I met some problems when using compute heat/flux in LAMMPS, version 30 Mar 2010.
First, when I run the example in its webpage, I found a mismatch of compute ID of ke and pe:
compute myKE all pe/atom
compute myPE all ke/atom
This will lead an error in the simulation.
Second, for this example, the integral I got was ~7e-11, same order but not exactly same value with what the authors provided.
I just used the input commands they showed at the bottom of the page. I wonder if there is anything wrong.
Also, for the “lamda”, it should be the thermal conductivity which the authors used kappa in the equations, right? Could the authors provide the number they used for “thermo” output frequency and timestep? For the current information, I can’t reproduce the result they got finally.
Mechanical Eng. & Material Sci. Department, MS 321
6100 Main Street, Houston