[lammps-users] Computing heat flux using Fix Langevin

Dear LAMMPS users,

I have a system including two parallel solid walls that Argon gas atoms are confined between them. In this system, gas atoms can freely exchange energy with each other, and the wall’s atoms. I want to study the heat transfer between walls and gas. To do so, I duplicated the system in the Y-direction in order to have a symmetric simulation box (as it can be seen in the attached figure). I used fix Langevin to fix the temperature of the middle wall at 500 and the temperatures of the side walls at 300. I want to compute the heat flux between the walls while keeping their temperature fixed. As I have seen in the example provided in LAMMPS documentation, the heat added to the heat source and subtracted from the heat sink, supposed to have negative and positive signs, respectively. But as you can see, in the attached log-file in my case, they have both negative values. I highly appreciate if someone could tell me what I am doing wrong here.

###########################################Input file
region whole block 0 100 0 450 0 100 units box

create_box 2 whole

region Au_B block INF INF 10 19 INF INF units box #Bottom Wall

region Au_M block INF INF 220 233 INF INF units box #Middle walll

region Au_T block INF INF 432 442 INF INF units box #Middle walll

create_atoms 1 region Au_B

create_atoms 1 region Au_M

create_atoms 1 region Au_T

group AuBfix region Au_B_fix

group AuTfix region Au_T_fix

group AuMfix region Au_M_fix

group AuB_thermo subtract AuB AuBfix

group AuT_thermo subtract AuT AuTfix

group AuM_thermo subtract AuM AuMfix

timestep 0.001

#----------------------------------------------FORCE FIELD--------------------------------------------------------------
#-------------------------------------------------------Adding Ar atoms

You are starting to accumulate data before your system is equilibrated so the amount of energy required to thermalize your system taints your observations.

I would also like to caution that your studying a gas is raising concerns of applicability of the methodology (which has been devised for condensed matter) due to differences in the dominant process of transfer of momentum and the possible need to use significantly larger systems and simulation times to get properly converged results.