reg high temperature and cooling rates solidification

[swetha]

Dear Bernd

This is the error message shown when I am trying to simulate for SLM conditions.This is because of the quasi-equilibrium problem you mentioned about earlier or some other problem the velocity used is 20 cm/s. At the start, after grain reaches full size ,it shows the following.

--> Force automatic start values
Forcing automatic start values
Automatic start values will be set
Thermo-Calc error 1611 MICRESS error 16 phases 1/ 0
trying hard phases 1 0 level: 3 zp= 304307 error= 16

Regards,
swetha

[Bernd]

Dear swetha,

The error message is an internal error of the calculation of quasi-equilibrium. More specifically it means that the backward equilibrium could not be calculated which is necessary for deriving the linearisation data.
But this information does not say much about why the error occurs. Typically, any type of numerical problem makes first the Thermo-Calc subroutines complain, because numerical problems always earlier or later lead to local concentrations or other conditions which are not physical any more (e.g. negative concentrations).
Are you sure that, when switching from "milder" conditions to SLM conditions, you properly adjusted the length scale (grid resolutions) and time scale (update intervals, intervals for nucleation checks, output times, etc.)?

Bernd

[swetha]

Dear Bernd,

I will try to simulate with 1d temp coupling for that condition.
1.With further adding the alloying elements in the ternary system , so that to simulate single phase for an inconel superalloy system,I made the following changes in the driving file,it is running but the phase disappeared at the first step itself.I don't know what changes/parameters i have to check further in this file,so that I can run the simulation and get the output.I have attached the driving file as PM.

2.I have seen simulation conditions for a similar superalloy which has subternay systems.How can I add such subternay systems for a complex alloy system?

Can you help me with this.

Regards,
swetha

[Bernd]

Dear Swetha,

1. There is nothing obvious in the driving file which explains why the initial grain is vanishing. Initial small grains (with radius 0) vanish sometimes if numerical conditions are bad. To find out what is bad, it may be helpful to start with a big grain instead (radius > Δx).
Apart from that, I stumbled over some details in your input file:
- interface thickness 15 cells: This is huge, and I don't see why you do that. A typical value is 3
- Your initial temperature is 1400 K which is far below the liquidus temperature of your alloy.
- The distance for the 1d-far-field solution is bigger than the moving frame distance. Typically, both values should be identical. Anyway, both values appear to be quite big...

2. In MICRESS, there is a multi-ternary extrapolation scheme for redistribution available which allows taking into account strong interactions between dissolved elements. This is done using the "interaction" keyword followed by the numbers of the two elements which interact. By this way, a more efficient redistribution of ternary subsystems is possible. To select ternary subsystems (which always include the matrix component and which are not allowed to overlap otherwise) one can use the partition matrix which is written in the .log file.
I am not sure whether this option could be helpful for you. You may consider to give it a try after your simulations are running properly otherwise.

Bernd