Hello,
I cannot help on your post-processing of LAMMPS simulation. If you still
want to use GPUMD and you are not sure if something is supported or not,
you can state your scientific research problem in detail, otherwise it is
hard to give you any further suggestions.
Zheyong
ymyang_2018 <ymyang_2018@xxxxxxx> 于 2020年5月6日周三 06:17写道：
Thanks for your kind reply. I notice in GPUMD you modified the heat
current in many-body potentials like Tersoff in your paper 'Force and heat
current formulas for many-body potentials in molecular dynamics simulation
with applications to thermal conductivity calculations'. If I dump the
force (force_x) and velocity (V_x) along X-axis of each atom in my system
from LAMMPs after equilibrium. Then I want to count the heat current along
X-axis by each slab (here, system is sliced into 20 slabs along x-axis).
For example, in one slab, the force of each atom is total force in cutoff
region, first the force should be divide by the number of atom in cutoff
region. Second, I fellow the code in GPUMD/tool to calculate the force from
atom K to atom i, which is contributed force from "the third" atom. After
that force_x should multiple V_x, that is potential term only by X_axis.
However, the heat current I calculated is so larger than the all energy
multiple with V_x in slab. I don't know what's wrong with the code or I
cannot calculate by using force of each atom from LAMMPs.Thanks again.
Mingyang Yang
YMYang_2018@xxxxxxx
At 2020-05-06 00:01:35, "Bruce Fan" <brucenju@xxxxxxxxx> wrote:
The grouping method currently does not support dynamic partition. The
current code assumed that a grouping method is fixed during the simulation.
That is, if atom i is in group 1 in the beginning, it will be alway in
group 1. We can consider including this new feature in the future.
Zheyong
ymyang_2018 <ymyang_2018@xxxxxxx> 于 2020年5月5日周二 18:53写道：
Dear developers
Thanks for doing GPUMD. I don't understand the group methods in
xyz.in. The manual mentioned that in 'xyz.in' ,the data for GPUMD, there
is number_of_grouping_methods for grouping different atoms. If I want to
heat up a region in system specially dynamic region like LJ fluid. How to
write the number_of_grouping_methods in xyz.in. How do I choice
different methods, I guess there should be a selection for checking
different atom moving to this region and further reset group number. Also
I notice an example 'ensemble heat_nhc 300 100 10 1 6', the last two
parameters 1 6, like reduced units in LAMMPs, which is 1 of
number_of_grouping_methods.
The half system has been sliced into 6 parts, first one part is being
heated up. If the atom in first part moves to others, which group method
will allow the number_of_grouping_methods be changed as atom moving? If
I am wrong please help me, thanks again!
Mingyang Yang
YMYang_2018@xxxxxxx
