# meshing problem

 Register Blogs Members List Search Today's Posts Mark Forums Read

 February 27, 2003, 21:05 meshing problem #1 danard Guest   Posts: n/a Good day everyone, I'm working on the rectangular curved diffuser (arc shaped) and I found that the mesh I generated in Build on two adjacent surfaces (where one plane has an agle of around 10 deg from the another plane) have different size. Mesh on one surface is obviously much more refined than the others. I try to merge these surfaces together to get a uniform mesh but it fails. Would anyone has the experience in this problem? Any suggestion is very welcomed. Thanks for your help.

 February 28, 2003, 14:09 Re: meshing problem #2 Robin Guest   Posts: n/a Hi Danard, You will not be able to merge two surfaces which face each other at a 10 degree angle. You should modify the mesh controls such that the mesh size at this location is similar for both surfaces. Alternatively, you could turn on the surface proximity to do this automatically. That said, no matter how you mesh it if you have surface normals 10 degrees apart and facing each other you are going to have problems. Best case would be if a single element filled that corner and gave you a 10 degree skew. 10 degrees is not very good, but tollerable. If you tet mesh it, you will end up with two elements in the corner, each with 5 degree skew, which is bad. The solver may tolerate this if the region is not very active, but if it is close to a open boundary, you may have trouble. What you shoud really do is modify the geometry. Ask yourself "Is this really a region of interest?". If not, cut off the sharp angle and put a flat face at the end of it (two 95 degree corners are much easier to handle than one 10 degree one). Chances are that this is only a CAD feature and the real geometry will not have a 10 degree sharp edge (no matter how you make it, it has to flatten out somewhere). Hope this helps. Regards, Robin

 March 1, 2003, 02:50 Re: meshing problem #3 danard Guest   Posts: n/a Thanks for your detailed explanation. I'm quite curious about the difficulty of merging two surfaces which face each other at more than 10 degree angle. Is that a common problem for the CFD software or just the limitation of CFX code? Will be grateful if anyone can tell me more about why it can hardly be done. I need to convince myself that there's no other alternative than changing the geometry. I mean whether there's way to make the size of the mesh elements on two surfaces looks alike. How can I use the mesh control in this case? People advise me that mesh control should be always be used as the last option. Is that true? Any suggestions will be very welcomed. Thanks for sharing your view.

 March 2, 2003, 20:17 Re: meshing problem #4 Robin Guest   Posts: n/a Hi Danard, By merging surfaces you are asking the code to take the mesh you can produce, however bad, and create a new underlying parameterization. This new 'merged' surface parameterization is then used by the mesher. This is good for getting rid of CAD edges which do not physically exist and de-featuring small scale geometry. In your case, a 10 degree internal angle, the merging algorithm is not appropriate and will fail (as appropriate). This is fairly common for many meshing applications but has nothing to do with CFD. As for mesh controls, they are an essential part of meshing, not a last resort. I have no idea what you geometry looks like, so I cannot say for certain what to do. Regards, Robin

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post [Other] Hex meshing problem DM12 patrick ANSYS Meshing & Geometry 7 January 9, 2015 08:21 ashtonJ CFX 4 August 15, 2014 20:55 brunob ANSYS Meshing & Geometry 3 June 21, 2011 20:03 Vidya Raja FLUENT 0 May 20, 2006 23:31 Gauthier Lambert Main CFD Forum 1 August 3, 2000 09:22

All times are GMT -4. The time now is 18:14.