Dear lammps users,
hello,
I am doing a simulation of the thermal conductivity of silicon nitride, using the non-equilibrium method (nemd), and the potential function is the tersoff potential, and after running the program, I found that the calculated thermal conductivity is in the 600’s (along the y-axis) which is much higher than the theoretical thermal conductivity (400), and I’d like to ask for a look at where the error is.
Attached is my code：
si3n4.in (4.9 KB)
Not getting the expected result is not automatically a LAMMPS problem, but can be a user problem, e.g. unconverged data. There have been plenty of previous discussions here about how to determine settings that yield converged data. Please note that examples that ship with LAMMPS are not automatically a good choice for that, since those are usually using parameters that make those runs quick. Also, the convergence can depend a lot on the material and its properties.