Author |
Message |
Richard Beeler (rmbeeler)
New member Username: rmbeeler
Post Number: 2 Registered: 11-2004
| Posted on Wednesday, January 18, 2006 - 09:31 am: | |
Using 4.2.16 I find that contact/jump conditions on both the side and one end of a cylinder embedded in a block give the "NONSYMMETRIC MAP" error. The error does not occur if the condition is given on only the side or only the end. It occurs even if the same jump condition is used on the side and on the end. Do I need to upgraded to version 5? |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 523 Registered: 06-2003
| Posted on Wednesday, January 18, 2006 - 05:19 pm: | |
This may be a confusion at the joining point of the two boundaries. Try cutting off the corner, so that there are two distinct nodes ending the two boundaries. If this doesn't help, I will need to see the script.
|
Richard Beeler (rmbeeler)
Junior Member Username: rmbeeler
Post Number: 3 Registered: 11-2004
| Posted on Thursday, January 19, 2006 - 01:31 pm: | |
A break at the corner (with no contact) does stop the error. But not if the contact is applied to the break. While this is good for diagnostics, it doesn't fix the problem for me. (As far as I can see, a break at the corner, creates a leak and making the break small makes many elements. Also, having to build in breaks at joints, negates some of the meshing simplifications that attracted me to the CONTACT BC in the first place. (Though I don't want to go back to building in finite layers of resistance!) I'm guessing this should be fixable. (Easily said when I'm not the one doing the fixing!) But, isn't it analogous to having adjacent surfaces each with a NATURAL BC applied? 2.6 K | test_6.pde "Example of problem with adjacent CONTACT surfaces" | | |
|