I'm new to simulation. Currently I'm using Material Studio to calculate solubility parameters and study the dispersion properties of nanofiller in polymers. I usually build up an amorphous cell (AC) with 2000 atoms with 5 output frames. I first do geometric optimization, then do a series of dynamic, finally calculate the bulk properties from the relaxed cells. The problem is I rarely get all the frams in AC converged after it is built. No matter how many steps of geometric optimization I do, what method I use, it just doesn't converge. I wonder if anyone know how to fix it. If I use a non-converged cell to do dynamic, it is possible to get all the frames converged enventrally? Any comments would help. Thank you.