|
[Sponsors] |
July 11, 2019, 08:19 |
Issue with wrapping
|
#1 |
New Member
Jalvathi
Join Date: Apr 2019
Posts: 7
Rep Power: 7 |
Hello all.
I am doing surface wrapping with the help of ANSA. I am giving area refinement for specific PIDS and a general coarse mesh for the rest fo the volume where not much details are required. I need the inner volume of the car to be wrapped. I have been running Octree wraps for the last three days and they seem to crash everytime or run in a long loop where I donīt know if they are running or they have crashed. Can any of you please let me know, why surface wrapping in ANSA occurs in two phases? The minimum refinement I have used is 0.0003m and it is only in a much smaller part. The rear volume and the other parts have refinement of 0.01m. I have also reduced the refinement a bit and tried to do the surface wrap. It worked fine, but right after when I try the same using a copy of the octree refinement, it simply doesnīt work. Any help in this area will be highly appreciated. Thanks. Jalvathi Last edited by jalvathi; July 12, 2019 at 10:39. |
|
October 2, 2019, 08:22 |
|
#2 |
Senior Member
Gerry Kan
Join Date: May 2016
Posts: 373
Rep Power: 11 |
Dear Jalvathi:
I ran into a similar problem, and I could see ANSA wrapping stuck at a relatively low percentage (8% in my case) but the memory usage kept going up and up (however sometimes not). I noticed that there are two issues: 1) My geometry has about 300 different non-intersecting entities. (Manually) assigning each entity to its own PID speeds up the self-proximity check immensely. 2) The geometry was about 2 x 2 km, and the geometry (STL) is represented in meters. When I scale it down to, say, mm, the wrapping goes through relatively quickly. Then I need to scale it back to meters afterwards. I don't know if these situations apply to you, but for a vehicle in wind tunnel it could. Hope this helps, Gerry. |
|
October 23, 2019, 03:33 |
|
#3 |
Senior Member
Gerry Kan
Join Date: May 2016
Posts: 373
Rep Power: 11 |
Dear Jalvathi:
I dug a bit more deeply on this issue. This has to do with the length units used in the wrapping operation. This is always set to the system unit and is by default in millimeters. If your CAD model is in a different unit (say, meters), you should either change the model unit or the default unit to reflect that, so that the wrapping operation does not produce a surface mesh that is too fine or too coarse. Gerry. |
|
October 24, 2019, 01:55 |
|
#4 |
Member
Maries
Join Date: Mar 2010
Location: Cologne, Germany
Posts: 75
Rep Power: 16 |
Hi,
I didn't notice this query for long time. What Gerry said is partially right. It is due to units. Also one more thing we need to take care while wrapping is the min. length. I assume you are all using latest ansa versions, which uses octree method for wrapping. In this method if we give the min. length less than 1, then ansa may crash or take long time. So always use the min.length greater than 1. Best Regards Maries |
|
October 24, 2019, 08:19 |
|
#5 | |
Senior Member
Gerry Kan
Join Date: May 2016
Posts: 373
Rep Power: 11 |
Quote:
Gerry. |
||
Tags |
ansa, octree, preprocessing, refinement, surface wrapping |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
CAMWA special issue on open-source numerical solver | feixu2019 | OpenFOAM Announcements from Other Sources | 0 | October 1, 2018 12:21 |
CAMWA special issue on open-source numerical solver | feixu2019 | SU2 News & Announcements | 0 | October 1, 2018 12:19 |
Convergence issue in natural convection problem | chrisf90 | FLUENT | 5 | March 5, 2016 09:30 |
Strange issue while launching ANSYS workbench in CentOS 6.4 | Philip_C | ANSYS | 11 | August 29, 2013 07:44 |
Meshing related issue in Flow EFD | appu | FloEFD, FloWorks & FloTHERM | 1 | May 22, 2011 09:27 |