1D Curve boundary conditions for Flow in Fractured Media

[SirGardener]

Dear Moose community and developers,

I hope you can help me out with this. What I am trying to accomplish is to reproduce the flow in fractured media example from the porous flow module (https://github.com/idaholab/moose/blob/next/modules/porous_flow/examples/flow_through_fractured_media/coarse_3D.i). However, I want to use an externally generated .msh File created in GMSH with the OCC Geometry Kernel. In the example, two intersecting fractures embedded within a cubic matrix are considered. These fractures, each box side and the total volume are assigned as physical groups in GMSH. Moreover, the intersection lines (1D) with the box sides are assigned as physical groups. By searching through the forum, I came across the hint that adding lower_dimensional_block in the .msh file in the respective lines under $PhysicalGroups fixed a first issue. This issue was that the fractures were not recognized in spite of having the correct ids. Afterward, the simulation successfully converged. However, inspection of the exported exodus II file revealed that lf1 and rf2 (the fracture intersections with the box side) were the entire left and right side of the cube instead of only the intersection lines. I am not sure whether this is a problem related to Moose or to GMSH. It might also be relevant to mention that while these lines were visible after creating the file in gmsh there are only points (no curves) visible for them when reopening the written .msh file with GMSHs fltk GUI. Does anybody have recommendations for me on what the best way would be to use mesh files generated in gmsh, e. g. in terms of file format? Is is perhaps necessary to combine the mesh with internal meshing capabilities?

Best Regards
Chris

[SirGardener]

I found out that is was just turner in the input file and now it works. Closed