. .
Here (Cut_Pipe&Vane) is a pipe with an internal vane. I want to analyse the flow through the vaned pipe. It is constructed as a pipe (Body_Pipe) from which a vane (Body_Vane) is cut out by means of a boolean operation (Part WB). The vane can be tilted at various angles by adapting the placement angle of Body_Vane. Here, a tilt angle of 5 degrees has been given.
In the CfdAnalysis we need two mesh refinements for the vane; i.e. one for its flat sides and one for its edges. In this model, the refinement was created in two different ways. One easy way, for which an unexpected trick appeared to be required. And a difficult one, for which such trick was not needed.
According to the easy, yet tricked method the two MeshRefinements were created by the following steps:
1/ To make the primary Body_Vane visible (space bar); (Note that the extends of Body_Vane beyond the diameter of Cut_Pipe&Vane have now become visible.)
2/ To start the task of creating a mesh refinement, followed by selecting and adding the appropriate faces of the primary Body_Vane in the 3D View. (Note that after the selection the refinement does NOT appear at the correct position).
3/ To additionally set the placement angle of the two MeshRefinements equal to the placement angle of the primary Body_Vane.
This operation gave these two mesh refinements of the attached model: MeshRefinement_VaneSides_(on Body_Vane) and MeshRefinement_VaneEdges_(on Body_Vane).
.
- View2.png (108.89 KiB) Viewed 3067 times
- View1.png (97.31 KiB) Viewed 3067 times
The difficult, untricked method, proceeds as follows:
1/ The primary Body_Vane remains invisible. In stead, the inside of Cut_Pipe&Vane is made accessible for face selection by placing a clipping plane longitudinally across its centre line (Menu/View/Clipping plane).
2/ The task of creating a mesh refinenment is started and followed by selecting and adding the appropriate faces of Cut_Pipe&Vane in the 3D View. (Combining a clipped view and carrying out other tasks is possible since FC19! Before that version, opening the clipping plane locked every other task. GREAT!) After the selection the refinement does actually appear at the correct position, so that further manipulation of the placement angles of the two MeshRefinements is not needed.
This operation gave these two mesh refinements: MeshRefinement_VaneSides_(on Cut_Pipe&Vane) and MeshRefinement_VaneEdges_(on Cut_Pipe&Vane).
I find this quite unexpected behaviour, and actually undesired.
1/ Unexpected, because the face selection was done on the basis of the already angled solid of Body_Vane. For some reason FreeCAD decides that the selected faces should be projected at an angle that does not exist in the model. That is to say, "does not exist in the model" as in, there is no way of finding that unangled face position displayable by investigating the primary bodies deeper inside the model tree.
2/ Undesired, because the first method is easy, and I like things easy. The required trick of manual adaptation of the mesh refinements can be a source of error. Further, this behaviour is certainly something that end-users should be aware of, to avoid mistakes.
I suppose that there are a number of other placement parameters that show the same behaviour. I did not investigate this.
PS:
After executing step 2 of the easy tricky method the Report says "Part::FeaturePython / MeshRefinement001: Links go out of the allowed scope". Well, FreeCAD is mistaken here. Cfd/OpenFOAM does not care and is perfectly happy with this messed meshing approach.
I am very curious about your view of the above experience. Greetz!
Roland
