Atom kinetic energy too high (for radiation damage)

LAMMPS LAMMPS General Discussion
1

Dear all,
I simulated the radiation damage in CuZr metallic glass with 10 Kev energy. (interatomic potential: mendelev 2007)
I used the hybrid/overlay eam/fs zbl with different inner and outer parameters for zbl but in some loops, I got this error:

“Atom kinetic energy too high for fix electron/stopping (…/fix_electron_stopping.cpp:178)”

inner and outer for example:
0.4 - 2.0
0.6-2.0
0.3-3.0
0.6-2.5

The loop was set to 50 and at the end of the simulation, I got just 15 or 20 completed simulations.

My energy table’s range is between 10 eV - 10 MeV.
How can I fix this error?
Thanks for your help.

2

Please attach all files required to replicate the issue.

4

I sent the files you needed.

5

The kinetic energy of one of your atoms is higher than the maximum energy supported by the Cu64_Zr36_table.txt file. So you either need to extend your table to support higher energies or need to set up your simulation so that such high kinetic energies do not occur.

6

Thanks Axel.

To set up my simulation which parameters should be changed ?? Can you help me, please?

7

Sorry, this is not my area of research. How to plan and construct your simulations is something you need to discuss with your adviser/mentor/tutor/etc. i.e. somebody that knows and cares about your research.

All I can offer you is the suggestion to do a test run without fix electron/stopping and instead add compute ke/atom and then use a custom style dump to monitor which energy individual atoms have. With the proper visualization/analysis tools, you can observe where the issues arise and then consider what to change.

8

Thank you, professor.

9

These are not all files needed to reproduce the issue - the datafile is missing. I also missed that the error occurs only after ~20 runs, so I wouldn’t be able to reproduce it fast anyway.

I would advise you to look at the log files, especially the kinetic and potential energy, to see if there are any inexplicable jumps.
You don’t have neigh_modify command in your scripts, so the default values are used. This means that the neighbour list is rebuild at most each 10 timesteps, which may be too infrequent for PKA simulations. Try adding neigh_modify every 1 check yes to your script, but first check the meaning of the line in the manual. If this does not solve your problem I think you need to find some local help.