CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ANSYS Meshing] High skewness at the intersection of a circle and a square (https://www.cfd-online.com/Forums/ansys-meshing/218609-high-skewness-intersection-circle-square.html)

Danmei June 26, 2019 15:38

High skewness at the intersection of a circle and a square
 
4 Attachment(s)
Hi all,

I'm a first year master student and am still learning the software.

I'm trying to mesh a full-scale membrane bioreactor unit.
Geometry: 16 sheets of membranes (symmetrical so only 8 are shown) with a gap of only 1.5mm between each other (so b/w each sheet there will be 1.5mm of gap, so you can imagine the gap at the symmetry plane will be 0.75mm (very small). The geometry is large in the height and length (2.6+m) but small in the width ( each sheet is about 2ish mm in width). there is a diffuser (also symmetrical) below the membranes. Water(fluid) flows from the inlet at the bottom left, through the gaps, and to the outlet at the top right.


Challenge:
1. The geometry divergence massively in sizing.
2. I was told that the bottom of the diffuser (the cylinder at the bottom) should be aligned with the inlet. But this requirement gives me a "sharp angle" for meshing which will bring a high skewness.


What I've done:
1. I managed to mesh every part in Hex by slicing the geometry to make it sweepable. Edge sized + face mapping.
2. For the cylinder part, I tried to mesh the face on the side by splitting face(to make it face mappable) and giving edge sizing. It doesn't work. I believe it was because of the "sharp angle" existed.
3. I also tried to reduce the defeaturing tolerance (suggested on a post on research gate)
4. Multizone: yes, I can get the mesh but somehow the high skewness won't go away.
5. I also played with global sizing.
6. Texahedral in that specific section. Didn't work.
7. I also looked up skewness and did a testing myself that high skewness, even tho there is very few (couple hundred) can still affect your simulation effectively. I will have to simulate with multiphase and turbulence models in the future. So I really want to control the skewness quality before I proceed.

My question: 1. Should I move the diffuser up for like 2mm (or even furthur to reduce the skewness) just to make it possible for me to do the classic face meshing for circle in a square? 2. Any suggestions/tips/advice on how to control the skewness in a geometry like this? 3. Is there anything outrageously wrong I did or the track that I'm completely off...

I tried to look up the threads, but didn't find ones that can answer my question completely. If there is, please let me know and I apologize for repeating threads.

As I'm the only one in my lab studying CFD, there really is very little help I can reach out to.

I attached some scn shots. Thanks ahead for any help from you!

MIDHUN@CFD July 20, 2019 03:26

the problem occurs where the half circle meets the square plate above.. there make small modification like a cut which will avoid tangency of circle.. mail me the geom for further help
midhun6189@gmail.com


All times are GMT -4. The time now is 08:12.