CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Blocking Philosophy Query (http://www.cfd-online.com/Forums/ansys-meshing/87971-blocking-philosophy-query.html)

aerospaceman May 4, 2011 07:36

Blocking Philosophy Query
 
Hi,

I'm really new to Icem, but I've done loads of work with Gridgen/Pointwise and also Star CCM+.

I'm thoroughly confused about the blocking in Icem.

In Gridgen, from your imported CAD you create mesh lines called connectors; then you distribute the desired number of points as required. From there you go on to making surfaces and consequently volumes.

With Star, you define a volume and stuff cells in there (obviously more complex as there is loads of treatments but essencially this is it).

Now with Icem I don't understand. You import your cad. Then you create create a block which is totally generic. Then you proceed to associate edges to curves, then change vertex locations....(from my experience with the 2D pipe tutorial)

I must be definatively getting it wrong, because this definatively seems like a very long-winded and inefficient method.

Could someone explain what the purpose behind blocking is? I've tried reading the manual, but it doesn't explain, only tells to you to do it.

I greatly appreciate and help.

BrolY May 4, 2011 08:55

The blocking is like rebuilding your geometry by splitting blocks and associating edges to your curves. Most of the time, it's better to think about your strategy first, then do the blocking. There are tools such as the O-grid, extrude face, merge mesh etc.. which can help you for complicated geometry.

Post some pictures or .tin fil on the forum if you need some help ;)

saisanthoshm88 May 4, 2011 09:19

Gridgen follows a bottom up approach , like you described you build the mesh connectors, surfaces and then volumes. ICEM rather has a top down approach where you build a block and then divide it into segments to capture the geometric feautures.

Association is meant to get the block topology to respect the geometric feautures

For instance when you associate a face to a surface you tend to say that the mesh on that face is mapped to the chosen surface and when you associate the edge to a curve you mean to respect the curve while meshing

aerospaceman June 15, 2011 07:09

Thanks a lot for the reply. But surely this is much more difficult and counter intuitive than the so called "bottom up" approach.

I struggle to find a situation when this is an advantage.

Certainly for external aerodynamics this is surely the long-winded route...

But I'm happy to be wrong! Thanks!







Quote:

Originally Posted by saisanthoshm88 (Post 306235)
Gridgen follows a bottom up approach , like you described you build the mesh connectors, surfaces and then volumes. ICEM rather has a top down approach where you build a block and then divide it into segments to capture the geometric feautures.

Association is meant to get the block topology to respect the geometric feautures

For instance when you associate a face to a surface you tend to say that the mesh on that face is mapped to the chosen surface and when you associate the edge to a curve you mean to respect the curve while meshing


Jonathan June 15, 2011 11:46

hi,

i am pretty sure the foremost reason for the blocking approach in ICEM is to allow users to build fully structured meshes over reasonably complex geometries which might otherwise be difficult to do this to - basically, instead of decomposing your geometry physically in your cad software, ICEM allows to to import it whole and then fit premeshed blocks of structured mesh to the geometry (ICEM hexa), and therefore realise the advantages of the structured hexa mesh over unstructured tet / poly meshes ...

obviously, ICEM tetra lets you follow the bottom up approach using a variety of schemes to produce unstructred and hybrid meshes as well ...

jon


All times are GMT -4. The time now is 20:37.