No updates in log.file and dump files

Dear LAMMPS users,

I am working on simulating ultrashort laser-matter interaction through TTM/mod fix in LAMMPS.
I release the periodic boundary in z direction, by applying (m) boundary on the front surface and (f) boundary on the rear surface.
Laser beam is shooted on the front surface.
Everything works fine until certain timestep that there will be no updates in log. file and therefore no further dumps.
According to the running status, lammps is still running with no error message.
I check the latest dump file, according to the snapshot in atoms, it seems that part of the material gets totally separated from the domain.
I am wondering if the boundary will go wrong when some part of the material is broken away. Also, what may cause this situation that lammps seems running but gives no feedback in log and dump as set.
Any advice and comment will be highly appretiated.
Thanks!

Xiao Jia

Most likely the run is producing NaNs for the thermodynamic outputs due to a dramatic change in the structure and box dimension. If you have atoms breaking away from the surface and reaching the boundary, a “m” boundary keeps changing to adapt to the moving atoms. As a result, you may get an unrealistically large box. If you have a restart file and you can restart your simulation so that it “halts” again soon, with “thermo 1” you would probably see the NaNs. This is when the energy/force evaluations run into trouble and the program doesn’t know to stop.

Ray