Regarding TDamp for NVE using units real

Please post to the list, not to me.

sorry, I thought your units were metal, for some reason.
For a tilmestep of 0.15 fs, a Tdamp of 5 (fs) should be
reasonable. What does the monitored temp of the system
say it is? And note that you have frozen atoms,
so you should be monitoring the temp of the atoms
you are thermostatting, not the full system.


Thank you for you reply. I ran the simulation using temp command. Are you suggesting that I use the temp/region compute temp of atoms?

Now I have looked through the mailing list archives but I am confused. I am trying to pull the copper and graphene in the z direction using the-- fix deform vel --command. Now I want to obtain force vs displacement values. I have tried using

compute force_x all reduce sum fx
compute force_y all reduce sum fy
compute force_z all reduce sum fz

to get the forces in the x y and z directions. The force along the z direction starts off positive and goes negative. Can somebody please give me an idea as to what I might be doing wrong?


Thank you for you reply. I ran the simulation using temp command. Are you suggesting that I use the >temp/region compute temp of atoms?

If you just use thermo_style … temp …

you will get the temp of the entire system, which in your case

includes frozen atoms. So it isn’t the temp of your thermostatted

atoms. See the doc page for fix nvt which explains what compute

stores the temp that is being thermostatted (for a group), or

define your own compute group temp and output that with the

thermo. Otherwise you have no idea what the thermostat is

actually doing.

How do you know the forces you are outputting are wrong?

Try dumping the forces to a dump file and summing them

yourself. I imagine they will be the same as what your 3 computes

are giving. If you think it’s wrong, then figure out which atom(s)

forces are diffferent than what you expect.

A LAMMPS script is like a computer program. It it doesn’t

produce what you expect, you have to debug it.