I want to use two phase method to predict the melting point. Here I construct one solid, divide the sample into two parts (solid and liquid). For one part I just raise the temperature to be very high to let it be liquid, and let the temperature of the whole system to be around Tm (around 2500 K). But when I raise the temperature of the solid part to be Tm, the solid part becomes amorphous. Is there any way to avoid the solid part becomes amorphous? Now my way is to create velocities to be Tm and fix npt. I also tried fix nvt and fix langevin, and they also cause amorphous.
Lisa
I want to use two phase method to predict the melting point. Here I
construct one solid, divide the sample into two parts (solid and liquid).
For one part I just raise the temperature to be very high to let it be
liquid, and let the temperature of the whole system to be around Tm (around
2500 K). But when I raise the temperature of the solid part to be Tm, the
solid part becomes amorphous. Is there any way to avoid the solid part
becomes amorphous? Now my way is to create velocities to be Tm and fix npt.
I also tried fix nvt and fix langevin, and they also cause amorphous.
well, you have to accept what your model predicts, even if you don't
like it, assuming that your parameters and other input are correct.
the important point is to understand _why_ you get the behavior,
not to enforce something else, only because you want it like that.
you should also check, if there is any evidence for how your
system should behave near the melting point.
cheers,
axel.
It seems very strange that even if I raise up the temperature to be 300 K, the struture becomes some kind of amorphous from the RDF, and when I create velocities to be 1000 K, it shows 500 K at the 0 timestep. I have checked that the potential energy and pressure computation is right for the potential.
2011/8/26 Axel Kohlmeyer <[email protected]>