You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, GPUMD is great to use for studying thermal transport using ML potentials, but I was wondering if the flexibility of two commands in the code be improved. The fix command only allows 1 group to be fixed per simulation run, i.e
fix 1
fix 9
will result in only atoms in group 9 of grouping method 0 to be fixed when the run command is called. This can be inconvenient for NEMD simulations where the ends of the structure are fixed to prevent translation. The issue is bypassed be adding the atoms in group 1 to group 9 in the model.xyz file, but allowing the command to fix multiple groups of atoms like
fix 1 9
would be more convenient, in my opinion. Similarly I was wondering if the same could be done for compute_shc. I understand that there is an option for allowing the SHC to be computed for ALL groups in the grouping method when
compute_shc 1 500 1 500 200.0 group 1 -1
is used. but could the command also allow
compute_shc 1 500 1 500 200.0 group 1 1 2 3 ...
for however many groups the user would like to find SHC of.
Thanks again for all the developers' work on GPUMD, it is fantastic!
Mayur
The text was updated successfully, but these errors were encountered:
Thanks for the questions. These could be improved of course. We aim to improve the commands systematcally and hopefully make them available in the next main release GPUMD 4.0.
Hi, GPUMD is great to use for studying thermal transport using ML potentials, but I was wondering if the flexibility of two commands in the code be improved. The fix command only allows 1 group to be fixed per simulation run, i.e
will result in only atoms in group 9 of grouping method 0 to be fixed when the run command is called. This can be inconvenient for NEMD simulations where the ends of the structure are fixed to prevent translation. The issue is bypassed be adding the atoms in group 1 to group 9 in the model.xyz file, but allowing the command to fix multiple groups of atoms like
would be more convenient, in my opinion. Similarly I was wondering if the same could be done for compute_shc. I understand that there is an option for allowing the SHC to be computed for ALL groups in the grouping method when
is used. but could the command also allow
for however many groups the user would like to find SHC of.
Thanks again for all the developers' work on GPUMD, it is fantastic!
Mayur
The text was updated successfully, but these errors were encountered: