Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Creating limited contact geometry

Please login with a confirmed email address before reporting spam

I am currently working on simulating conduction within a small packed bed of uniformly sized spheres but having problems with building the geometry and meshing. If I try to form a union, some of the domains become unselectable. And while the alternitive option of forming an assembly doesn't have this problem, it introduces other issues and still refuses to mesh. The file I have attached is has 11 of the 60 spheres I use in my model enabled. 2 of these 11 spheres refuse to be selected if Form Union is used, and show weird things between them in the case of Form an Assembly. The goal is to be able to implement thermal contact BC's at the faces of contacts and either insulated/diffuse surface BC pretty much every where else. Preferably without haveing to manually use the GUI to go through and specify each of the 240+ contact locations manually. Ivar, if you can hear me, I am fully willing to sacrifice an undergrad from our lab upon an alter to you in exchange for your help.



1 Reply Last Post Mar 23, 2018, 6:46 p.m. EDT

Please login with a confirmed email address before reporting spam

Posted: 7 years ago Mar 23, 2018, 6:46 p.m. EDT
Updated: 7 years ago Mar 24, 2018, 6:36 p.m. EDT

Thanks to a wonderfully helpful COMSOL employee who contacted me almost imediately after I posted my initial post, I found that much of my initial problems were due to the boundarys of the faces of the individual sphere's lining up with one another. Fixing this for 60 spheres wasn't easy, but eventually I was able to build a geometry using the app builder, save it, and import it back into another mph file. By using the app builder, I was able to both create each sphere individually and add a rotation transform for each. The rotation angles and axes were randomly generated for each sphere. However, while this seems to help the geometry engine, and I can no longer find un-selectable domains pre-union, the geometry still doesn't seem to be coming out quite right. When looked at to assign boundary conditions, it becomes evident that some domains are still merged and others un-selectable - even though the face boundarys on these nodes are now not aligned. And it won't mesh. Does anyone happen to have any other suggestions or have seen similar problems before? The problem still looks to have something to do with face singularities, but I'm not sure exactly what.

Thanks to a wonderfully helpful COMSOL employee who contacted me almost imediately after I posted my initial post, I found that much of my initial problems were due to the boundarys of the faces of the individual sphere's lining up with one another. Fixing this for 60 spheres wasn't easy, but eventually I was able to build a geometry using the app builder, save it, and import it back into another mph file. By using the app builder, I was able to both create each sphere individually and add a rotation transform for each. The rotation angles and axes were randomly generated for each sphere. However, while this seems to help the geometry engine, and I can no longer find un-selectable domains pre-union, the geometry still doesn't seem to be coming out quite right. When looked at to assign boundary conditions, it becomes evident that some domains are still merged and others un-selectable - even though the face boundarys on these nodes are now not aligned. And it won't mesh. Does anyone happen to have any other suggestions or have seen similar problems before? The problem still looks to have something to do with face singularities, but I'm not sure exactly what.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.