How to generate an orthorhombic cell instead of a triclinic pattern?

[sfmnewbie]

Hello! Due to modeling needs, I need a reasonable orthogonal CSH model, but the model generated by pyCSH is a triclinic crystal model, I tried some orthogonalization methods, but the simulation still shows errors similar to Angle atoms, Bond atoms missing. Can there be a way to make pyCSH output the orthogonalized CSH model directly? Thank you very much!

[jlopez141]

We are aware that it might be interesting to provide orthogonalized simulation cells. But for now pyCSH does not have that capability. We might try to do it in future updates of the code, but right now it is impossible to get the orthogonal cell.

[sfmnewbie]

Thank you for your reply! Best wishes for further progress in your work.

[casarziga]

There are two 'easy' ways to make an ortho box. You can change the simulation box parameters (lammps data file) so that they represent an orthogonal box (e.g. you can delete the xy xz yz row). Then you can use either 1) LAMMPS and its remap command (if I remember correctly), or 2) Ovito where you can use 'wrap at periodic boundary conditions'.

Both will remap the atoms from outside of your ortho box into inside. It works, I did it before. However, you will probably need to redefine angles and bonds, so be careful. In any case, your simulated system needs to be of an appropriate size for this (at least 4x4 in a-b axis direction)

[sfmnewbie]

Yes, thanks a lot, that's very useful advice!