CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Relation and understanding of pre-mesh quality criterion "Angle" and "Min ortho" (https://www.cfd-online.com/Forums/ansys-meshing/194033-relation-understanding-pre-mesh-quality-criterion-angle-min-ortho.html)

 Mads October 9, 2017 09:01

Relation and understanding of pre-mesh quality criterion "Angle" and "Min ortho"

Hi.
The following is an introduction to my problem. The question starts after the "Now for the question(s):" section header.

Intro:
I'm doing my master thesis where I'm modeling a bioreactor using CFX and ICEM CFD. I have constructed a structured mesh for the bulk of the reactor fluid, and I'm in the process of checking the mesh quality. From what I have learned from my supervisors and this forum (e.g. this thread: https://www.cfd-online.com/Forums/an...eria-icem.html), the following pre-mesh quality criterions indicate a good mesh quality:

"Determinant 3x3x3" > 0.3
"Aspect ratio" < 1000 (When using double precision)
"Angle" >18deg

The extremes of these criterions for my pre-mesh have the following values:

min("Determinant 3x3x3") = 0.456
max("Aspect ratio") = 18.7
min("Angle") = 31.41

So far so good... However, from what I can read from the documentation on this page: https://www.sharcnet.ca/Software/Ans.../i1323480.html the official requirements for the CFX solver are related to the "mesh orthogonality". For instance, the criterion: "Orthogonality Angle Minimum" should be >10deg. Unfortunately, if I compute the "Min ortho" criterion, I get:

min("Min ortho")=4.06e-012

Now for the question(s):

-Have I misunderstood the documentation, and is the mesh quality actually still good?

It seems weird to me that two mesh quality criterions that are, as far as I understand, very closely related, can indicate so different mesh qualities. Secondly:

-What exactly is the orthogonality angle?

As far as I understand, the orthogonality angle(s) is 90deg minus the angle between the edge that connect two nodes and the surface normal vector(s) on the corresponding surfaces to the edge, i.e.

orthogonality angle = 90deg - angle(edge, surface normal)

For nice cubic elements, we should therefore get:

angle(edge, surface normal)=90deg => orthogonality angle=0

I would therefore think that a low orthogonality angle was desirable. However, the documentation states that the orthogonality angle should be >10deg, which really confuses me.

Phew, sorry for this long post.
I hope somebody can shed some light on this problem. :)

 All times are GMT -4. The time now is 11:57.