Partition-no-partition transition temperature during pearlite-to-austenite transformation

solid-solid phase transformations, influence of stresses and strains
Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by Bernd » Mon May 23, 2022 11:12 pm

Dear Pina,

That sounds good. If I understand you well there is only one problem left, namely that the PNTT-temperature is not quantitatively correct, right?. And do you think 2K is a lot?

The next question is where it could come from. Could it be that your theoretical value has uncertainties? Or could it be a curvature effect, e.g that cementite shrinkage is favored by its curvature?

Another possible source of deviations is that the C-composition in ferrite could be out of equilibrium at the beginning. You could hold some time for equilibration or reduce the heating rate, but there also remains an intrinsic problem at the grain boundaries between the colonies, where you cannot have full equilibrium.

There could also be an inexactness of the MICRESS nple-model. Perhaps you just need to update thermodynamic data more often? Maybe you should show me the input file (attached or as PM), perhaps I find something else which could be the reason.

Bernd

pmecozzi
Posts: 23
Joined: Mon Apr 15, 2013 3:20 pm
anti_bot: 333

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by pmecozzi » Wed May 25, 2022 3:17 pm

Dear Bernd

I did again the PNTT calculation, this time I did it not including Si, which is in the alloy but was not in the MICRESS simulations. The new PNTT value is lower than the previous one (with Si) and equal to 1021 K. Maybe, this new result can explain the growth of austenite in colony 3.

In addition, I repeated the simulation by changing the redistribution behaviour of Mn from “LENP” to “ParaTQ” and, in another one, to “Normal”. In the attached PowerPoint file there are the phase maps at different times in the three thermodynamic conditions. I understood that the different driving forces induced different kinetics. However, the behaviour of the growing austenite in the simulation with “ParaTQ” surprised me. Some austenite grains, that started to grow at 1021 K, shrank at the next step (see blue arrows in the second and third slide). Can this behaviour be explained by the fact that happens at a temperature close to the Partition/No-partition Transition Temperature?
In addition, do you know the reason for the growth of cementite lamellae in simulation with “ParaTQ” (red arrows in the last slide)?

Kind Regards,

Pina
Attachments
Phase_morphology for ACCESS.pptx
(180.29 KiB) Downloaded 147 times

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

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by Bernd » Wed May 25, 2022 8:30 pm

Dear Pina,

Using "normal" in cases with negligible diffusion of substitutional elements typically leads to undefined behavior which is somewhere between "nple" and "para". In so far, the comparison between "nple" and "normal" shows the correct trend, while "paratq" should be similar (or even faster) than "normal".

When using "paratq" you must be aware that this model drastically changes the equilibrium condition itself, while "nple" and "para" only change the redistribution inside MICRESS. Thus, "paratq" can lead to inconsistencies at triple junctions if you use "paratq" not for all phase interactions which are involved in this triple junction. If you did do that, this could explain the unexpected behavior of "paratq"...

Bernd

pmecozzi
Posts: 23
Joined: Mon Apr 15, 2013 3:20 pm
anti_bot: 333

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by pmecozzi » Fri May 27, 2022 10:45 am

Dear Bernd,

Many thanks for the clarification.
I realised that I set the “paraTQ” option only for the γ/α and Υ/θ interactions but not α/θ interaction.
I repeat the simulation setting “paraTQ” for all three interactions. However, the austenite growth is still weird (see the attached pictures). I will send the driving file as a private message.

Kind Regards,

Pina
Attachments
ParaTQ_0.08s.png
ParaTQ_0.08s.png (15.91 KiB) Viewed 5045 times
ParaTQ_0.06s.png
ParaTQ_0.06s.png (14.21 KiB) Viewed 5045 times
ParaTQ_0.05s.png
ParaTQ_0.05s.png (14.21 KiB) Viewed 5045 times
ParaTQ_0.04s.png
ParaTQ_0.04s.png (13.02 KiB) Viewed 5045 times

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

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by Bernd » Tue May 31, 2022 4:45 pm

Dear Pina,

I still do not understand why you want to apply the higher Mn composition only to the lamellar region and not to the whole area of the colony - what is the reasoning behind?

Apart from that, I found a problem in the linearisation parameter sets of fcc/cementite and bcc/cementite with paraTQ (unrealistic Mn composition in cementite, huge values for dS2 and dcdT/2), I need to have a look at that. Unfortunately, I am a bit busy right now, but I will come back to you as soon as possible...

Bernd

pmecozzi
Posts: 23
Joined: Mon Apr 15, 2013 3:20 pm
anti_bot: 333

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by pmecozzi » Wed Jun 01, 2022 10:29 am

Dear Bernd,

Many thanks for your support. Please take all time you need to solve the problem of linearization. It is not urgent.
As regards your question about Mn concentration in cementite lamella, I took the values from some experimental data (see the attached picture); then, I assumed some inhomogeneity in Mn redistribution during pearlite formation.

Kind Regards,

Pina
Attachments
Initial concentration.pptx
(821.55 KiB) Downloaded 124 times

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

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by Bernd » Wed Jun 08, 2022 8:13 pm

Dear Pina,

Meanwhile, I intensively checked your problem with the "paratq" model, and I came to the conclusion that there is a fundamental problem: This model uses the tqgdf2 function of the Thermo-Calc TQ-interface which also allows for calculation of para-equilibrium. However, this model requires the definition of a matrix composition, while the second phase is treated like a precipitate. But in cementite C is fully stoichiometric (i.e. no solubility range for C), which causes problems when defining it as matrix for the backward direction.

Thus, my current understanding is that "paratq" generally does not work together with stoichiometric phases, at least not in the symmetric way we implemented that in MICRESS. The problem is the backward linearisation step which is not possible under para-equilibrium conditions.

Thus, my suggestion in your case is to use "paratq" only for fcc/bcc, while using "para" for fcc/cementite and bcc/cementite. Alternatively, you could also use "para" for all three interfaces.

Bernd

pmecozzi
Posts: 23
Joined: Mon Apr 15, 2013 3:20 pm
anti_bot: 333

Re: Partition-no-partition transition temperature during pearlite-to-austenite transformation

Post by pmecozzi » Tue Jun 14, 2022 11:25 am

Dear Bernd,

Many thanks for the clarification. I will keep it in mind also for future simulations.

Kind Regards,

Pina

Post Reply