Author |
Message |
c_bailey (c_bailey) Member Username: c_bailey
Post Number: 9 Registered: 01-2010
| Posted on Thursday, February 18, 2010 - 01:59 pm: | |
On line 27 of this script, a trivial value is defined for "Seebeck", which is either overwritten in some regions or only part of an INACTIVE() equation in the only other region. This value should never matter. Yet, setting it to zero causes a funny error message about "Equation is Empty". Part of the error message appears to be in unprintable characters. This descriptor is very similar to another I recently posted, but the issue is different this time. I do not think the equations are physically correct, but still, a value of zero for this "Seebeck Coefficient" parameter is physically possible (it occurs in superconductors and is close enough to zero in the metal lead that this is sometimes assumed). Moreover, the value on this line should not actually get used anyplace, should it? |
rgnelson Moderator Username: rgnelson
Post Number: 1330 Registered: 06-2003
| Posted on Thursday, February 18, 2010 - 10:26 pm: | |
1) the unreadable part of the diagnostic is an erroneous attempt to report a layer name, which doesn't exist in a 2D problem. 2) the diagnostic check for empty equations is done in the equation processing pass, before any computations are done, and we forgot to test for inactive variables before reporting an error. In the meantime, you can get around the error report simply by writing equations with non-zero coefficients, even in regions with inactive variables. We will correct both of these items in the 6.13 maintenance release.
|
|