Solid state transition

solid-solid phase transformations, influence of stresses and strains
Post Reply
shaojielv
Posts: 100
Joined: Thu Apr 01, 2021 2:16 pm
anti_bot: 333

Solid state transition

Post by shaojielv » Thu May 27, 2021 10:54 am

Dear Bernd,
1. In solid-state phase transitions, there are several ways for each element to be redistributed, such as:Nple, para, paratq, normal [mob_corr], atc [mob_corr], etc., but in the process of learning, I did not see the explanation of the allocation mode in the manual, so I still have a lot of doubts about the redistributed mode, and I hope to get your answer.
2. In case T012, I ran the case given on the official website, but after time t=90s, the result of the phase changed from round to polygon, which became more and more strange as time went on (in my opinion), and there were a few blue dots in the phase. I didn't know whether this was the phase boundary or not.In addition, why did the phase transition begin to change very slowly (basically unchanged) from time t=20s to time t=80s? Why does this process have all of these phenomena?
3. Case T012 selected Para as the redistribution method of manganese element, and I adjusted the redistribution method of Mn element to Normal, but the results made me very confused.As shown in the figure, I have compared the simulation results under different redistribution methods, and the simulation results are very different. I hope you can give me a professional explanation.
lvshaojie
Attachments
T012_Gamma_Alpha(question_3).png
T012_Gamma_Alpha(question_3).png (420.19 KiB) Viewed 2632 times
T012_Gamma_Alpha(question_2).png
T012_Gamma_Alpha(question_2).png (456.72 KiB) Viewed 2632 times

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

Re: Solid state transition

Post by Bernd » Fri May 28, 2021 12:30 am

Dear lvshaojie,

1.) All the options you mention are special options for redistribution and available by use of the "redistribution_control" keyword. nple, para and paratq are redistribution options for elements which are diffusing slowly and cannot be spacially resolved in a MICRESS simulation. They are of special interest in steels and have been discussed e.g. here.
"mob_corr" and "atc" are options for thin-interface correction, which is needed to correct kinetics and redistribution for artefacts of the diffuse interface. They have been described here. The underlying equations are published here:

A.Carré, B.Böttger, M.Apel, Implementation of an antitrapping current for a multicomponent multiphase-field approach, Journal of Crystal Growth 380(2013)5–13

I don't understand what you mean with "allocation mode".

2.) This is the normal behaviour. The example presents an isothermal quenching process with a gamma to alpha phase transformation. It is normal that the transformation gets slower and eventually stops under these conditions. Transformation to polygons is natural if particles grow together leaving a narrow gap of the vanishing phase, just like in case of globulitic solidification.
Indeed, there is a numerical issue towards the end of the simulation, which has been already corrected (for version 7.1).

3.) It is not possible to switch the redistribution mode of Mn to "normal" because the pile-up of this slow-diffusing element before the phase front cannot be resolved properly, leading to numerical artefacts. nple is very different to para because most of the driving force is consumed by redistribution.

Bernd

kmukherj
Posts: 21
Joined: Wed Aug 12, 2009 3:22 pm
anti_bot: 333

Re: Solid state transition

Post by kmukherj » Fri Jun 18, 2021 7:01 am

Dear Bernd,

Can you kindly let us know what is the difference between the choices "para" and "paratq" ?

Thank you,
Regards,
Krishnendu

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

Re: Solid state transition

Post by Bernd » Fri Jun 18, 2021 3:50 pm

Dear Krishnendu,

When using "paratq", a para-equilibrium model of Thermo-Calc is used when calculating quasi-equilibrium in MICRESS. That means, paraequilibrium is fully treated within Thermo-Calc, while MICRESS interprets the results as quasi-equilibrium.
"para" is a MICRESS-internal model which changes the redistribution behaviour by adding/substracting a corresponding amount of the component before redistribution (older MICRESS versions used a correction of the phase fractions instead). The condition for "para" is that the element has no contribution to the driving force.

Bernd

shaojielv
Posts: 100
Joined: Thu Apr 01, 2021 2:16 pm
anti_bot: 333

Re: Solid state transition

Post by shaojielv » Fri Jun 18, 2021 5:08 pm

Dear Bernd,
Hello, I would like to ask you some questions. When I calculated the case T010_GAMMA_ALPHA.through the influence of different carbon content on the aging, I calculated that carbon content suppressed the solid phase transition (gamma-alpha).I hope to get your guidance, if correct, then I hope you can help me to explain the principle of carbon content inhibiting transformation.I would appreciate it very much.
I sincerely look forward to your reply. Finally, I wish you a happy life and smooth work.
Lv Shaojie
Attachments
碳浓度的影响.png
碳浓度的影响.png (210.37 KiB) Viewed 2546 times

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

Re: Solid state transition

Post by Bernd » Sun Jun 20, 2021 3:56 pm

Dear shaojielv,

The system behaves like expected, as carbon shifts the equilibrium temperature between austenite and ferrite to a lower value, so that the driving force for phase transformation is decreased.
You could guess that also from the fact that carbon is accumulated in the vanishing phase.

Bernd

shaojielv
Posts: 100
Joined: Thu Apr 01, 2021 2:16 pm
anti_bot: 333

Secondary development

Post by shaojielv » Fri Jul 02, 2021 10:15 am

Dear Admin,
I am very sorry that I cannot reply your message on the original post I set up for the time being, so I replied here. I am sorry for the trouble caused to you.As shown in the following:viewtopic.php?f=24&t=727
So far what I'd like to change is the phase field model itself, the thermodynamic model, the diffusion model and the nucleation model which I think would be a perfect thing to do if the user could also redevelop it. I believe that we will pay more attention to this aspect in the future research.
Lv Shaojie

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

Re: Solid state transition

Post by Bernd » Fri Jul 02, 2021 12:33 pm

Dear Lv Shaojie,

I understand your point. However, this would require an open source development of the software and an extremely modular structure.
In view of performance and user friendlyness this was not our approach.

Bernd

Post Reply