Structure optimation by bop potential

Dear lammps users,

I am trying to optimize grain boundary structure of CdTe by bop potential, and i have an error.

In my simulation cell, I have 4 types of atoms, 1:Cd, 2:Te, 3:Cd, and 4:Te (see attached “gb.atomin” to this mail). The reason why I separate Cd and Te atoms into two types, 1/3, and 2/4, is that I want to freeze the type 3 and 4 atoms during optimization (not freezed yet in this simulation).

When I run the calculation, I got an error message like below:

segmentation fault (core dumped) lmp_daily -i gb.in > gb.out

I know this message can be output by many causes as indicated by the lammps mailing list archives, but I couldn’t find the cause.

To make sure my input file is correct, I deleted the types 3 and 4 atoms (see “gb-simple.atomin”), then calculated.

As a result, my code successfully ran.

Anyone can resolve my errors?

To easy follow, I attached these input files and potential files.

“gb-4types.atomin, gb-4types.in” : error occured

“gb-2types.atomin, gb-2types.in” : error didn’t occur

Version of my lammps is “Large-scale Atomic/Molecular Massively Parallel Simulator - 5 May 2020” installed by “sudo apt-get install lammps-daily” on Ubuntu 18.04.3 LTS.

Thanks in advance.

Shin

gb-2types.atomin (2.86 KB)

gb-2types.in (727 Bytes)

gb-4types.atomin (2.95 KB)

gb-4types.in (726 Bytes)

CdTe.bop (381 KB)

shin,

thanks for reporting and providing suitable test inputs.
i have been able to reproduce the issue and it looks like this is something deep inside the pair_bop.cpp code.
i will forward your e-mail with my comments to the authors of the pair style and hopefully they can find out what is causing it and provide a fix.

will report back to this thread as soon as we have something to test.

axel.

Dear Axel,

Thank you for your quick reply.

I’m looking forward to pair_bop.cpp code fixed.

Best,

Shin