Question about custom collision resolve function

[xczics]

Hi, I'm using rebound to simulation the collision stage during planet formation. And I want to record the relative velocities during the impacts.
I'm not sure when the rebound code (in python) call my custom collision resolve function, the simulation time is the collision time, or is the time before/after the step when collision happens.
I ask because I use the "line" method to detect collisions, and I'm not sure whether should I add extra correction to the relative velocities, or just take the difference between two particles.

[hannorein]

Hi Zicong,

REBOUND first integrates all particles forward to the end of the timestep. It then checks for collisions. In your case, because you are using the LINE algorithm, it will detect collisions that occurred during the last timestep. However, the positions and velocities that you see in your collision resolve function are those at the end of the timestep. If this is not good enough for your use case and you need the coordinates at the exact time of closest approach or first contact, then you'd need to calculate that yourself in the collision resolve function. The calculation would basically be identical to what the LINE algorithm uses itself. See: https://github.com/hannorein/rebound/blob/main/src/collision.c#L215

I hope this helps.
Hanno