|
[Sponsors] |
August 6, 2013, 13:02 |
3D free blocking problems
|
#1 |
New Member
Mathieu Trottier
Join Date: Aug 2013
Posts: 3
Rep Power: 13 |
I am trying to use 3D free blocks to reduce the number of cells around a square cylinder mesh. As the mesh will be used in DDES I need to be able to expand the size of my elements outside the fine zone in the wake of the cylinder without using an interface. Here’s a look at the blocking I use. The zone near the cylinder is an O-Grid and the zone corresponding to the wake of the cylinder is a high density cubic region. Both of these regions are structured meshes.
1.jpg When I try to build 3D free blocks, I start by changing all exterior blocks (all except the two structured zones) to type free. Note that the inside edges already have the correct dimension and because it’s a mapped blocking exterior edges have the same number of nodes. 2.jpg After I change all exterior free blocks faces to free with “Convert free block face to free”. 3.jpg While regenerating the pre-mesh, I receive a license error. Is this a feature that requires another license? 4.GIF The pre-mesh obtained after the regeneration seems to only contain one free block while others are empty. I exported the pre-mesh to mesh and then to fluent and there is no mesh in the empty zones. Also, depending on the regeneration I do, it is not always the same block that shows up! I would like to know if there is another way to do the mesh I want or how to solve my 3D block problem? Thanks in advance. 5.jpg |
|
August 7, 2013, 11:54 |
|
#3 |
New Member
Mathieu Trottier
Join Date: Aug 2013
Posts: 3
Rep Power: 13 |
The current 3D mesh that I use is a 2D extruded mesh shown below.
7.jpg 8.jpg The inside zone of the 2D mesh is a smoothed structured mesh and the outside zone is composed of free blocks in order to reduce the total number of cells. Since I want cubic cells in the area of interest which is around the square cylinder and in its wake, the spacing between layers needs to be very small. This results in cells with very high aspect ratio near the boundaries of the computational domain (very thin cells in the spanwise direction). With the current mesh used, the size of the cells in the outside zone (with 2D free blocks) expands in the x and y directions while remaining contant in the z direction. The goal is to be able to expand in all 3 directions. I thought this would be possible by changing the 3D blocks to free as it is in 2D. I also tried to use an unstructured mesh in the outside zone but I was not satisfied with the expansion ratio next to the structured zone (area of interest) and I was not able to obtain conformal matching between faces of the structured and unstructured meshes resulting in the need of using an interface. The aim is to study the wake of the square cylinder and so we would prefer to avoid using an interface which could deteriorate the informations in the wake. |
|
August 7, 2013, 12:53 |
|
#4 |
Super Moderator
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 29 |
Ok, i suggest:
- Blocking the hole geometry, and use "block refinement" to independently refine blocks - take a look a this here may be it is more suitable for you : http://www.cfd-online.com/Forums/ans...tml#post365865 |
|
August 8, 2013, 17:17 |
|
#5 |
New Member
Mathieu Trottier
Join Date: Aug 2013
Posts: 3
Rep Power: 13 |
I read the thread you suggested and found out about the merge node option which solved my need for a conformal interface. I reduced my mesh from 13.6M to 11M elements which is a really good gain considering that my internal mesh has 8M elements alone.
The only problem that remains is that I change from a cubic mesh (interior) to a tetra mesh (exterior). Therefore i have a sudden change in resolution. It is not critical because i only study the wake inside my cubic zone but it still creates more cells than needed. As suggested in my first post could it be possible to use free 3D blocking to create quad in the outside zone. It seems to work for only one block while the others remain blank. Also, last time I tried to use quad in 3D free blocking i received a liscense warning as you can see in my first post. Is it a method that requires another type of liscense? Thanks a lot for the help. It is really appreciated. |
|
August 9, 2013, 01:09 |
|
#6 |
Super Moderator
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 29 |
I am not very familiar with free blocks. i never had to use them actually. But i still think that what you need if "refine block" look for it under the blocking, it lets you refine a block independently from the others, have your tried ?
|
|
Tags |
blocks, cylinder, free, square |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[ICEM] Problems with coedge curves and surfaces | tommymoose | ANSYS Meshing & Geometry | 6 | December 1, 2020 12:12 |
Bubblze induced turbulence at a free surface ???? | thomas | Main CFD Forum | 6 | April 21, 2012 10:48 |
Free Air B.C problems | Riaan | FLUENT | 20 | January 29, 2005 16:52 |
free RANS code with free surface ability | Lou jing | Main CFD Forum | 1 | July 11, 2003 15:43 |
problems with free surface | seba | FLUENT | 1 | October 11, 2001 19:32 |