Convective heat transfer in a 2D nanochanel

Hi Qiangqiang,

That’s a pretty nice and clean set-up. Remember to always include the list - they will probably be able to spot errors in your set-up faster than myself. My idea, when I responded to you last, was that you might be getting a large amount of viscous heating near your walls. To evaluatue this I suggest to simply do an analysis on your fluid.

a.) Look up closures for the thermal conductivity of a 2D Lennard Jones fluid in terms of your chunk averages, density, temp, etc.
b.) Evaluate your heating from your chunks in MD - (strain rate tensor):(stress tensor) + (\grad p)*(div v).
c.) Solve a model fluid system and see if this seems reasonable. (From this model, I expect your temperature to be fully 2D.)

It doesn’t look to me like you’ve made a LAMMPS error, what you expect might need to be adjusted.

Pay no attention to the micro-canonical language, you should be using NVE for this model. The time-step you could solve on your own.

obviously check your equations (p div v)*

Eric Murphy, PhD
Mechanical Engineer
murphyericjames@…24… | (563)449-6661