List close approaches to the Sun with the other close encounters

[AstroEloy]

Summary

The Sun has no Hill radius, so REBOUND.py tracked its distance but never listed it as a close encounter, and it could not be hit either. This adds both, with a small change.

• Close approaches. A perihelion passage closer than SUN_ENCOUNTER_AU = 0.1 AU goes into the same encounter list as the planets, in time order. The Sun gets a fixed distance threshold, not a fake Hill radius, so its entries have hill_radius_au and n_hill set to None. The terminal and report lines show the distance in solar radii instead of R_Hill, and the headers state both rules:

  Close encounters (< 3 Hill radii, Sun < 0.10 AU), in the order they happened:
  

• Impacts. SUN_RADIUS_AU (695,700 km) is added to the default body_radii, so an orbit that dives into the Sun ends in an impact.
• The WHFast warning counts Sun passages too, since WHFast degrades at small perihelia.
• The JSON gets sun_encounter_au next to n_hill_threshold. Readers of "encounters" must allow n_hill to be None, and the example in REBOUND.md is updated for this.

Tests

• Replaced the test asserting that the Sun is never an encounter with two new tests: a Sun passage at 0.05 AU is listed and one at 0.15 AU is not, and a perihelion inside the Sun is an impact.
• 84 passed on rebound 5.1.1 / reboundx 5.1.0.
• Mutation-checked: removing the Sun gate, the Sun radius, or the record builder's acceptance of the Sun each fails one of the new tests.
• Ran the CLI on the example trajectory (forward, with a temporarily wider threshold) and checked the terminal, report file and JSON output.

Notes

• The threshold is a module constant, like n_hill.
• A stream with perihelion below 0.1 AU gets one Sun line per perihelion passage. The δ Aquariids (q ≈ 0.07 AU) would get about 20 over 100 years; this is estimated from the orbit, not run.

🤖 Generated with Claude Code