|
[Sponsors] |
November 13, 2009, 05:48 |
C Grid for U bend Pipe
|
#1 |
Senior Member
Join Date: Mar 2009
Location: Indiana, US
Posts: 186
Rep Power: 17 |
Need help to block a C-grid around the U bend of a pipe. Seems simple and straightforward, but i tried everything and doesn't seem to work.
Thanks |
|
November 15, 2009, 16:30 |
Top Down...
|
#2 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Look at the shape of the upipe and you can see that it is a Cblock with the central block deleted... That is a key to using ICEM CFD Hexa, use the OGRID tool to capture your overall blocking topology.
I will attach an image to illustrate the two main ways to capture this geometry... The first one, top down like a sculptor, is the best way with the ICEM CFD tools. Try the Sphere Cube workshop to drive this principle home. After you have done this first step to use Ogrid to capture the general topology, you will need OGrid again at the end to capture the boundary layer... |
|
November 16, 2009, 01:19 |
Thanks Simon
|
#4 |
Senior Member
Join Date: Mar 2009
Location: Indiana, US
Posts: 186
Rep Power: 17 |
Thanks a lot. The top down approach looks lovely. WIll try it.
|
|
November 29, 2010, 07:29 |
|
#5 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
hello Simon
I too have stuck in the same problem. I saw your reply but could not make out how to capitalize that "top-down topology creation". When i delete the central block, i get an L shape (without using O- grid method.) Can you please elaborate (involved steps), how we can get 2nd & 3rd block from the 1st block & then mapping order. Is there some pre-requisite(e.g. no. of curves along the bend..etc), that should be full-filled in the geometry for topology to work properly. I have been trying to find out proper topology for an U bend pipe for last 4-5 days, but without success. Now it seems like..I am at the right door..!!! Please help me ..i am in dire need. Thanks in advance.. |
|
November 29, 2010, 14:23 |
|
#6 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
If anyone else has..some important piece of information..then plz share..!!
|
|
November 30, 2010, 01:13 |
|
#7 |
Senior Member
Join Date: Mar 2009
Location: Indiana, US
Posts: 186
Rep Power: 17 |
Pradeep, you need an O-grid first, and then 'delete' the central block. You will then have a U to capture your geometry. Post some pics of your blocking so we could help you better.
Regards |
|
November 30, 2010, 01:39 |
|
#8 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
I have seen your geometry, & i m doing exactly the same problem as you have done i.e. generation of structured hexa mesh in an U bend tube. I was trying the top down topology as stated above, but couldn't succeed.
You please just list me the involved steps on the basis of your own blocking including... 1. splitting to O grid block from the initial block, 2. deleting some of the blocks & then, 3. association of block edges to curves of the geometry.... for the same blocking & geometry as you have posted above here. Please correct my procedure if above seems to be wrong. If possible please attach some images. Thanks a ton for your response..i was really feeling helpless. ---------------------- regards Pradeep |
|
November 30, 2010, 01:45 |
|
#9 |
Senior Member
Join Date: Mar 2009
Location: Indiana, US
Posts: 186
Rep Power: 17 |
Post some pics of your blocking ... I followed the top-down approach exactly and it worked.
|
|
November 30, 2010, 02:00 |
|
#10 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
|
|
November 30, 2010, 02:31 |
|
#11 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
this is the picture
|
|
November 30, 2010, 02:49 |
|
#12 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
this is the picture...of two kind of blocks ..that i think i can use..
|
|
November 30, 2010, 06:51 |
|
#13 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
I could not under stand..whether first delete the central block and then split in to O-grid block ..or first split the original block in to O-grid block & then delete the central block.
Also after splitting in to O-grid block association between edges and curves under Associate tab becomes very confusing. |
|
November 30, 2010, 06:54 |
|
#14 |
Senior Member
Join Date: Mar 2009
Location: Indiana, US
Posts: 186
Rep Power: 17 |
First create O-grid then delete the central block
|
|
November 30, 2010, 07:00 |
|
#15 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
but after creating O-grid, the association between edges and curves under Associate tab becomes very confusing as there are many blocks and many edges now.
|
|
November 30, 2010, 07:16 |
|
#16 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
The main problem is of association (between block edges & geometry curves )or you may call it mapping, unless which one can not proceed further for generating mesh.
|
|
November 30, 2010, 07:54 |
|
#17 |
Member
jeevan kumar
Join Date: Mar 2009
Posts: 88
Rep Power: 17 |
hello pradeep
can you post the image of the c topology which you made before association. since for this geometry there will be only 3 blocks after deleting one. i will try to make a ppt and send you later seeing your topology. Thank you |
|
November 30, 2010, 12:51 |
|
#18 |
Member
pradeep
Join Date: Jul 2010
Posts: 31
Rep Power: 15 |
Hello jeevan..i have already posted two pics showing the same..!!
thnx forur response |
|
November 30, 2010, 13:51 |
Not an Ogrid at all...
|
#19 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
pradeeppandeygbpec, your images show an HGRID split, not an OGRID at all.
You should start over... 1) Re initialize blocks to get one single block. 2) Ogrid, select the block, select the three faces as shown in my illustration. 3) Apply. You will now have a CGrid with 4 blocks. The three outer blocks are your Pipe, the inner block is not needed. Your index control should have 4 indices, i, j, k, and O3. 4) Delete the inner block. 5) Associate the remaining blocks to the pipe. ... |
|
November 30, 2010, 13:52 |
Alternative...
|
#20 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Alternatively, you can just take that single initial block and fit it to the pipe... (project the edges at one end to the curve and the inlet of the pipe, and then project the opposite edges to the curves at the outlet of the pipe. Then add splits for control points so you can steer the block along the pipe.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Pipe bend erosion problem | John Yang | FLUENT | 2 | December 12, 2007 04:06 |
Grid Independence problem for simple pipe geometry | chanchala | FLUENT | 13 | November 20, 2003 08:24 |
Terrible Mistake In Fluid Dynamics History | Abhi | Main CFD Forum | 12 | July 8, 2002 09:11 |
Combustion Convergence problems | Art Stretton | Phoenics | 5 | April 2, 2002 05:59 |
fluid flow fundas | ram | Main CFD Forum | 5 | June 17, 2000 21:31 |