mesh-indep. study (!pics inside!)
Hello. I'm trying to mesh (ansys mesher) an axial turbine (a flow inside it) here:
(yellow = hub, red = blade, external cylinder side = shroud)
As far as i get it, i must perform a mesh-independence study to be sure the output is correct. When i make the mesh finer (increasing relevance), i got the following (presrise - output parameter - outlet to inlet pump pressure rise):
The mesh is created automatically with ASF "On Curvature"/"Prox. + curv." + tried adding inflation of blade,hub,shroud - similar results. Approx. number of elements: from 400k to 800k.
The question is: is the auto-generated mesh so bad (it doesn't "converge" to any result value (presrise), while i make it finer) , or i'm just doing it wrong? Thanks in advance.
Refining only a part of the mesh in a complex geometry or using the somewhat arbitrary "relevance" function in the Ansys mesher can not lead to the type of convergence you expect.
Usually when checking for mesh dependence, the variable is plotted against the cell size.
In general, with complex turbulent flows the convergence with cell size does not need to be monotonic.
When refining the mesh, details of the flow field resolved only on fine meshes affect the flow field at other places, leading to non-monotonic convergence.
|All times are GMT -4. The time now is 21:38.|