Solution to periodic boundary problem
 

Recently i am working on a geometry which requires rotational periodic boundaries. I have encountered some problems and hopefully this will help those who have similar problems.

1. After I have done the mesh which is a combination of structured and unstructured mesh. The geometry is split into several different bodies to have better control of the mesh.The structured mesh was done using 'Multizone' and unstructured mesh for complicated regions was done using tetrahedron. I then exported the mesh and read it into Fluent. When I tried to make periodic boundaries. Fluent give me an error: Failed to make zones periodic. Error: #f. In my case, I think it means that those two periodic zones do not have the identical mesh.

2. Then i went back to Ansys Workbench mesher. Tried to apply 'Match Control' on the faces in the unstructured mesh region. It failed. The problem is I have to create a new cylindrical 'Coordinate System' becasue the axis of rotation for my geometry is not 'Z' axis. For the new coordinate system, I have to set the principal axis is 'Z' which is defined by the global X axis. And more importantly to set the orgin to (X=0, Y=0, Z=0). Otherwise it won't work. Then set up the 'Match Control' using this new coordinate system.

3. After the mesh is done with the modifications, the mesh export was failed giving the error: The meshing translation to fluent failed due to non-matching periodic nodes. The solution to this is to create a 'Symmetry' folder under the 'Model' tree and then insert 'Periodic Region' and select two set of periodic zones. The reason for this is because I have split the geometry into many bodies and one periodic zone is consisted of several small faces. I need to generate the mesh again after this modification and the mesh export was successful. When the mesh has been read into Fluent. The periodic boundary is already there and no need to set it up in Fluent.

Hope this will help and save some time for people who are having similar problems.

well done, in addition to that failed to make periodic zone when the rotation axis direction is wrong in the cell zone. hope it will help