|
|
Orbit of a test particle around a central mass.
|
|
|
|
|
|
Can do this in two ways: Impose an external potential or put two particles in a box but fix one of them in place.
|
|
|
Central mass = 1.0 Solar masses
|
|
|
Semi major axis = 1.0 AU
|
|
|
Eccentricity = 0.99
|
|
|
|
|
|
Difference between the two cases is that the latter includes softening and a different time-step condition.
|
|
|
|
|
|
Here I show plots of the orbits, comparing to the analytic case.
|
|
|
|
|
|
Semi-major axis is fixed to be 1AU, and eccentricity varies.
|
|
|
|
|
|
Softening is 0.01 AU and 'eta' is 0.01
|
|
|
|
|
|
Evolved for 100 orbits
|
|
|
|
|
|
![orbit_plot_11](/uploads/4a0a68e5fa4a387ecfa2cf61476ed4c0/orbit_plot_11.png)
|
|
|
|
|
|
![orbit_plot_14](/uploads/29057b568f49fc5199048bd743db5573/orbit_plot_14.png)
|
|
|
|
|
|
If we reduce the semi-major axis for the highly eccentric orbit, then things go a bit crazy for the 'self-gravity' case.
|
|
|
|
|
|
Could be due to softening or the different time step condition.
|
|
|
|
|
|
![orbit_plot_16](/uploads/3defb52207e33275ca272404c6e49c04/orbit_plot_16.png)
|
|
|
|
|
|
Reducing the softening to 0.001 AU does not seem to fix the issue
|
|
|
|
|
|
![orbit_plot_17](/uploads/2b60bf26f29d45f0a3b08fea372468a0/orbit_plot_17.png)
|
|
|
Fails if softening is too large
|
|
|
![orbit_plot_52](/uploads/a7d74411dbc4cbba4c43a5079b9b7424/orbit_plot_52.png)
|
|
|
|
|
|
Reducing softening fixes the issue
|
|
|
![orbit_plot_53](/uploads/4c586c85292d7beb0f6088084f202d0a/orbit_plot_53.png)
|
|
|
|
|
|
How to define the error?
|
|
|
|
|
|
Run for 10 orbital times, find the distance between initial and final position, and see how this changes with 'eta'
|
|
|
|
|
|
![eta_test_plot](/uploads/d5d23c553f45cd21bfcbee79cbd9f8e9/eta_test_plot.png) |