[Wenqiang]

Hi,


Hope everything is good. Now I try to run a very simple test case with multizone.



I have attached the screenshot of the domain in this post. The domain is composed of two ducts, every duct has 6 surfaces: inlet and outlet, then the other 4 surfaces are all wall. What I did is to attach the inlet of zone 2 to the outlet of zone 1, and I can make sure that these two surface are exactly the same, they both have the same axial coordinates.


I generate the mesh in SU2 native format, and use markers like NZONE, IZONE as the tutorial online: https://su2code.github.io/docs_v7/Multizone/



I name the surface from 1 to 12. Face 1 is the inlet surface of zone 1, surface 2 is its outlet surface. Surface 3-6 are wall; Face 7 is the inlet surface of zone 2, Face 8 is the outlet surface of zone 2 and surface 9-12 are wall. In the .cfg file, I have following setup:


MULTIZONE = YES
MARKER_INTERNAL=(face2,face7)
MARKER_EULER= ( face3, face4, face5,face6,face9,face10,face11,face12 )


and for boundary conditions:
INLET_TYPE= TOTAL_CONDITIONS
MARKER_INLET= ( face1, 288.6, 102010.0, 1.0, 0.0, 0.0 )
MARKER_OUTLET= ( face8, 101300.0 )


This is a single physics simulation, both zones are stationary and there is special setup, so I guess I do not need any subcfg files.
I use "OUTER_ITER= 50" and try to run it, but it always diverges and crashes after about 10 iterations.



I do not know if my setup are suitable. I tested each single zone mesh and run it using exactly the same boundary condition, I can get reasonable results. So I think there is nothing wrong with the mesh. Could anyone tell me if I need to change anything in the .cfg file? I also attached it in this post (I rename it as two.txt, or it cannot be uploaded). Thank you.


kind regards,
Wenqiang

[pcg]

Hi,
You need to define the fluid interface between your zones using MARKER_ZONE_INTERFACE and MARKER_FLUID_INTERFACE.
MARKER_INTERNAL does not do anything, it is a "fictitious" boundary condition.
You can have a look at the examples in SU2/TestCases/sliding_interface.

[Wenqiang]

Originally Posted by pcg

Hi,
You need to define the fluid interface between your zones using MARKER_ZONE_INTERFACE and MARKER_FLUID_INTERFACE.
MARKER_INTERNAL does not do anything, it is a "fictitious" boundary condition.
You can have a look at the examples in SU2/TestCases/sliding_interface.

Thank you so much for your reply. I checked the sliding plane example as you suggested. Then I put these two lines in my .cfg file as well as "MULTIZONE_SOLVER=BLOCK_JACOBI". But even I tried that, it still diverges. The density becomes very large.



This time I generated a much smaller mesh as following picture. So it is much quicker to test it. I also attached my .cfg file again, named as "two.txt". I also uploaded the mesh file I used here:


https://drive.google.com/open?id=1mE...A5R8wQhZERg6CE


As I mentioned, this zone1+zone2 mesh just comes from zone1, and I tested running zone 1 alone, everything is fine. Zone 2 mesh is just a copy of zone1 but the coordinates has been moved along the x direction to match surface 2 and surface 7.


So is there still anything wrong with my setup or mesh? Thank you so much.