error #40300 and 105 while initiating Liquid/FCC_L12#2 interface

technical aspects of .dri file generation (e.g. debug mode ) etc...
Post Reply
Atur
Posts: 33
Joined: Tue Apr 13, 2021 11:39 am
anti_bot: 333

error #40300 and 105 while initiating Liquid/FCC_L12#2 interface

Post by Atur » Sat Feb 26, 2022 11:43 am

Dear Bernd,

I recently saw that there is L12 phase in my as-built samples produced by SLM and I wonder whether it formed during initial solidification or cyclic heating and cooling during deposition of subsequent layers/tracks. Therefore I am trying to involve L12 phase in my solidification simulations to see whether it will really form during initial solidification, however I receive following error:

Force automatic start values
initialisation failed, error = 40300 interface LIQUID/FCC_L12#2
Force automatic start values
initialisation failed, error = 105 interface LIQUID/FCC_L12#2

my concentration solver setup reads as following: Phase number 1 is disordered FCC (named as FCC_L12 in database) and phase number 3 is L12 phase (names as L12#2 in database) .
stoichiometric 3 2-3
diagonal
disordered 1 1 2
ordered 3 1 2
diff_comp_sets 1 3
vol_comp_sets 1 3
tic_tq_segments
no_more_stoichio

I thought since these are different composition sets in the database, the simulations should run without an issue but here is the case. Can you recommend any to this error?

Lastly, is it somehow possible to build up a simulation in this predictive manner?

Regards,
Ahmet

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: error #40300 and 105 while initiating Liquid/FCC_L12#2 interface

Post by Bernd » Sat Feb 26, 2022 7:31 pm

Dear Atur,

FCC_L12 is a thermodynamic description for the fcc-phase which allows for an order-disorder transition. Thus, FCC_L12#2 (or phase 3) should be the ordered phase. The problem with composition sets in MICRESS (and also other software like DICTRA) is that even if you have defined 2 phases in MICRESS, each one can switch to the wrong composition set (or state of order). For this reason, in your input file, you put the check that phase FCC_L12 should be disordered and FCC_L12#2 should be ordered. If not, the equilibrium is rejected, and the error number 4xx00 appears in the MICRESS output (with xx being the phase number).

Error 105 does not give more information in this case, it just means that iteration of quasi-equilibrium failed (due to error 40300)...

Thus, in your case the initial equilibrium ends with a disordered FCC_L12#2, and therefore is rejected. This can have a large number of reasons, e.g. when trying to calculate (quasi-)equilibrium above the spinodal temperature or outside the critical concentration range. Without further information about the specific system and the conditions for initialization I cannot give concrete recommendations here.

I did not get your last question about "predictive manner"...

Bernd

Atur
Posts: 33
Joined: Tue Apr 13, 2021 11:39 am
anti_bot: 333

Re: error #40300 and 105 while initiating Liquid/FCC_L12#2 interface

Post by Atur » Sun Feb 27, 2022 12:23 pm

Dear Bernd,

Thank you for your reply. I will try to provide you more information about the system. Moreover, I sent you my GES file and drive file via direct message.

Yes, in database description they state FCC and L12 phases are modelled with single gibbs free energy curve which describes order-disorder transition. So in my case at lower temperatures, e.g. below 650°C, disordered FCC_L12 becomes L12#2. The solidification is mainly two phase, namely FCC and BCC_B2, where BCC_B2 is the ordered interdendritic phase. This is already observed and validated from experiments. However, recently, we saw that L12 phase is present in the as-built sample after SLM, which is actually not a stable phase in the given solidification interval (according to Scheil). At given composition, the L12 phase forms below 650°C (way below the solidification interval) and calculations do not indicate that L12 would be stable for an unchanged FCC composition. So now I wonder if this phase is a formed during solidification (might be possible, short range diffusion), or it is in-situ precipitated during subsequent cyclic heating. The latter should be possible if material stay for extended time at 500°C to 600°C range during the process. This is why I would like to construct a solidification simulation to screen it to see if it will form during solidification or not (if it is possible to do this anyway). In addition, within the temperature range for the given concentration where L12 phase is stable, the disordered FCC is not stable, it decomposes to L12 and another B2 composition.

I read some of the old threads in the forum and tried to amend the phase description by setting the major constituents for the composition set L12#2. I set Ni and Al as main elements in the sublattices of L12#2 and defined the phase as ordered in micress concentration solver, stating that it is a different compsition and volume set with respect to disordered FCC. I see in the log file that start composition values sset for this phase as well, yet I get the same error over and over again.

I hope the the input is sufficient to proceed :)

Regards,
Ahmet

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: error #40300 and 105 while initiating Liquid/FCC_L12#2 interface

Post by Bernd » Mon Feb 28, 2022 4:10 pm

Dear Ahmet,

If you try to initialize any phase interaction with FCC_L12#2 at a temperature above the critical point/decomposition temperature, it is impossible to get a stable equilibrium, because the FCC_L12#2-description will always switch into disordered FCC_L12 composition, and you get an error (because you required correctly that FCC_L12#2 should be ordered).

This means that the ordered phase cannot be stable at 1588K. If at all, this phase can be stable only at lower temperature. Hence you should try to start checking for nucleation at lower temperature.

Bernd

Post Reply