I am using lammps for md simulation. I am wondering if there is way for md simulation in Gibb's ensemble. One way lammps manual suggest is to use " fix gcmc command " which requires chemical potential to be known. I would be grateful if anyone could suggest me the way when chemical potential is not known. I am particularly interested in calculating vapor pressure .
You may find the fix NVT helpful to simulate a canonical ensemble http://lammps.sandia.gov/doc/fix_nh.html
You cannot, for Gibb's ensemble you need to exchange volume between 2 or >2 boxes, feature not included in LAMMPS as far as I know.
Check Towhee
Thank you for your response. Do compute stress/atom and fix ave/spatial commands give correct pressure profile in NVT ensemble? I got fluctuation of normal component of pressure tensor at the liquid-vapor interface which is supposed to be constant through the interface.
Hi,
In my experience, the global value of the stress tensor are reliable, they give elastic tensor coefficients equal to other codes. Thus, the per atom quantities should be reliable. However, keep in mind that the quality of the force field will be the key factor, and that local fluctuations may be quite large at a given atom and time step, so average properly will be important.
- Badges
- Science topic