Repository navigation
Is the PML wave condition available in 1D3V geometry? #7022
|
Dear Warpx team, boundary.field_lo = pml and some guard cells with particles so reflecting particles does not mirror the wave back warpx.do_pml_in_domain = 1 but running it will crash without very meaningful diagnostic messages. If I remove the three lines about do_pml_in_domain/has_particles/j_damping, A possible walk around is to use 2D coordinate with nx=1, where the pml can work fine. Therefore, my question is that: is the PML possible to be implemented for 1D geometry? |
Replies: 1 comment 1 reply
|
Thanks for your question. (The Silver-Mueller boundary condition is less ideal for 2D and 3D simulations for waves in oblique incidence, and this is why one would use PML instead in those cases.) |
Hi @YifanWu-USTC
Thanks for your question.
You are correct that the PML is unavailable in 1D, but you can instead use the Silver-Mueller boundary condition, which should be near-optimal for 1D simulations:
(The Silver-Mueller boundary condition is less ideal for 2D and 3D simulations for waves in oblique incidence, and this is why one would use PML instead in those cases.)