that makes an .xyz file so that the orbit of an object incorporates gravity?

Michael Kilderry

[Pierebean]

So if I understood well, all non-analytically described trajectories cannot be incorporated in celestia (apart xyx trajectories). As the trajectories of three masses cannot be described analytically. the only way is the xyz?
in that case, this pointless to think about it currently, because I understood that xyz are not periodic so the motion won't be described "for ever".

did I understand well? Am I clear?

[t00fri]

So if I understood well, all non-analytically described trajectories cannot be incorporated in celestia (apart xyx trajectories). As the trajectories of three masses cannot be described analytically. the only way is the xyz?
in that case, this pointless to think about it currently, because I understood that xyz are not periodic so the motion won't be described "for ever".

did I understand well? Am I clear?

I think there is quite a bit of confusion, really (c..f Michael above!).

We have to clearly distinguish 3 kinds of gravity effects:

1) gravity effects on massive bodies leading to periodic motion
2) gravity effects on massive bodies (e.g. spacecrafts, asteroids,...) leading to non-periodic motion
3) gravity effects on the observer.

(3) is consistently ignored in Celestia. The observer rather acts like a "test particle" in theoretical physics, i.e. he/she is considered massless (no gravity attraction from massive bodies) and kind of a "Superman/woman", able to travel at a speed much bigger than that of light!

(1) is taken into account in a most sophisticated manner by means of the VSOP 87 theory. For each planet, it involves hundreds of Tchebichef polynomial terms, accounting for the subtle gravity perturbations from all other orbiting bodies! So here effectively multibody effects due to gravity onto the body under consideration are accounted for. The underlying philosophy is quite analogous to the Hartree-Fock method in atomic physics...The resulting accuracy is phenomenal!

(2) This is largely the domain of the xyz data files and if less accuracy is required, of parabolic or hyperbolic solutions of Kepler's eqs.

But of course, Michael, xyz orbits take into account through numerical data all known gravity effects that might influence the object's motion.

Bye Fridger

[maxim]

BTW, a periodical XYZ trajectory wouldn't be a bad idea. Good for modelling periodic but unusual 'tumbling' orbits.

Shouldn't that be implementable with just a few lines of extra code, dealing with a modulo operation and delta-t time calculation?

maxim

[jestr]

Yeah this would be very useful if implemented,Jestr

[Pierebean]

i'm quiet relieved that n masses trajectories are ploted with accuracy.
hence, the trajectories of 3 masses in a binary system is reachable.
I will look for the ploting of such trajectories.
thx to make me understand.