conductors have to be closed Surfaces. Is this also the case for dielectrics ?
Example:
I have 2 conductor strips (some distance apart)on a SiO2 layer. The dielectric layer is below botth stribs (e.g. z<0), the strips both in x-y-plane.
I create the conductor strips (Cuboids) by 3 Q panels with outperm 1.0 and 1 Q panel with outperm 3.9 each, join them in a List file.
Now i want to add the dielectric layer. Is it now enough to implement the layer as a few Q panels in the x-y-plane surounding the 2 Metal strips or do i have to make a Cuboid expanding in z-direction.
To be pricise:
Only panels in x-y-plane: ht*ps://imagizer.imageshack.com/img921/1974/bhCbVG.png
Cuboid as dielectric : ht*ps://imagizer.imageshack.com/img924/4240/YVMMrQ.png
In your case, both options may be good, if the dielectric surface is large and thick. In the case of a dielectric represented as a 2D object and not a 3D cuboid, it should be large enough to have a negligible effect on the fringing fields (in your pictures it seems large enough, but you can see the effect only by simulation).
Do Dielectrics have to be closed surfaces
Posted - Aug 12 2019 : 15:56:17
