Posted Thu, 24 May 2007 09:08:46 GMT by H. Clausen

whenever I'm trying to run my modell with 'automatic time step control', I receive the log message "WARNING: Math: overflow range error in function pow" and Feflow hangs up. This is especially strange as I'm currently not using a single .pow-function. However, it does work with constant time steps. How come?


Posted Thu, 07 Jun 2007 09:30:52 GMT by Kerstin Kernbach
Dear Holger,

This warning is unrelated to the FEFLOW power functions. Instead, it is a math (system) warning that refers to an exponential expression that evaluates to a value greater than the largest flaoting-point number that can be expressed (~10^308). With FEFLOW 5.3, parameter overflows are reported much earlier than before, but in most cases that's not critical.
Unless the consequences are obvious, it's hard to say where the error specifically originates.
I think, you can probably ignore that. It would be better to run the model with 'automativ time step control' in spite of warning as with constant time steps. Because that makes the simulation more instable, beside you have choosen really small time steps.


Posted Tue, 12 Jun 2007 14:17:48 GMT by H. Clausen
Thank you, Kerstin,

I would also prefer to run the model with automatic time step control, but FEFLOW just hangs up. I have to close it by closing exceed, because FEFLOW does not react at all.
Anyway, as I am still using the demo version, I hope it will not happen again when the full version is installed.



You must be signed in to post in this forum.