Per atom temperature

Dear all:

I am trying to obtain temperature distribution of my system (actually temperature/atom).

According to the previous discussions in the mail list, I found that researchers are using different ways around this issue:

1- Defining a constant parameter to the “compute ke/atom”;

2- Using “compute temp/chunk” command;

3- Using “compute temp/profile” command.

Nevertheless, I cannot manage to find the proper procedure. I would be thankful to receive your guides.

Many thanks,


If you want just the temperature/atom, you can do a compute ke/atom and then divide out the appropriate degrees of freedom in a per-atom variable, and dump this to a dump file. Then you can post-process this dump file to plot the info.

If you want a temperature profile, you’ll need to use chunks.