Add hybrid boundary condition in Custom Equation mode

Dear Precise Simulation,

I'm trying to solve Maxwell induction heating problems in 3D multidomain systems. I'm using custom equation mode with A-J formulation. The internal boundary conditions for the system are called "Hybrid". I need to set them to internal boundary surfaces between the air and a not perfect conductor (not infinite conductivity). The system offers only Dirichlet/Neumann conditions for internal and external boundaries.

Is there a way to set customized hybrid boundaries? I tried also to implement them as part of Dirichlet conditions (i.e. Ax=Ax+"HB" where HB are the hybrid boundaries=0).

I can maybe divide the domains and solve the weak formulation in two steps?

Best Regards,

eio

I'm not quite sure how you define hybrid boundaries? If you want a linear Robin type boundary condition you enter it like a Neumann type condition, and it should automatically be linearized as for example described here:

https://www.featool.com/tutorial/2018/02/12/featool-multiphysics-mixed-robin-boundary-conditions

Hi, thanks for your reply.

The boundary condition i would like to set is an internal interface between a conductive region and the air in induction heating problem. It has particular interface conditions. i.e. for perfect conductors it is n x (H-Hc)=J where n is the normal to the surface, H and Hc are the magnetic fields outside and inside the interface and J is the surface current density. This doesn't work for non perfect conductor. The interface condition for non perfect conductors involves the second derivative of the magnetic field H with respect to the normal direction and the skin depth parameter. This should be done in A-J formulation.

Here it is for some detailed explainations:
https://webhome.phy.duke.edu/~rgb/Class/phy319/phy319/node59.html

I'm asking if there is a way to set a custom interface condition (in this case for internal boundary) in custom equation mode.

Thanks :)

Not exactly sure how it would look like. Do you know how it would be implemented in a general FEM PDE code (in the framework of usual Dirichlet and Neumann BCs)?