Non-selected node affects end-results

Thread index  |  Previous thread  |  Next thread  |  Start a new discussion

RSS FeedRSS feed   |   Email notificationsTurn on email notifications   |   4 Replies   Last post: October 29, 2013 11:45am UTC
Machiel Kolstein

Machiel Kolstein

October 25, 2013 4:53pm UTC

Non-selected node affects end-results

Hi,

I have various physics nodes in one single project. (Both are Transport of Diluted Species nodes.) One of them works with a variable "n" and another "h". I also have a Study, with a Step which is Time Dependent. Here, in "Physics and Variables Selection", I only select one of the two nodes (let's say, with output "n"). However, when I plot out the results (i.e variable "n"), these seem to be influenced by both nodes. When I only plot "h", I get 0, as it should be.
When I disable the second node (the one with "h"), and redo the study, then the result, plotting "n" is as expected.
In other words: the "n" result depends on whether or not the "h"-node is enabled, even though the "h-node" is not included in the time-dependent Study Step (nor in any other Study Step)! How is this possible?
Why does a node influence my result, even when the node it is not selected in the Study section?

Best regards,

Machiel

P.D. (I have COMSOL multiphysics 4.3a)

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Machiel Kolstein

Machiel Kolstein

October 28, 2013 5:22pm UTC in response to Machiel Kolstein

Re: Non-selected node affects end-results


Hi again,

After investigating some more, I found that this problem only occurs, when the "non-selected" node has a variable "h". Whenever I change the name of the variable, the problem is solved.
Actually, when I switch on the node, with variable "h", the program gets stuck. However, changing the variable name, the program gives a good result.
From this I deduced that "h" was some kind of proteced variable (Planck's constant). Fair enough. However, when I use "c" instead of "h", the problem does not occur.
In any case, shouldn't COMSOL give me a warning when I accidently use a constant as a variable name?

Best regards,

Machiel

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Jeff Hiller

Jeff Hiller
COMSOL Employee
USA
Moderator

October 28, 2013 6:55pm UTC in response to Machiel Kolstein

Re: Non-selected node affects end-results

h is the element size, per Reference Manual, version 4.3b, page 244.
h_const is the Planck constant, per Reference Manual, version 4.3b, page 257.
c_const is the speed of light, per Reference Manual, version 4.3b, page 257.
c is the default name for concentration if you have a mass transfer equation in your file, per Reference Manual, version 4.3b, page 714.
Best,
Jeff

Reply  |  Reply with Quote  |  Report Abuse

Machiel Kolstein

Machiel Kolstein

October 29, 2013 10:27am UTC in response to Jeff Hiller

Re: Non-selected node affects end-results


Hi,

Thank you very much for your reply and confirm this issue. I don't have version 4.3b, but 4.3a, but in the User Guide, on pages 153 and 166, I found the information as you stated. I use two Transport of Diluted Species nodes, so I can't just stick to the default "c". Anyway, I will be more careful in selecting variable names in future.
Still, I think it might not be a bad idea if COMSOL would provide some warning when a user tries to use certain reserved variables (like h or t or x) in the inappropiate place.
Best regards,

Machiel

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Jeff Hiller

Jeff Hiller
COMSOL Employee
USA
Moderator

October 29, 2013 11:45am UTC in response to Machiel Kolstein

Re: Non-selected node affects end-results

Hi Machiel,
Thank you for your feedback.
Best regards,
Jeff

Reply  |  Reply with Quote  |  Report Abuse


Rules and guidelines