Some problems about the grid point of pot.cube file in the EXTERNAL_POTENTIAL subsection
Tianshu Jiang in Beijing
jts2t... at gmail.com
Sat Dec 2 12:52:17 UTC 2017
Hi, Nico, sorry to bother you again.
I have checked the PW grid as you have suggested, and in my simulation the
finest grid has 108 grid points.
PW_GRID| Information for grid
number 1
PW_GRID| Grid distributed over 12
processors
PW_GRID| Real space group dimensions
12 1
PW_GRID| the grid is
blocked: NO
PW_GRID| Cutoff [a.u.]
100.0
PW_GRID| spherical
cutoff: NO
PW_GRID| Bounds 1 -54 53
Points: 108
PW_GRID| Bounds 2 -54 53
Points: 108
PW_GRID| Bounds 3 -54 53
Points: 108
PW_GRID| Volume element (a.u.^3) 0.1025E-01 Volume (a.u.^3)
12909.5421
PW_GRID| Grid span
FULLSPACE
PW_GRID| Distribution Average
Max Min
PW_GRID| G-Vectors 104976.0 104976
104976
PW_GRID| G-Rays 972.0
972 972
PW_GRID| Real Space Points 104976.0 104976
104976
And the simulation box is a cube with length 12.4138.
&FORCE_EVAL
METHOD QS
&SUBSYS
&CELL
ABC 12.4138 12.4138 12.4138
&END CELL
The next thing I need to do is to distribute 108^3 grid points in the
simulation box uniformly. And specify the potential atoms can feel at each
grid point.
Is it right ?
Expecting your reply!
在 2017年12月1日星期五 UTC+8下午6:31:14，Nico Holmberg写道：
>
> Hi,
>
> You need to supply a cube file that contains as many grid points as the
> grid that holds the real space electron density in your simulation system
> (depends on the PW_CUTOFF and cell size). You can check the size of this
> grid by turning on the print key FORCE_EVAL/PRINT/GRID_INFORMATION (see
> example below). When you print out a cube file, unless you explicitly
> specify otherwise, only every second value along each axis is printed out
> (keyword STRIDE). You need to take this into account when you supply a file
> for READ_FROM_CUBE, because as far as I understand no
> interpolation/extrapolation from tighter/looser grids is possible.
>
> An example output is given below. In this case, the grid contains 350^3
> points at a planewave cutoff of 500 Ry. So to summarize, if your system
> contains 54 data points along each direction (at STRIDE 1), then you need
> to provide a cube with 54^3 grid points. You need a larger cube file with
> (2*54)^3 grid points if you printed out the cube file with STRIDE 2.
>
> PW_GRID| Information for grid
> number 1
> PW_GRID| Grid distributed over 192
> processors
> PW_GRID| Real space group dimensions
> 192 1
> PW_GRID| the grid is
> blocked: NO
> PW_GRID| Cutoff [a.u.]
> 250.0
> PW_GRID| spherical
> cutoff: NO
> PW_GRID| Bounds 1 -175 174
> Points: 350
> PW_GRID| Bounds 2 -175 174
> Points: 350
> PW_GRID| Bounds 3 -175 174
> Points: 350
> PW_GRID| Volume element (a.u.^3) 0.2459E-02 Volume (a.u.^3)
> 105442.7279
> PW_GRID| Grid span
> FULLSPACE
> PW_GRID| Distribution Average
> Max Min
> PW_GRID| G-Vectors 223307.3 223650
> 223300
> PW_GRID| G-Rays 638.0
> 639 638
> PW_GRID| Real Space Points 223307.3 245000
> 122500
>
> RS_GRID| Information for grid
> number 1
> RS_GRID| Bounds 1 -175 174
> Points: 350
> RS_GRID| Bounds 2 -175 174
> Points: 350
> RS_GRID| Bounds 3 -175 174
> Points: 350
> RS_GRID| Real space distribution over 4
> groups
> RS_GRID| Real space distribution along
> direction 1
> RS_GRID| Border
> size 29
> RS_GRID| Real space distribution over 6
> groups
> RS_GRID| Real space distribution along
> direction 2
> RS_GRID| Border
> size 29
> RS_GRID| Real space distribution over 8
> groups
> RS_GRID| Real space distribution along
> direction 3
> RS_GRID| Border
> size 29
> RS_GRID| Distribution Average
> Max Min
> RS_GRID| Planes 145.5
> 146 145
> RS_GRID| Distribution Average
> Max Min
> RS_GRID| Planes 116.3
> 117 116
> RS_GRID| Distribution Average
> Max Min
> RS_GRID| Planes 101.8
> 102 101
>
> BR,
>
> Nico
>
>
> perjantai 1. joulukuuta 2017 11.19.13 UTC+2 jts2... at gmail.com kirjoitti:
>>
>> Hi everyone,
>>
>> I have asked several questions about the same subject. Until yesterday, I
>> have figured out the meaning of the *.cube file generated by CP2K.
>> What I want to do is to add potential to my simulation system, but the
>> potential is calculated from Poisson equation using finite element method.
>> Hence I only have grid point values of the potential.
>>
>> So the solution is adding the potential from pot.cube file using
>> READ_FROM_CUBE option. I will try this in a simple simulation to figure the
>> structure.
>> The simulation box in my simulation is a cube, whose length of a side is
>> 12.4138 (Angstrom, I think). And from the *.cube file I know the length
>> of the
>> vector along x, y and z direction is 0.434420(Angstrom, maybe). What's
>> more, the number of data points along each direction is 54.
>>
>> Now I need to supply the pot.cube file to CP2K. I wonder how many grid
>> point value should be given?
>> (54*12.4138 / 0.434420) ^3 ?
>>
>> I major in mathematic, so the easy thing for you may be a problem for me.
>>
>> Thanks for your reply !
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20171202/4a5dc759/attachment.htm>
More information about the CP2K-user
mailing list