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/)
-   -   [ICEM] Meshing 3 holes geometry (https://www.cfd-online.com/Forums/ansys-meshing/216887-meshing-3-holes-geometry.html)

Pablotf12 April 23, 2019 06:40

Meshing 3 holes geometry
 
1 Attachment(s)
Hello Guys, im new to ANSYS ICEM, and im struggling wiht the meshing of a 3-holed Baffle , I have tried with an individual block to mesh the whole volumen, but i want to refine the mesh close to the holes edges, i have also tried with 4 blocks, 3 for the holes and 1 for the big tube, but i donr know if there would be any interference between mesh elements.

I attach the image of the baffle.

Please help and thanks for your time.

Pablo

bluebase April 29, 2019 15:06

2 Attachment(s)
Hi Pablo,


as i always write: do all the (blocking) tutorials which are provided for ICEM.
Your description of your solution attempts indicate possible misconceptions of how you you have to create blocking structures - or how this works in general. When you have done the tutorials, read the guide "Simons Tips & Tricks" to be found in the sticky thread of this sub-forum.



Anyhow, here would be my attempts to properly block your geometry. The major difficulty is the three-fold axisymmetry, while the enclosing cylinder calls for an o-grid. The center block would be triangular. But, this is not suitable for hexa blocking. Therefore, you need some kind of Y-block in the center. I see to path going from here. Using a "three" edged y-block or a hexagonal y-block. From there on, it's more or less filling the gaps. The circular holes need on o-grid inside and outside. The outside perimeter of these o-grids then have to fit in somehow.


I'd design the mesh in either of these following structures:
Attachment 69685Attachment 69684
I'd first try the hexagonal y-block center, because of a hunch that the resulting structure will yield higher orthogonality.


Best regards,
Sebastian

bluebase April 30, 2019 13:04

1 Attachment(s)
Playing a bit further on my ideas: rotating the main y-block by 60 degree - splitting the circles - could yield a good mesh too.

Attachment 69709

Pablotf12 May 1, 2019 06:33

Thank you guys, im trying now creating a surface mesh at the inlet of the tube and extrude it along the tube, but i have problems creating the volume inside the baffle holes, in fluent it creates me walls on the holes and in ICEM there are problems such as uncovered faces.

Its possible to do it using the extrusion method?

bluebase May 1, 2019 09:04

Yes, it is possible to use the extrusion method. It might be even the simplest.
Of course, you have to split the extrusion at the surfaces of the baffles, and delete all blocks which will be outside of the fluid domain. The topology of the "free cylinder" contains the topology (an additional o-grid) of the baffle's holes. Though, i only drew it once in each my figures.



the uncovered faces issue is a very regular one for beginners. You probably should search the forum for respective threads.


Anyhow, the warning really means what is says. There are (block's) faces which are uncovered (in the boundary setup). ICEM wants you to know, it needs boundary conditions for every boundary face. Failing to do so will result in an exported mesh which fluent (and others) will not accept. Very often there are edges left which are not associated sufficiently.

Pablotf12 May 17, 2019 12:49

Thank you very much for your help, sorry for the late message but i was too busy with my thesis, you have helped me a lot.

Kind regards.

Pablo


All times are GMT -4. The time now is 10:42.