Hi all,

We have been working on modifying Jonathan's gist* code 
(crazycoupling2.ipynb<https://gist.github.com/guyer/bb199559c00f6047d466daa18554d83d#file-crazycoupling2-ipynb>)
 of how to couple the PDEs along the upper boundary of a 2D mesh, i.e. wherein 
one PDE is along the x-axis and the other acting along the y-axis, whereas the 
two exchange information along the top edge of the 2D grid.  The original 
question is posted here**.

As per Jon's advice, we'd benefit from solving the two equations together in a 
tightly coupled fashion, even though we "waste" resourced by sweeping the 1 
dimensional PDE over the entire 2D range.

The problem is that when we try the .solve() method instead of the current 
.sweep, this crashes Python , and we hit a memory fault error.  "Process 
finished with exit code -1073741819 (0xC0000005)."  This happens after a 
warning is issued by Pysparse's linearLU solver. "VisibleDeprecationWarning: 
using a non-integer number instead of an integer will result in an error in the 
future.   b = b * (1 / maxdiag)"

The code with the .solve() implementation is here. 
https://gist.github.com/krishnakumarg1984/f1ef1864c503c4eb66dbf83a072fdd2d   
(essentially, the only change is to comment out the while loop which sweeps the 
PDEs, and instead replacing them with a single .solve(dt=dt) code.

We'd appreciate any help in solving this issue.

Best regards

Krishnakumar




* https://gist.github.com/guyer/bb199559c00f6047d466daa18554d83d

** https://www.mail-archive.com/[email protected]/msg03731.html
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
  [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]

Reply via email to