Simulate with complex geometries and complex physics
Illustration
illustrate the idea of boundary point activation
Suppose there is a local point configuration of boundary, free surface, and interior points.
We establish sidewise ghost points for free surface and boundary points. The ghost points mimic the same type of point as their origin.
For free surface points, we establish one additional ghost point in normal direction. The mimic regular interior points.
In order to judge activation of boundary points, we establish a local tetrahedralization around potentially activated boundary points.
We neglect tetras/triangles whose corners only touch boundary or free surface points (marked in red).
We measure the opening angle of the remaining regular tetras/triangles (marked in green) and give a nondimensionalized functional spanning: 1=full half sphere, -1=zero opening angle.
The user can give the number of ghost points to be used for regular walls (see FurtherOptions ).
If 0 is given, we speed up the computation, then the terehedrization looks like this:
The user can also give the number of ghost points to be used for free surface points (see FurtherOptions ).
If 0 is given also here, we speed up the computation even more. In this case, the tetrahedralization looks like this:
In order to judge activation of boundary points, we establish a local tetrahedralization around potentially activated boundary points.
We neglect tetras/triangles whose corners only touch boundary or free surface points (marked in red).
We measure the opening angle of the remaining regular tetras/triangles (marked in green) and give a nondimensionalized functional spanning: 1=full half sphere, -1=zero opening angle.
The user can give the number of ghost points to be used for regular walls (see FurtherOptions ).
If 0 is given, we speed up the computation, then the terehedrization looks like this:
The user can also give the number of ghost points to be used for free surface points (see FurtherOptions ).
If 0 is given also here, we speed up the computation even more. In this case, the tetrahedralization looks like this: