Attached are two photos of the same model and settings.
Photo: results global view of mesh - no enhancement layers seen on impeller.png
Photo: mesh shown with added plane - enhancement layers shown on impeller.png
In my view, this global mesh view is buggy - requires some further analysis. Firstly, the global view might give the wrong impression (ie implies that there are no enhancement layers). In this sense, the view is wrong as the unstructured mesh could not 'reach' the surface boundary in this manner. What we should see are enhancement layers (rectangular elements) shown on the lower surface (the yellow portion). These rectangular enhancement layers are only present on the added plane, but should also be shown on the lower surface in the global mesh view.
Going deeper...the true geometry involves an intersection of enhancement layers as there are both enhancement layers on the lower yellow colored surface as well as the impeller blade. The only way in which this can be resolved is by the shape of the enhancement layers that intersect the two enhancement layers as being that of an extruded triangle, in which case the view could be correct as what we are seeing on the lower yellow surface are in fact that sketch outlines of enhancement layers - NOT the unstructured mesh. In this view of the geometry, the impeller blade surface enhancement layers have been applied after the lower (yellow colored) surface received it's enhancement layers - hence it does not 'intersect' the lower (yellow) surface and thus appears to be not enhanced.
However, if this is the case (ie that the lower yellow surface was enhanced before the impeller surface) then in fact we should see the tracings of the lower yellow surface's enhancement layers appear on the impeller surface just as they have appeared on the green surface?
So Royce...? (I discussed this with him prior ...)
This mesh building geometry with enhancement layers gets tricky in a real hurry with intersecting meshes...I'm thinking further on this and I find it hard to construct this geometry in my head while making sure nodes are common. (I'm guessing that mesh enhancement layers create many more odd sized mesh elements just to get around the complications caused by the intersection of enhancement meshes). I guess that's why we get 20 enhancement layers and Ansys gets 60 or so.
Hi Doug,
Is this geometry from ProE/CREO?
If not, try reducing your minumum refinement length in the CFD meshing diagnostics to the smallest value available in the slider and see if that helps generate the mesh enhancement as expected.
If it is let use know and I can provide some more details on another solution.