I’m studying flow in nanochannels. It’s similar to openchannel flow, a fixed wall reflect is used at the upper z boundary. The flow is along x axis. The width is along y axis. It’s basically a flow boiling setup.
Now, I’m using copper as substrate, using the ‘phantom heating’ method to boil the flowing argon. I’m applying force at the very starting region of the flow, and then using thermostat to reset the temperature of the flow, and then data collection region starts (periodic flow along x).
I’m surprised then even after increasing the applied force, which clearly substantially increases velocity, the heat flux is almost identical. The increasing velocity should also increase convection heat transfer, but it doesn’t.
I’m using these commands to calculate flux:
compute ke conducting_layer ke/atom
compute pe conducting_layer pe/atom
compute stress conducting_layer stress/atom NULL virial
compute heat_flux conducting_layer heat/flux ke pe stress
variable volm equal 353*32.5*16 # (353, 32.5 & 16 are lenght, width and depth of conducting layer only)
variable heat_flux equal c_heat_flux[3]*1.6*10^13/(${volm})
fix flux_vs_time conducting_layer ave/time 1 5000 5000 v_heat_flux file flux.txt
Here the substrate is divided into 3 layers:
- fixed layer at the bottom to support the structure
- phantom heating layer (used langevin to keep at constant temp)
- conducting_layer that transfers heat from phantom layer to argon
I’ve attached the velocity profiles and heat flux profiles (NOTE: The temperature profile is also almost identical when I plotted)
