erroneous initialization
how to solve errors with erroneous initialization? and also the errors regarding temperature out of range?
3 replies
-
Hi Faisal,
"ERRONEOUS DATA INITIALIZATION" errors are a result of having specified a thermodynamically invalid set of primary variables in blocks PARAM.4, INDOM, and/or INCON. Check the primary variables for the desired phase combination (accounting for the occasional "+10" and "+50") and make sure mass fractions are thermodynamically valid for the given pressure and temperature.
In general, for single-phase conditions, mass fractions have to be smaller (or greater, depending on which mass fraction you need to specify) than the equilibrium mass fraction. To find out what the equilibrium mass fraction is, initialize a single element under two-phase (or three-phase) conditions, run it for a single, small time step, and check the equilibrium mass fractions in the output file (the local thermodynamic equilibrium assumption assures that you get equilibrium mass fractions whenever the corresponding free phase is present).
The following slides, which can be found in this Forum under "Help Articles", "TOUGH Tips & Tricks", may be helpful as well:
http://esd1.lbl.gov/files/research/projects/tough/support/initial-boundary-conditions.pdf
Regarding "TEMPERATURE OUT OF RANGE". It means just that! If it happens during a Newton-Raphson iteration, the time step will be automatically reduced, and the simulation may continue just fine. In this case, you may ignore the message.
However, there are obviously cases where the temperature just goes out of range because of the physics you specify (either you inject too much heat, or the pressure goes too low, evaporation occurs, and the temperature goes below 1 degree). In these cases, check your model set-up - or acknowledge that you try to solve a problem that the chosen EOS cannot handle.
Stefan