Anonymous user
Talk:N-body problem: Difference between revisions
→a simple case with an explicit solution
| (One intermediate revision by the same user not shown) | |||
Line 23:
Since this is just for a gravitational interaction simulation, collisions should not be worried about, right?
I propose we leave the method of numerical integration up to the user as long as their error is within a given amount.
== a simple case with an explicit solution ==
I think two separate problems are being discussed: an abstraction in which point masses interact gravitationally and non-relativistically, and a more concrete problem considering actual interactions between bodies in earth's solar system. I recommend that this task be split into two: 'n-body problem' which is the abstract one, and 'orbital mechanics' which is the concrete one. My remaining comments are with regard to the abstract one.
I recommend that the requirement is a solution to the n-body problem for an arbitrary number of bodies. The verification case should be the simplest case for which there is an explicit solution. The J solution is a candidate, a radially symmetrical system of three bodies orbiting a central point at constant velocity. Assuming that the agreed verification case is some sort of system with stable orbits, the time span should be at least one revolution of the system. Additional cases might be presented, at the discretion of the programmer.
| |||