How to add the thermophoresis term to the soot transport equation?

[grabtheNS]

Under normal gravity, the thermophoresis effect may be negligible compared to convection effects. However, under zero-gravity diffusion-dominated conditions, relevant reports suggest that neglecting thermophoresis could lead to errors of an order of magnitude for soot svf. Therefore, I want to simulate the soot distribution considering thermophoresis under zero gravity. Does anyone know how to do this?

[baperry2]

A simple approach would be to compute the thermophoresis source term and add it as a user-defined external source term using the ProblemSpecificFunctions::modify_ext_sources function defined in pelelmex_prob.H following the example here: https://github.com/AMReX-Combustion/PeleLMeX/blob/development/Exec/RegTests/EB_ODEQty/pelelmex_prob.H#L215. That would treat the thermophoresis source term with an effectively explicit Euler numerical approach, but let you get something pretty quickly. A more rigorous approach would be to dive into the source code in PeleLMeX_Diffusion.cpp and associated files and use AMReX linear operators to compute the necessary gradients and perhaps even solve the system implicitly. You could use the Soret diffusion implementation to guide your work.