I have been writing a MATLAB code to reproduce the results of Verlets paper - Computer "experiments" on classical fluids, 1967.

Here is how I go about it:

- I initialize all the velocities (uniform distribution about [-1, 1])

- Initialize all positions  (NOT random allocations).

- I use Verlet's algorithm to update the positions.

However, in the very second iteration (with a non-dimensional time step of 0.032 as suggested by the paper), My force goes to the order of 10^41! This has been eating me alive for the last 4 weeks!

Any ideas?

Aneet

dnaneet@mtu.edu