Time integration failed

826 views
Skip to first unread message

Rahul Kumar

unread,
Feb 25, 2015, 6:34:16 PM2/25/15
to canter...@googlegroups.com
i am calculating the energy terms. I am using new python 2.7.9
error :Traceback (most recent call last):
  File "F:\Book\VIII Project\new python\adiabatic_flame.py", line 39, in <module>
    f.solve(loglevel=loglevel, refine_grid=True)
  File "onedim.pyx", line 662, in cantera._cantera.Sim1D.solve (cantera/_cantera.cpp:47931)
Exception: 
***********************************************************************
CanteraError thrown by OneDim::timeStep:
Time integration failed.
**************************

i donot know what to do. please tell me
adiabatic_flame.py

Ray Speth

unread,
Feb 25, 2015, 9:00:02 PM2/25/15
to canter...@googlegroups.com
Rahul,

Two changes that I would recommend, which seem to fix the problem, at least in this instance:

1) Set your initial grid to be evenly spaced:

initial_grid = linspace(0, 0.03, 7) # m

2) Keep grid refinement disabled (refine_grid=False) while you're solving with the energy equation disabled.

Regards,
Ray

Rahul Kumar

unread,
Feb 26, 2015, 3:41:57 AM2/26/15
to canter...@googlegroups.com
i cannot change the grid, Ray. uneven grid is required 

Ray Speth

unread,
Feb 26, 2015, 2:22:40 PM2/26/15
to canter...@googlegroups.com
Rahul,

The final call to 'solve' with refine_grid=True will give you a non-uniform grid that resolves the flame structure appropriately, regardless of initial grid. The initial grid can be chosen arbitrarily, with convergence of the solver on that grid being the only consideration. In this case, a uniform initial grid works while the non-uniform grid you initially supplied does not.

Regards,
Ray
Reply all
Reply to author
Forward
0 new messages