CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Boundary condition that only allows a normal component, and no tangential component (https://www.cfd-online.com/Forums/openfoam-solving/255143-boundary-condition-only-allows-normal-component-no-tangential-component.html)

ErikNij March 21, 2024 09:31

Boundary condition that only allows a normal component, and no tangential component
 
Hi all,


I am trying to create an electromagentic solver. To do this, I am solving the following equation:


https://www.cfd-online.com/Forums/da...AASUVORK5CYII=


Where T is the electric vector potential. It is useful because if you take the curl(T) it will give you the induced eddy current J. This is not super relevant tho, because this is all working.



However, T comes with a rather odd boundary condition. Officially it is as follows: T the interface between conductors and non-conductors, the tangential component of the electric vector potential T is constrained to 0.



I've poured over the standard boundry conditions here. However, I am never quite able to find something that fits. At first glance, surfaceNormalFixedValue may seem promising, however, for that you must specify if you want in or out of the patch. You might also need to specify the magnitude, none of which I know. I need a boundary condition that just limits the direction of the vector, but lets the value be any scalar. Even if I need to write a python program to assign the normal vector for each surface cell / patch cell, I can do that. If there is b.c. that restricts the solution to a nonuniform list of restriction vectors, which would be the normal vectors in my case.

Or perhaps there is something else. In one way I am just looking for a noSlip b.c, that allows for a normal component as well. I am not super fa miler with the noSlip b.c, but perhaps this is a good starting point, however I don't think it will work without some modification.



I am interested to hear if anyone here has any creative solutions to my problem, or any experience here! If something is unclear, please drop a question and I can hopefully make it clear.



Best,


Erik


All times are GMT -4. The time now is 20:53.