Learning Forum

[Antang]

So I'm trying to model an ejector but I cannot seem to get a structured mesh in the region between the ejector (so between the two airfoil-like surfaces) mainly because the nose is basically a semicircle.  I've gotten it to look "decent" with a face sizing but I cannot control its first layer height because it is not a structured mesh.  Does anyone know how I can get this to be structured and with appropriate y+ value (<1).  Also if I left it how it is (with appropriate height on outer surface and not appropriate on inner surface), is it safe to say the result from the k-w SST results are bad even if I'm getting okay convergence?  Would it be more accurate with k-e wall function?

 

Using a classical C-mesh the domain looks like this but I divided it into different faces.  The area between the airfoils is not structured but rather just face sized. 

So the outside of the airfoil is good but the inside is not of an appropriate y+.

This flat surface between two curved surface is a velocity inlet

[peteroznewman]

If you offset the boundary of each "airfoil" an imprint that onto the fluid region so you have a "ring" around the boundary, you might succeed in getting an inflation layer to go all around the airfoil boundary.

[Rob]

I'd also look at mesh growth rate and aspect ratio: just because the cells are good quality doesn't mean the mesh is OK. 

As an aside, if this is an ejector, is it a 2d planar or would 2d axi-symmetric be more appropriate?