Equilibration for Iron

Hi,

I am following the script for aluminum in tension and am replicating with iron. I have attached the script. The script is attempting to equilibrate to 300K, which works for aluminum, but the iron for some reason appears to be settling out at around 160 after 20000 steps.

I experimented with maybe minimizing the structure first. However, the minimization didn’t change much and, as far as I know, is not necessary for this basic unit cell. When I add in atomic hydrogen however I then use the above minimization.

Thoughts on why it won’t equilibrate? Should this effect stress strain values?

Thanks very much

ironpaperEAM.txt (2.99 KB)

[log.lammps temp|attachment](upload://r15ElO6Zo1Bw9Sf8iOIxR59jDL9.lammps tem) (11.6 KB)

Hi,

I am following the script for aluminum in tension and am replicating with iron. I have attached the script. The script is attempting to equilibrate to 300K, which works for aluminum, but the iron for some reason appears to be settling out at around 160 after 20000 steps.

​you are simulating for 2ps, that is not much time for equilibration. the comments in the script hint, that it was initially used with a larger time step. 20ps isn’t great either, but much better than 2ps. also, your temperature relaxation time constant is 1ps​, that doesn’t give the nose-hoover algorithm much time. for a more thorough equilibration, it is usually better to start with a fixed volume and use a dissipative thermostat first, so that you enforce energy equipartitioning and then switch to a nose-hoover thermostat.
you should read up on thermostats and their properties and equilibration in general in MD text books and discuss with your adviser. this forum is meant to discuss LAMMPS specific issues and this is not one.

I experimented with maybe minimizing the structure first.

​that makes no sense.​ if you start from an ideal lattice structure, your forces will be zero due to symmetry. besides, if you want to bring your system to equilibrium for a given temperature, minimization with have the opposite effect of quenching your system down to 0K. again, please note, that this is not a LAMMPS issue, but about simulation methodology in general, so you should seek advice from your supervisor/adviser/tutor/colleagues and text books.

​axel.​