Replication command

Dear Lammps users,

Please I have a question.

I used a CIF file for generating a unit cell by VESTA and then used “replicate” command to create my supercell, however when I ran my system for I always get (-nan ) as an output for all my parameters and when I replicated the unit cell by packmol I started to get results.

Is there any reason for that?

Please find the attached cif,data and log files.

Thanks in advance

crystal.cif (2.85 KB)

data.crystal (2.34 KB)

log.lammps (7.36 KB)

Apologies, please find the correct log file

Thanks

log.lammps (32.9 KB)

Dear Lammps users,

Please I have a question.
I used a CIF file for generating a unit cell by VESTA and then used
"replicate" command to create my supercell, however when I ran my system for
I always get (-nan ) as an output for all my parameters and when I
replicated the unit cell by packmol I started to get results.

Is there any reason for that?

LAMMPS strictly follows the GI-GO principle (garbage in, garbage out).
so your input must be wrong, when you get NaNs. it is not a LAMMPS
problem, but how you generate that input.

axel.

Thank you Axel for your reply, so does that mean I used replicate command in a wrong way as I get answers with another data files?

Thank you Axel for your reply, so does that mean I used replicate command in
a wrong way as I get answers with another data files?

no. you cannot use the replicate command in the wrong way.

axel.

Is there any way to detect where is my fault? I am pretty sure it is in the data file but how can I detect it?

Is there any way to detect where is my fault? I am pretty sure it is in
the data file but how can I detect it?

NaN energies are usually obtained when certain atoms overlap. You should
check your data file for this.

Of course, you wil get these NaN values with or without replication.

Kristof

Kristof, thank you very much, this was really helpful for me as I am new to LAMMPS and this was new info to me.

Thomas, thank you for that it is really narrowing my search to how to solve this problem but may I ask please some clarification for what you mean by it starts with atoms 1+2 and goes on with 3+4. I am sorry for asking so much but I am just trying to get the point and where to search for a solution.

I am appreciating your help…

Regards

2017-07-14 14:08 GMT+02:00 Meral Sharkas via lammps-users <[email protected]>:

Is there any way to detect where is my fault? I am pretty sure it is in the data file but how can I detect it?

NaN energies are usually obtained when certain atoms overlap. You should check your data file for this.

Of course, you wil get these NaN values with or without replication.

Kristof

Hi Thomas,

Apologies, I got you now, atoms 1,2 and 3,4 overlapped

I guess I should use delete atoms command to avoid that…

2017-07-14 16:30 GMT+02:00 Meral Sharkas <[email protected]...>:

Hi Thomas,

Apologies, I got you now, atoms 1,2 and 3,4 overlapped

I guess I should use delete atoms command to avoid that..

Or just, you know, make a data file without overlapping atoms...

Kristof

Dear Kristof,

It seems I am missing something. My original data file has no overlapping atoms however the dump file is showing overlapping and Thomas mentioned that this is because of the periodic boundary conditions, so what am I missing in this situation plz?

Thank you all for your great help…

Dear Kristof,

It seems I am missing something. My original data file has no overlapping
atoms however the dump file is showing overlapping and Thomas mentioned that
this is because of the periodic boundary conditions, so what am I missing in
this situation plz?

how about having too small a box, so that atoms overlap due to PBC?

axel.

No, it is just the same atom - translated by one box length - defined again.

Your conversion is wrong here, because those atoms should be in the box only once. Guess it happens for all atoms with one of the x,y,z components being zero in your file.

That messes up your cell composition by the way. Please check how many atoms you would expect in your cell.