During LAMMPS simulation, especially during first equilibration, temperature suddenly rises to given value in first step, then again drops to much lower value at next step, then then gradually attains equilibrium temperature at next steps. Why does this happen? LAMMPS implement temperature by assigning random velocities on a group of molecules, drawn from Maxwell-Boltzmann distribution at that given temperature. Can not all atoms be reassigned temperature in single-step equilibration? Or the temperature change is attended by uniform heat generation,uniform heat flux from boundary, or some other physical way?
If periodic boundary condition in three dimensions are assumed, then can the boundary be taken to be symmetry adiabat? can lumped capacitance method for temperature rise in the simulation can be obtained inside a periodic cell? (I obviously note there is no convection involved here).
Following is a simulation result based on temperature rise.
- Badges
- Science topic