[CPMD-list] On periodic boundary conditions

[Luca UvA]

Dear all,
            it is clear that the TRAJECTORY file does not apply PBC, for good reason (e.g. calculation of RMS displacement). I am working with an fcc cell, and recovering the PBC (e.g. for pair-correlation function), in NVT ensemble, is easy, applying an algorithm based on Allen-Tildesley one, found at CCP5 library. But, for NPT ensemble (since I am working on liquid phase I allow only isotropic deformations), to compare data I am trying to simulate still in a fcc cell, just setting the proper triclinic parameters (a 1. 1. 0.5 0.5 0.5). Clearly the cell is defined by only one parameter. Then, trying to recover the minimum image, still applying the former algorithm, I find only crazy results.
Any suggestion?

Luca Ghiringhelli.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cpmd.org/pipermail/cpmd-list/attachments/20020920/daf70d95/attachment.html