|
[Sponsors] |
[ICEM] Two boxes connected by a cylinder, blocks merge where I don't want them to merge. |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
April 29, 2014, 12:37 |
Two boxes connected by a cylinder, blocks merge where I don't want them to merge.
|
#1 |
New Member
Join Date: Jan 2014
Posts: 19
Rep Power: 12 |
Hello folks!
I'm currently looking into a flow case where the flow is in two large boxes with a small cylinder connecting them. I figured out that this is a great geometry to do hex meshing, with O-grid in the cylinder. However when I try to create three blocks for each separate part, I run into problem. First I create the block for the cylinder, associate the edges and then do and O-grid split, so far so good. Then I create a block for the first box and associates the edges, ICEM asks me if I want to Merge it which I want to. So now I have two blocks. I then associate the edges to the curves on the large box(see first picture). The problem arises when I add the third block to the other box. ICEM asks me if I want to Merge or Replace the blocks, I press Merge since Replace will remove the other two blocks. Then the vertices of the two big blocks for the two big boxes merge and thus creating something that doesn't look like the original geometry(see second picture). Note that I initialize the 3D blocking by choosing the appropriate surfaces. I must have surely overlooked something very simple but I can't figure it out. Thanks in advance Last edited by Polarbear; April 29, 2014 at 13:06. Reason: Spelling |
|
April 29, 2014, 13:52 |
|
#2 |
Senior Member
Join Date: Feb 2013
Location: Germany
Posts: 200
Rep Power: 24 |
First start with some basic tutorials on blocking as it seems you got the concept behind it wrong. Then start with one big block around the whole geometry and make splits, associations,... . This is called the top down approach is the most common way of blocking. Your blocking should be similar to the one in att2.
|
|
April 29, 2014, 15:20 |
|
#3 |
New Member
Join Date: Jan 2014
Posts: 19
Rep Power: 12 |
Holy smokes it's working! Thanks so much for your help, I think I understand more the idea of the blocking. One last question, I see that you have double O-grids in the boxes and just a single one in the cylinder? Whats the reason for that? Is it to be able to control the hex elements?
Rep to you! |
|
April 30, 2014, 09:14 |
|
#4 |
Senior Member
Join Date: Feb 2013
Location: Germany
Posts: 200
Rep Power: 24 |
First it is always a good idea to have your grid somehow aligned with the flow. This gives you less numerical errors. Second it is because of the structure itself or like you called it control (see attachments). It makes refinement easier and helps to avoid degenerated cells. Maybe you just try it on your on and you will easily see the difference.
|
|
April 30, 2014, 16:21 |
|
#5 |
New Member
Join Date: Jan 2014
Posts: 19
Rep Power: 12 |
Okay, thank you very much for your help kad! I'll try to toy with it and see what happens.
Cheers! |
|
Tags |
association, block, icem |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[Commercial meshers] converting Fluent mesh to openfoam standard mesh | deepesh | OpenFOAM Meshing & Mesh Conversion | 31 | March 29, 2017 06:59 |
Tecplot 360 2011 R2 preview with OpenFOAM data loader available for evaluation | scott_rumage | OpenFOAM Post-Processing | 50 | January 23, 2017 03:49 |
dsmcInitialise - dsmcFoam | archymedes | OpenFOAM Pre-Processing | 94 | July 15, 2016 17:14 |
CheckMeshbs errors | ivanyao | OpenFOAM Running, Solving & CFD | 2 | March 11, 2009 03:34 |
CheckMeshbs errors | ivanyao | OpenFOAM Running, Solving & CFD | 0 | March 10, 2009 23:07 |