Seed Number in Velocity Command

Hello,

I had a basic question related to seed number in the velocity command. If I use the same seed number in the velocity command for slightly different structures, for example:

  1. Increasing number of atoms of a certain type in a simulation which also consists of other types of atoms and their number remains the same
  2. Simulations of composite system in which polymers are different
  3. Simulations of different polymers

Will the simulations/results have any correlations in these cases? Or is it fine to use the same seed number for them?

Thank you

MD is a chaotic system, meaning small perturbations will cause exponential divergence over time. In other words, any small difference should lead to big differences after a few thousand timesteps. You should equilibrate your system and throw out the first part of the run anyway, so I don’t see a problem with using the same random seed to initialize velocities for different simulations.

1 Like

Thank you for the response. This answers my question.