Austenite to pearlite transformation in steel

solid-solid phase transformations, influence of stresses and strains
sudi
Posts: 4
Joined: Wed Sep 14, 2011 6:09 am
anti_bot: 333

Austenite to pearlite transformation in steel

Post by sudi » Sat Nov 12, 2011 4:53 am

Hi,
Although I've got many literature for other types of solid-state transformations using micress, but I havn't got enough material on eutectoid transformations(pearlitic in steel) . How to select appropriate interface mobilities in the case of pearlite evolution? Any information on eutectoid transformation would be helpful.
Thank You,
sudi

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

Re: Austenite to pearlite transformation in steel

Post by Bernd » Mon Nov 14, 2011 2:57 pm

Hi sudi,

there are different ways to treat pearlite in MICRESS, and there are different users who do it already!
If treating pearlite traditionally (i.e. fully resolved), I would assume diffusion-limited growth, although I don't know whether this is really the case. Otherwise one would need experimental information for calibration of the interface mobility. Hamid is one of our forum members who works on this field, as far as I know. There is also literature from our house:
I. Steinbach, M. Apel: „The influence of lattice strain on pearlite formation in Fe-C”, Acta Materialia 55 (2007), pp.4817-4822.

Apart from that, one can treat pearlite as a binary pseudo-phase. This has been done e.g. by kmukherj. There has also been a publication from that group:
J. Rudnizki, B. Böttger, U. Prahl, and W. Bleck, "Phase-Field Modeling of Austenite Formation from a Ferrite plus Pearlite Microstructure during Annealing of Cold-Rolled Dual-Phase Steel", METALLURGICAL AND MATERIALS TRANSACTIONS A 42A (2011) 2516.

Finally, we developed a new model for description of spatially unresolved eutectics as an effective phase (MICRESS 6.0 released), which also may be interesting for simulating pearlite at a bigger scale (without resolving the individual lamellae) and which is available in Version 6.0.

I personally do not have own experiences with pearlite, but I would appreciate if other users of the MICRESS forum shared there experiences here!

Bernd

sudi
Posts: 4
Joined: Wed Sep 14, 2011 6:09 am
anti_bot: 333

Re: Austenite to pearlite transformation in steel

Post by sudi » Thu Mar 22, 2012 9:22 am

Dear Bernd,
I have been trying to simulate pearlite using the conventional approach.However, I am quite uncertain about the types of nucleation seed that would be required for this simulation. Is it appropriate to use cementite as substrate phase in addition to ferrite as the phase of new grains? I have taken growth mode as stabilization.
Yours sincrely,'
Sudi

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

Re: Austenite to pearlite transformation in steel

Post by Bernd » Thu Mar 22, 2012 12:05 pm

Dear sudi,

it is correct to use cementite as substrate phase for nucleation of ferrite at the cementite/austenite interfaces, and ferrite as substrate phase for nucleation of cementite at the ferrite/austenite interface. The choice of the substrate phase (if it is different from the matrix phase) restricts the interfaces for nucleation to those between the substrate and matrix phase, and determines the sign for the additional curvature contribution to the nucleation driving force.

Austenite should be the matrix phase in both cases (i.e. the phase in which the undercooling of the new grains is calculated). Never use cementite as matrix phase, because it is stoichiometic, and no undercooling could be calculated!

Bernd

OksanaOM
Posts: 6
Joined: Thu Jun 23, 2016 12:17 pm
anti_bot: 333

Re: Austenite to pearlite transformation in steel

Post by OksanaOM » Thu Jun 23, 2016 12:42 pm

Dear Bernd,

I would like to ask a qusetion about pearlite formation from austenite. I use this simulation as a preliminary one to simulate the initial ferritic-pearlitic microstructure. As geometrical input for the simulation I use the micrograph of the boron steel, where I artificially prescribe to the perlite region the phase of austenite. Following the procedure of the standart MICRESS example GammaAlphaPearlite_TQ_dri.txt I simulate the cooling and subsequent perlite formation.

My question is how to evaluate the simulation results?

As a result of simulation I obtain the completely new microstructure with the phase of ferrite and the unresolved mixture phase of the perlite and cementime (which I can basically observe only in *frac*.mcr files). The only one criteria to evaluate the results, which comes on my mind is to compare the initial and simulated fraction of cementite. If there is a way to obtain the simulated microstructure visually similar to the input one?

I would appreciate any comment on this problem.

Best regards,
Oksana

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

Re: Austenite to pearlite transformation in steel

Post by Bernd » Thu Jun 23, 2016 8:51 pm

Dear Oksana,

Welcome to the MICRESS Forum.

If I understand correctly, your problem is that you would like to visualize pearlite grains not only as a diffuse cloud of cementite phase fraction. Pearlite grains are regions of the cementite/ferrite eutectoid where the cementite has the same orientation. Therefore, in case of the "unresolved" model, we always write the orientation of the lamella phase to the orientation (.orie) output even when the fraction of ferrite is bigger. Thus, the .orie output will give you a nice view on the pearlite grains if cementite is defined as anisotropic.

I am not sure whether this solves your problem ...?

Bernd

OksanaOM
Posts: 6
Joined: Thu Jun 23, 2016 12:17 pm
anti_bot: 333

Re: Austenite to pearlite transformation in steel

Post by OksanaOM » Fri Jul 01, 2016 3:34 pm

Dear Bernd,

thank you very much for your prompt answer. The visualisation of orientations is very helpfull to evaluate the results. In comparison to the concentration field it allows better to follow perlit formation on the former regions of austenite.

Thank you once more for the hint.

Best regards,
Oksana

OksanaOM
Posts: 6
Joined: Thu Jun 23, 2016 12:17 pm
anti_bot: 333

Re: Austenite to pearlite transformation in steel

Post by OksanaOM » Thu Jul 14, 2016 1:13 pm

Hello Bernd!

I faced another difficulties during my simulation. The Austenite to pearlite transformation simulation works now properly and I would like to consider two other processes: recrystallization and phase transformation from the recrystallized ferritic-pearlitic microstructure to austenite. Thus, I should couple 3 following simulations:
1) Austenite to pearlite transformation -> I obtain the initial ferritic-pearlitic microstructure. (edited GammaAlphaPearlite_TQ_dri.txt file)
2) Simulation of the recrystallization - > as initial data I would like to use the restart file from the first simulation. (edited ReX_mean_dislocation_dri.txt file)
3) Furthr heating and simulation of the phase transformation to austenite. (edited Gamma_Alpha_TQ_dri.txt)

Now I have difficulties with the restart option (with the "reset time"). I was a bit inconsistened and firstly tried the 3. simulation with the results of 1. sim. as initial data. This works without issues.
But recrystallization simulation does not able to read the data from 1. sumulation and I obtain the following error:

# Reading restart file: Results_00_Gamma_Alpha_Pearlite_TQ/GammaAlphaPearlite_T
Qv9FECMN_AnisPh_rest.mcr
# Concentration field read
# Grains field read
# grain properties read
# IFace list read

Unable to read from file Results_00_Gamma_Alpha_Pearlite_TQ/GammaAlphaPearlite_TQv9FECMN_AnisPh_rest.mcr
STOP in routine initRestart
Reading iFace-lists


Could you understand from the given information what can be a problem? I adjusted the number of phases (all are anisotropic), number of components, geometrical parameters, type of coupling...
I would be very gratefull for your help.

Best regards,
Oksana

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

Re: Austenite to pearlite transformation in steel

Post by Bernd » Thu Jul 14, 2016 5:10 pm

Dear Oksana,

nice to hear that you could proceed successfully with the austenite to pearlite transformation simulation.
When using restart without or with the option "reset_time", there are quite a lot of compatibility requirements, because the .rest file is a sequential file which has a different structure depending on the parameters of a MICRESS simulation. If e.g. the number of phases is different, the number and structure of the interface lists is different, and you will not be able to read the .rest file. I cannot know the exact reason in your case without having seen the individual input files, but I imagine that there are several of them. If you e.g. would just take the input file for step 1 and put an additional nucleation type for recrystallization, you would have problems because the .rest file does not contain the information corresponding to the extra seed type.
Fortunately, there is another another option ("structure_only") which removes most incompatibility issues and which probably can solve your problem. I took the chance to add some fundamental information on restart and its options to the "new functionalities" board.

Apart from that, there is also the possibility to include all 3 steps into one single simulation. This may be of advantage if you plan to perform several modified runs of the whole chain. While the temperature-time curve for the whole process can be easily constructed using "connection points", recrystallization due to deformation can be included by introducing stored energy or dislocation density values during run time using the nucleation option "add_to_grain". Furthermore, many numerical parameters can now be controlled as function of temperature or time in order to allow different values during different process steps. But it is up to you whether you want to go for that...

Best wishes
Bernd

OksanaOM
Posts: 6
Joined: Thu Jun 23, 2016 12:17 pm
anti_bot: 333

Re: Austenite to pearlite transformation in steel

Post by OksanaOM » Wed Aug 10, 2016 2:33 pm

Hello Bernd,

I answer with a delay, but wanted to try all proposed by you methods. The best way for my simulation is to create a single smulation, containing pearlite formation, recrystallization and austenitization. In this case there are no problems with data consistensy. I initially defined all nucleation seeds reuired for 3 steps and for the defined step "switched" some of them off by means of unrealistic nucleation conditions (the easiest way is a temperature range). So it needs sveral adjustments, but I would highly reccomend this solution for multistep simulations to avoid the problems with the data transfer for restart.

Thank you for your help.

Best regards,
Oksana

Post Reply