Trace:

periodicity

This shows you the differences between two versions of the page.

— |
periodicity [2014/02/09 15:50] (current) oschuett converted from doc/tutorialCp2k.html, Credit: Fawzi Mohamed |
||
---|---|---|---|

Line 1: | Line 1: | ||

+ | <note warning>This text is probably out of date and needs to be revised.</note> | ||

+ | |||

+ | ====== Periodicity ====== | ||

+ | The Cp2k program is optimized for periodic calculations, and those are the default, still it is possible to perform also other kinds of calculations. First an important thing to note is that by convention the cell goes from 0 to h, not from -h/2 to h/2, thus if you want to put a cluster at the center of the cell you need to put it in h/2 not in 0. | ||

+ | |||

+ | The periodicity of the neighbor list, and thus of the gaussian collocation/integration is controlled by ''FORCE_EVAL%SUBSYS%CELL%PERIODIC''. | ||

+ | |||

+ | The poisson solver (electrostatics) is controlled in the section | ||

+ | ''FORCE_EVAL%DFT%POISSON''. The possible poisson solver are chosen with the ''POISSON_SOLVER'' keyword, that can have the values | ||

+ | |||

+ | * ''PERIODIC'', which can do only fully periodic solvers, | ||

+ | * ''ANALYTIC'' which can do 0, 1d and 2d periodic solutions using analytical green functions in the g space (slow convergence), | ||

+ | * ''MT'' Martyna Tuckermann decoupling, that interacts only with the nearest neighbor, beware these are completely wrong in the cell is smaller than twice the cluster size (with electronic density), | ||

+ | * ''MULTIPOLE'', uses a schema that fits the total charge with one gaussian per atom | ||

+ | |||

+ | The periodicity of the poisson solver (electrostatics) is set in ''FORCE_EVAL%DFT%POISSON%PERIODIC'', and the poisson solver chosen should support it. | ||

periodicity.txt ยท Last modified: 2014/02/09 15:50 by oschuett

Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-ShareAlike 4.0 International