Celestia Forums

Hey ! I'm new here,

I would like to know if there was developpement for including physics in celestia : to include gravitation laws and let the simulation run to see how it evoluates, try to put objects in orbits from earth for example or even visualize tidal forces (display of gravitation force vectors on bodies)... And with collision detection, reproduce events such as the collision between a comet and Jupiter...

I could to help in including such capabilities.

leFab wrote:Hey ! I'm new here,

I would like to know if there was developpement for including physics in celestia : to include gravitation laws and let the simulation run to see how it evoluates, try to put objects in orbits from earth for example or even visualize tidal forces (display of gravitation force vectors on bodies)... And with collision detection, reproduce events such as the collision between a comet and Jupiter...

I could to help in including such capabilities.

There is already quite a lot of physics contained in Celestia;-).

We had a lot of discussions before about including gravity effects (just search for them with the forum engine).

But since a /proper/ inclusion of
gravity is very hard (please note, I know what I am talking about!), we decided against it. Celestia is NOT a naive space game, but supposed to be a most accurate, sophisticated space simulation...

Of course the more or less trivial approximation of widely separated gravitational point sources is getting entirely inadequate, when you want to manoeuver e.g. among the Jovian moons in the huge and complicated field of (non-spherical) Jupiter...

The general philosophy has been so far to rather ignore a particular physics effect entirely rather than including it /incorrectly/.


It would be very challenging and interesting, however, to address /relativistic/ effects. Notably shifts on the field of view and colour shifts. Unlike gravity, relativistic effects can be accounted for more or less rigorously. The main question here is whether all this can be coded /fast/ enough.

Bye Fridger

the relativistic effects can certainly be coded and quicly t00fri, just check out http://www.starstrider.com

[quote="t00fri]
Of course the more or less trivial approximation of widely separated gravitational point sources is getting entirely inadequate, when you want to manoeuver e.g. among the Jovian moons in the huge and complicated field of (non-spherical) Jupiter...
[/quote]

Thank for your answer

Are you sure that the gravitational point sources approximation introduces too much inaccuracy ? Wouldn't it be accurate enough to see how the system evolves with this laws (and if it remains stable) or to send space probes ? When Voyager probes trajectory were calculated, this approximation was not used ?

[quote="t00fri]
Of course the more or less trivial approximation of widely separated gravitational point sources is getting entirely inadequate, when you want to manoeuver e.g. among the Jovian moons in the huge and complicated field of (non-spherical) Jupiter...
[/quote]

Thank for your answer

Are you sure that the gravitational point sources approximation introduces too much inaccuracy ? Wouldn't it be accurate enough to see how the system evolves with this laws (and if it remains stable) or to send space probes ? When Voyager probes trajectory were calculated, this approximation was not used ?

EDIT : I wrote the above "guest" message, how can i delete it ?

Celestia uses formulas with precalculated terms to determine the positions of all of the planets and most of the major moons. For the planets, it uses a set of Chebychev polynomials known collectively as "the VSOP87 theory." The polynomial terms were determined by someone actually doing all the gravitational calculations. Evaluating the polynomials (each of which includes more than a thousand terms) is much faster than doing the gravitational calculations to comparable accuracy.

Guest postings cannot be edited or deleted except by the forum manager, who generally has more important things to do. Sorry.

selden wrote:Celestia uses formulas with precalculated terms to determine the positions of all of the planets and most of the major moons. For the planets, it uses a set of Chebychev polynomials known collectively as "the VSOP87 theory." The polynomial terms were determined by someone actually doing all the gravitational calculations. Evaluating the polynomials (each of which includes more than a thousand terms) is much faster than doing the gravitational calculations to comparable accuracy.

Guest postings cannot be edited or deleted except by the forum manager, who generally has more important things to do. Sorry.

Do you mean that those analytic equations gives you the position of a body given the date parameter ??? And that it internally takes into account every other body influence ?? Or does it only take the sun into account and neglect other influences ? It's hard to imagine that a simple equation (even if it's a 1000th order polynomial) can modelise the positions of a whole set of interdepending bodies !!!

Yes, that's exactly what I mean.

All the relevant gravitational effects were numerically integrated for each of the planets. VSOP87 consists of polynomials which were fit to JPL's DE200 numerical integration for 8 of the 9 planets, as well as for the Earth-Moon barycenter. A separate ephemeris is available for Pluto. The results are supposed to be better than 1 second of arc for the period from about 1000BC to 3000AD.

VSOP87 is the ephemeris used by most "consumer" planitarium software packages. More accurate integrations and ephemerides have been computed since VSOP87 was released and are used by JPL's Horizons ephemeris service. Chris has mentioned that he's considering upgrading Celestia to one of those at some point.


Ref:

Bretagnon P., Francou G., : 1988, Astron. Astrophys., 202, 309.

Hi all, and sorry to bump an old thread, but I have what I think is a related question.

There was an old Mac program out there in the mid 90's called Gravitation Ltd that was a fairly basic approximation of gravitational interactions. In short, you could set up your own solar system, with masses and initial velocities, then let it go and see how objects collide, slingshot, orbit, etc. I found it a great way to visualise Keplers Laws in action, and it was a lot of fun to play with.

Does anyone know of a program out there these days which does a similar thing, but is a little more modern?

There's Gravity Simulator by Tony Dunn, but it may be only for Windows.

See http://www.orbitsimulator.com/

drewbert wrote:Does anyone know of a program out there these days which does a similar thing, but is a little more modern?

One of the guys here - tony(somethingorother) - has a program called Gravity Simulator. Unfortunately the user interface is rather nightmarish. I think this is it:

http://www.orbitsimulator.com/gravity/a ... /what.html