mirror of https://gitlab.com/QEF/q-e.git
30 lines
1.2 KiB
Plaintext
30 lines
1.2 KiB
Plaintext
This example tests the use of images in ph.x. It computes the phonon
|
|
dispersions of AlAs as in example02 and the dispersions of Al as in example14.
|
|
|
|
NB: This example requires PARA_IMAGE_PREFIX and PARA_IMAGE_POSTFIX to be
|
|
set in environment_variables file. The image calculation must run with
|
|
a number of processors equal to the number of images multiplied by the number
|
|
of processors used in the pw.x run.
|
|
|
|
The calculation (run_example) proceeds as follows:
|
|
|
|
1) make a self-consistent calculation for AlAs without images
|
|
(input=alas.scf.in, output=alas.scf.out).
|
|
|
|
2) make a phonon dispersion calculation with images. Each image calculates
|
|
a part of the q points and of the irreducible representations.
|
|
(input=alas.ph.in, output=alas.ph.out, out_1_0, ..., out_N_1, N is
|
|
the number of images).
|
|
|
|
3) make a phonon dispersion calculation without images recovering the
|
|
information from outdir/_ph0/{prefix}.phsave. This is a fast calculation
|
|
where the dynamical matrices are collected, diagonalized, and written
|
|
on the working directory.
|
|
(input=alas.ph.in1, output=alas.ph.out1).
|
|
|
|
4) The rest proceeds as in example02.
|
|
|
|
|
|
Similarly for example14. The two phonon calculations are done using
|
|
the image features and then collected.
|