Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Domain selection bug when working in Weak Form

Please login with a confirmed email address before reporting spam

Hi all,

Sorry in advance for my english, it's not my native language.

My model has 3 domains in union. When I add a PDE Weak Form node to my model, the default setting is to include
all domains in the first Weak Form PDE 1. But when I add 2 other Weak Form PDEs under the same node (Weak Form PDE 2 and Weak Form PDE 3), it's impossible to unselect domains 2 and 3 for the Weak Form PDE 1.

If I add a node of Coefficient Form PDE or General Form PDE, it's possible to unselect domains for the first equation after adding extra equations.

Could it be a bug? See the attached jpg and model.


4 Replies Last Post 13 janv. 2011, 10:09 UTC−5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 12 janv. 2011, 14:05 UTC−5
After looking again at my model, what I mean is that domains 2-3 should be overridden in Weak Form PDE 1, just like they are overridden in Coefficient Form PDE 1 and General Form PDE 1.

Francois
After looking again at my model, what I mean is that domains 2-3 should be overridden in Weak Form PDE 1, just like they are overridden in Coefficient Form PDE 1 and General Form PDE 1. Francois

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 janv. 2011, 06:14 UTC−5
Hi

your node tree should be read from top to bottom, if you select a n element in the top node (one layer higher than your image) then ALL HAVE TO BE part of the first default. Actually you should go one level higher and unselect it there.

You can also add a "later" level node and redefined something existing previously, this gives the overridden or unavailable entries, that will fall-back here if the later node definition is i.e disabled or removed/deleted

the node icon >< defines if these are relative summing or absolute definitions (read through the doc)

--
Good luck
Ivar
Hi your node tree should be read from top to bottom, if you select a n element in the top node (one layer higher than your image) then ALL HAVE TO BE part of the first default. Actually you should go one level higher and unselect it there. You can also add a "later" level node and redefined something existing previously, this gives the overridden or unavailable entries, that will fall-back here if the later node definition is i.e disabled or removed/deleted the node icon >< defines if these are relative summing or absolute definitions (read through the doc) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 janv. 2011, 08:48 UTC−5
Hi Ivar,

Thanks for the follow up.

Adding a "later" level node and redefine something existing previously gives the overridden or unavailable entries only in the Coefficient Form PDE or General Form PDE modes, but that doesn't work in the Weak Form PDE mode. Look at the model file attached on my 1st thread, this is the difference (bug?) that I wanted to point out.

I know that I could do 3 separate nodes and select only 1 domain on the highest level of each node, but it's not an elegant way to proceed since I would have to link the internal boundaries manually with pairs in order to solve for the same dependent variable in the 3 nodes/domains, etc.

Francois
Hi Ivar, Thanks for the follow up. Adding a "later" level node and redefine something existing previously gives the overridden or unavailable entries only in the Coefficient Form PDE or General Form PDE modes, but that doesn't work in the Weak Form PDE mode. Look at the model file attached on my 1st thread, this is the difference (bug?) that I wanted to point out. I know that I could do 3 separate nodes and select only 1 domain on the highest level of each node, but it's not an elegant way to proceed since I would have to link the internal boundaries manually with pairs in order to solve for the same dependent variable in the 3 nodes/domains, etc. Francois

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 janv. 2011, 10:09 UTC−5
I hadn't noticed that Weak Form PDE nodes are "Contributing nodes" and not "Exclusive nodes", that's why they don't override previous selections. I guess it's for keeping the Weak Form mode the most flexible possible.

Thanks Ivar for pointing me out the >< icons.
I hadn't noticed that Weak Form PDE nodes are "Contributing nodes" and not "Exclusive nodes", that's why they don't override previous selections. I guess it's for keeping the Weak Form mode the most flexible possible. Thanks Ivar for pointing me out the >< icons.

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.