Changes in mesh type
 

Hello guys.

I wonder if changes in mesh type produce some numerical inaccuracy? For example this mesh (hope that the author will not have it against me ;)):
http://d2bm3ljpacyxu8.cloudfront.net.../screen002.jpg
Isn't it a "bad practice" to mix different kind of mesh types? On the other hand I heard about "Chimera grid".
What possible difficulties, problems, errors could produce such a methodology?

who is the author of this mesh :p

Having the mixed mesh is the next ambitios project at ANSYS. In this way you have more control on mesh that you are using high quality mesh (hexa) in critical region and tetra in less important regions. In this way you save large no of nodes in farfield.

But does such transition affect or not on the results?

You wrote that it's for using "high quality mesh (hexa) in critical region and tetra in less important regions" i'm planning to do exactly the opposite. I want to do some propeller flow analysis, but I don't know how to do structured mesh around it, so I'm thinking of creating small domain around my propeller and mesh it with unstructured mesh, and bigger domain, in which there is any complex geometry with structured mesh.

Therefore I'm wondering if it would affect the results in such way that they will be unacceptable.