Timeline for Simulating solar system with Newton's law
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 11, 2021 at 15:13 | comment | added | RBarryYoung | @LuisALberto But Mercury is not the biggest error, the Earth and Moon are, which is what really puzzles me about your question. Normally Mercury always has the biggest error (for multiple reasons) so if something else is the biggest (Earth-Moon) that eliminates most of the normal single-cause culprits (non-symplectic integration, bad G, bad mass of the Sun, etc.) What's distinct about the earth-moon system is the high angular tilt of the moon's orbit and the high mass of moon compared to it's primary (earth). My gut is that there's something wrong with the source data or the z-calcs. | |
| Dec 1, 2021 at 19:08 | comment | added | Luis ALberto | Dake, u are right the main error comes from walking in the wrong way. I'm again taking the data from Horizon files and error is related to distance from sun, (Mercury the biggest). So I think it's the mass, because it is not accurate. I will first simulate some mass suns, and take the best with min error to Horizon. Then go planet by planet doing the same. | |
| Nov 30, 2021 at 15:12 | comment | added | David Hammen | You don't need to use a symplectic integrator. JPL does not use one, for example. They use a variable step size, variable order Adams family integrator. Maybe Cowell or Gauss-Jackson? This are in the Adams family of numerical integrator, and both that position and velocity differently. | |
| Nov 30, 2021 at 2:01 | comment | added | Dale | I doubt that gmp will help. You will just have the inaccurate positions to high precision. | |
| Nov 30, 2021 at 0:56 | comment | added | Luis ALberto | Thanks, I tryed parsing Horizon files but got bigger errors!. Also its minimun time step is 0.5 seconds. Since some planets use the same code(in Don Cross source), and the ones with bigger errors have smaller constats, I will try with a precission library like gmp, but it is tedious. | |
| Nov 30, 2021 at 0:27 | comment | added | Dale | I would recommend using the JPL ephemerides for checking your computations: ssd.jpl.nasa.gov/horizons | |
| Nov 29, 2021 at 23:57 | comment | added | Luis ALberto | Thanks, but I used The Verlet code from wikipedia and got the same results | |
| Nov 29, 2021 at 13:09 | history | edited | Michael Seifert | CC BY-SA 4.0 |
Removed mobile links
|
| Nov 29, 2021 at 3:46 | history | answered | Dale | CC BY-SA 4.0 |