Updating of diffusion data from database...

[SuperX]

Dear Bernd:

It's me again. I used the model Janin gave me for the solidification of ductile cast iron, it works, and I have got some results. As you have said, the process of graphite forming is a very complex problem. So I am not very confident about my resuls....can you please help me take a look and tell me if it is right? The picture is in the attatchments.

There is another question about the solid state transformation. After the cementite has formed, the Micress software start to get some problems with "updating of diffusin data from daabase". The software won't crash out, but it just keep on "updating of diffusin data from daabase...". Six hours have passed for the updating, I guess there may be some problems, I checked all the diffuion settings, but I still can't find out the problem,so I turn to you for help.
Best wishes
SuperX

[Bernd]

Dear SuperX,

The question whether the results of your cast iron simulation is reasonable or not you should address rather to Janin than to me as I don't have experiences in this topic...

Your second problem, as far as I understand, consists in the observation that MICRESS does not provide more output than "updating of diffusion data from database" (twice) during 6 hours. The most typical reason if MICRESS is running but gives no more output is that something it is doing takes long time. Simulation can get very slow when a new phase appears (like cementite in your case) and the simulation time step gets very small. This appears reasonable in your case because
- you defined a quite high interface mobility of phases 1:5 (austenite-cementite)
- you have not defined any lower limits for time stepping
- the warning about demixing in interface FCC_A1-cementite for element SI indicates further trouble with this phase interaction
- there is no reason to assume that the problem is linked to the updating of diffusion data. Updating for cementite has to be done when it appears for the first time, and general updating is requested all 10 seconds. You can check the .diff output whether this updating has been successfully finished.

I would advise you to do the following:
1.) define SI as stoichiometric in cementite (Si has very low solubility in cementite, and calculation of solvus slope is numerically dangerous)
2.) check the tab_log outputs (.TabL, .TabT, .TabP) whether there has been response after the last output on screen (after 47.1 s). If there is no response, decrease the output interval and re-run MICRESS, until you see response ("tab_log" in "Selection of the outputs")
3.) In the .TabT file you now can see whether it is really the small time step which causes the symptoms, and which interface it provokes.
4.) Try to find out which parameters are those which determine this time steps (interface mobility, thermodynamics (e.g. extreme slopes due to missing definition of stoichiometric element), possible cutoff of small time step values in case they are due to local extreme values).

Bernd

[SuperX]

Dear Bernd:

Thank you for your help. I checked the TabT file and found out that it was the small time step that cause the problem. After I set the diffusion of Si in cementite as stoichiometric, the problem was solved.

But there is another problem. When the phase ferrite and cementite occur, an error takes place: "Force automatic start values initialisation failed, error= 105 interface FCC_A1/CEMENTITE " . Maybe I have set something wrong, but I failed to find out.

Also I want to ask, if it is possible to use micress to simulate the shrinkage cavities. Can you please give me some guides if it is possible? Thanks.

Best wishes
superX

[Bernd]

Dear SuperX,

the error messages clearly say that initialisation of the FCC/CEMENTITE interaction is failing. This initialisation is performed when nucleation of cementite is checked for the first time. Herefore, the local composition and temperature is used. Typical problems can be in general:

- nucleation is checked too early so that the new phase is far from being stable (especially phases with composition sets)
- local composition is far from realistic (e.g. negative compositions). Check .c*pha1 output.
- internal start compositions for iteration ("major composition") are not set properly when creating .ges5 file (especially applies to phases forming composition sets)
- input compositions (using "input" option for initial composition, also serve as MICRESS-internal start composition in case of numerical trouble) are not set or unreasonable values are given.
- definition of stoichiometric phases is wrong

I think in your case, the last point is the culprit: You intended to set Si stoichiometric in phase 4. However, you have set a composition limit to exactly 0.00, which means that initial equilibrium calculation is exited with error if the database predicts any small solubility of SI in CEMENTITE!
You should rather define SI as stoichiometric, thus allowing any constant composition:

# Phase diagram - input data
# ==========================
#
# List of phases and components which are stoichiometric:
# phase and component(s) numbers
# List of concentration limits (at%):
# <Limits>, phase number and component number
# List for ternary extrapolation (2 elements + main comp.):
# <interaction>, component 1, component 2
# Switches: <stoich_enhanced_{on|off}> <solubility_{on|off}>
# End with 'no_more_stoichio' or 'no_stoichio'
4 2
no_more_stoichio
#


I guess, you also got the error number 10402 somewhere which indicates that there was a violation of the user limit for element 2 in phase 4...

With respect to formation of cavities, I am not sure how the experimental MICRESS version which you got allows that. As vacuum in not a phase, you would have to include a gas phase and a corresponding gaseous element like hydrogen which is soluble in FCC and LIQUID. Whether this works would be a question for Janin...

Bernd