|
[Sponsors] |
March 22, 2012, 13:16 |
|
#61 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
4) taking another look at that squished disk, I think I see both an Ogrid and an HGrid overlapping... Perhaps you tried to put an Ogrid only within that pipe? You should run that OGrid thru the entire model instead...
If I get a chance, I will come back and show you later. Simon
__________________
----------------------------------------- Please help guide development at ANSYS by filling in these surveys Public ANSYS ICEM CFD Users Survey This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)... CFD Online Users Survey |
|
March 22, 2012, 15:08 |
|
#62 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
@Far: forget about question no. 2 - obviously I had forgotten the Elbow-Tutorial...argh; thx for the reminder.
@PSYMN: "3) At the interface between the pipes, is there a surface? If it is open flow, there shouldn't be and the blocking should be continuous without trying to project that between face to anything... If there is a surface, a continuous blocking will ignore it anyway..." --> Yes, I set them when I created the model to show the area where the fluid enters...also a case of low concentration... The rest I must check/try tomorrow. I wish you a good night respectivly a good start in the day...well, whatever time it is at your place right now |
|
March 23, 2012, 04:44 |
|
#64 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
I did copy the 2D-Blocking so that I got half of the top- surface of cylinder blocked and it worked quite well (model g6). But the extrusion in Z-direction by -70 produces a somehow twisted structure on the bottom side which gives me a bad quality.
--> model g6 mistakes: a) the copy of the 2d-Blocking "almost" worked correctly. There are only some minor problems: in the copied part there are more bad elements than there are in the original part (picture: copying 2d model g6) b) I didn't copy the geometry from the top to the bottom - is this maybe the reason for the 3d - extrusion problems? The edges going through the cylinder from top to bottom also need curves or surfaces? I also experimented a little bit with the periodicity options and improved the mesh a little (model g14) for the 90 degree 2d blocking...well, and then the copy procedure of the 2d blocking doesn't work as it did before... -->model g14 mistakes: a) I wasn't able to copy the entire blocking (of the 90° segment) as I was before - Error message (picture 1). b) The blocking in the copied parts was different to the original blocking. One can see that by the bad elements of the Determinant 2x2x2 (picture: wrong copy blocks). |
|
March 28, 2012, 08:05 |
|
#66 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
Hey Far,
I still have problems with the rotate block function even when I follow the instructions: rotate geometry, rotate blocks, update associations. It's the same with the translate function for creating 3D blocking out of my 2D topology (translate blocking). In both cases the copied mesh isn't equal to the original mesh. You said that you already created the 3D Mesh. Would you give me your 2D topology you used to create your 3D blocking? I want to test it on my computer. If it works I know that the mistake is something I did before, if it doesn't work the mistake must be somewhere in my settings. Best Regards Last edited by despaired student; March 28, 2012 at 11:05. |
|
March 28, 2012, 10:14 |
|
#67 |
Super Moderator
|
Ya, Sure.
I am working on similar problem here, this may be helpful to you. http://www.cfd-online.com/Forums/ans...-fuel-rod.html |
|
March 28, 2012, 11:17 |
|
#68 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
Hi,
I've read the thread. When the association-problems appear after "rotate blocks" or "translate block" then it is very likely that these are the same problems. Could this be a problem of too low computational power or could the amount of copied blocks be too big? |
|
March 28, 2012, 12:11 |
|
#69 | |
Super Moderator
|
I am thinking to copy few blocks at moment and then associate them properly. And repeating that procedure for other blocks. I shall post those files within day or two.
Quote:
|
||
March 28, 2012, 15:17 |
|
#70 | |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
Quote:
|
||
March 29, 2012, 03:49 |
|
#71 | |
Super Moderator
|
Quote:
This is exactly the method we are using. |
||
April 3, 2012, 04:19 |
|
#72 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
I've tried this procedure several times but I always had negative results. I'm very sure that the problem is somehow connected to the size of the blocking because it works with smaller blocking.
|
|
April 17, 2012, 05:57 |
|
#73 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
Hi Far and other helpers,
I received good results by dissociating all edges and vertices before using "mirror blocks" or "rotate blocks". Afterwards you must/can re-associate the vertices and edges. In this case there was no pre-mesh shown after extrusion...don't know why? But I'm sure that a pre-mesh had been created because I could create a mesh and by quality control one could see the chosen elements within the geometry so somehow there was a pre-mesh... It was also possible to extrude the mesh - not the pre-mesh. I'm not sure if this is what you were talking about. In this case I misunderstood the meaning. Maybe these information help if someone else has similar problems. Best regards |
|
April 17, 2012, 08:46 |
|
#74 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Thanks for getting back to us on this.
Your "workaround" of dissociating the blocking from the geometry may be a clue to what ever the problem is. I will mention it to development. I know they recently (in the past release or two) added an option so that copy/rotating the blocking would bring a copy of the associated geometry with it also... I wonder if it is an issue in that code. As for extrude, we were talking about extruding the blocking faces. It is more flexible than a mesh extrusion. But since it didn't work for you, good job figuring out a work around. Best regards... Simon
__________________
----------------------------------------- Please help guide development at ANSYS by filling in these surveys Public ANSYS ICEM CFD Users Survey This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)... CFD Online Users Survey |
|
April 17, 2012, 09:27 |
|
#75 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
I should clear my statement at one point: rotate and mirror blocks worked quite good with all edges and vertices dissociated but Create Block >> 2D to 3D >> Method: translate still does not work properly.
Here I had to use the real mesh for extrusion. @PSYMN: You mentioned the last two versions of ICEM. I'm currently using Icem 14.0 but started with ICEM 13.0 First I tried a way recommended by Far 1. rotate geometry 2. rotate blocking 3. update associations --> Here most problems occurred within the associations in the O-grids. Other parts were also affected but mainly O-grids. Last edited by despaired student; April 18, 2012 at 09:31. |
|
April 18, 2012, 09:39 |
hopefully last correction
|
#76 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
The function create blocking>>2D to 3D>>Method translate worked for a value of 70 which is the length of the geometry... but I had to dissociate edges and vertices before doing this...
|
|
April 18, 2012, 21:36 |
|
#77 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
I was not able to reproduce any problems on my end. Can you be more specific about what was going wrong for you?
__________________
----------------------------------------- Please help guide development at ANSYS by filling in these surveys Public ANSYS ICEM CFD Users Survey This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)... CFD Online Users Survey |
|
April 25, 2012, 03:44 |
|
#78 |
Senior Member
Sören
Join Date: Mar 2012
Posts: 102
Rep Power: 14 |
I've send you the link to a video-download where you can see the problems.
enclosed are the .tin, .blk files etc. |
|
March 4, 2013, 10:05 |
|
#79 |
Senior Member
starter
Join Date: Sep 2012
Posts: 125
Rep Power: 15 |
hi far
I could not find these tutorials. Message I get is that they are removed from website. Can you please see if URL is correct. regards |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
complex number operation in udf? | hongxiao | FLUENT | 5 | June 14, 2017 14:11 |
Error: Operation would result in Non-Mainfold geometry | fergal | ANSYS Meshing & Geometry | 1 | December 9, 2013 08:04 |
operation pressure and pressure outlet BC | Lilly | FLUENT | 4 | May 8, 2013 09:57 |
Operation to READ_IF_PRESENT field | sega | OpenFOAM Programming & Development | 6 | February 15, 2011 09:57 |
Initalization fails with Operation Pressure = 0 Pa | Tobias | FLUENT | 3 | October 31, 2005 09:02 |