used simple dirichlet boundaries, implemented interpolation at first to be consistent but there is some bug in it. Planning to make a separate source function for the potential equation, such that the implementation can be verified with an analytical solution for the 0 source term and just assuming 1 for magnetic moment and electron density (== heat equation with dirichlet boundaries, can do grid convergence study). Will make implementation more neat, then test with a nonlinear electron temperature function and find a better solution for the way electron pressure is calculated (have a separate array for Tev and p), probably not very efficient. After that, planning to study the Hara papers a bit more to actually understand the discretization and why he would calculate potential in the first place. And will implement boundary at anode using Bohm diffusion and recombination on surface, as we talked about on friday.
used simple dirichlet boundaries, implemented interpolation at first to be consistent but there is some bug in it. Planning to make a separate source function for the potential equation, such that the implementation can be verified with an analytical solution for the 0 source term and just assuming 1 for magnetic moment and electron density (== heat equation with dirichlet boundaries, can do grid convergence study). Will make implementation more neat, then test with a nonlinear electron temperature function and find a better solution for the way electron pressure is calculated (have a separate array for Tev and p), probably not very efficient. After that, planning to study the Hara papers a bit more to actually understand the discretization and why he would calculate potential in the first place. And will implement boundary at anode using Bohm diffusion and recombination on surface, as we talked about on friday.