Recrystallisation

ripening phenomena, dislocations, grainboundary topology
Post Reply
Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Recrystallisation

Post by Bernd » Wed Nov 25, 2009 7:22 pm


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

Re: Recrystallisation

Post by OksanaOM » Thu Sep 15, 2016 9:14 am

Dear Bernd,

The simulation of microstructure evolution due to forming process is partially possible in MICRESS by means of application of Temperature-Time profile as a boundary conditions. The Temperature-time profile in this case is taken from FEM or Experimental measurements.

I was wondering if the dislocation density profile could be used in MICRESS in the similar manner. As it is written in Manual, dislocation density can be considered for the recrystallization model rather as initial parameter (be read from file ). The point is, if there is a way to prescribe its evolution during time.

I would be very gratefull for your answer.

Best regards,
Oksana

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Recrystallisation

Post by Bernd » Mon Sep 19, 2016 12:08 pm

Dear Oksana,

in the recrystallization model in MICRESS, dislocation densities are used in a similar way as stored energies. As you said, you can read in an initial dislocation density distribution or assign per grain values in the initial microstructure. During the recrystallization process, the dislocation densities constitute the driving force and are altered by growth of new dislocation free grains.
Apart from that behavior, there is no evolution equation available in MICRESS which would describe changes of dislocation density due to ongoing deformation or relaxation processes e.g. like during dynamic recrystallization. I think it would be interesting to have that in future, we already have thought into that direction...

At present there are only "tricks" how the dislocation density can be changed during run time:
1.) The nucleation model allows assigning dislocation densities or stored energies to newly formed grains (the default is 0).
2.) The nucleation option "add_to_grain" allows a redefinition of the dislocation density or stored energy. In principle, this permits to alter these values by step-wise "nucleation" events.

I don't know if this information can be helpful for you. In principle, we could also implement some simple time evolution laws like exponential decay or linear change if this makes sense. I would like to hear your opinion on that.

Best wishes
Bernd

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

Re: Recrystallisation

Post by OksanaOM » Thu May 11, 2017 4:02 pm

Dear Bernd,

I was wondering if it is possible to distinguish in the
# Recrystallised fraction output? ('TabR')
between the fraction recrystallised due nucleation and fraction recrystallised due to grain growth.

It seems that the result in the table relates only to the fraction recrystallised due to nucleation, the fraction due to grain growth is not consedered.

I would be grateful for your answer.

Best regards
Oksana

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Recrystallisation

Post by Bernd » Fri May 12, 2017 5:02 pm

Dear Oksana,

you are right: The recrystallized fraction information in the .TabR output gives you the fraction of grains with reX energy=0, which typically have been formed by nucleation.
Describing recrystallisation by the movement of interfaces between deformed grains was not the basic idea of our reX model, but is also possible. In this case, at least in the present implementation no region with reX energy=0 is formed, so this "grain growth" process will not change the recrystallized fraction in .TabR. However, there is also information about the distribution of reX energy classes which should be altered by grain growth of deformed grains, because the grains with higher reX energy are shrinking while those with a lower reX energy are growing.
To be honest, I personally did not work with this type of reX modelling, so my understanding is limited.

If you still have doubts about the concrete meaning of your TabR output, it may be helpful to check again with a concrete application case.

Bernd

deepumaj1
Posts: 104
Joined: Fri Feb 13, 2015 12:00 pm
anti_bot: 333
Contact:

Re: Recrystallisation

Post by deepumaj1 » Mon May 07, 2018 7:57 am

Hello Bernd,
Can you give an example of a driving file input (the nucleation data region), where the add_to_grain option can be used to modify the stored energy value as you said in the post below? I was not successful in doing this. I tried this option by giving a specific grain number. The stored energy for that grain got modified only when the new stored energy was zero. It was not getting modified for other values of stored energy. Thanks.

Kind Regards,
Deepu

Bernd wrote:Dear Oksana,

in the recrystallization model in MICRESS, dislocation densities are used in a similar way as stored energies. As you said, you can read in an initial dislocation density distribution or assign per grain values in the initial microstructure. During the recrystallization process, the dislocation densities constitute the driving force and are altered by growth of new dislocation free grains.
Apart from that behavior, there is no evolution equation available in MICRESS which would describe changes of dislocation density due to ongoing deformation or relaxation processes e.g. like during dynamic recrystallization. I think it would be interesting to have that in future, we already have thought into that direction...

At present there are only "tricks" how the dislocation density can be changed during run time:
1.) The nucleation model allows assigning dislocation densities or stored energies to newly formed grains (the default is 0).
2.) The nucleation option "add_to_grain" allows a redefinition of the dislocation density or stored energy. In principle, this permits to alter these values by step-wise "nucleation" events.

I don't know if this information can be helpful for you. In principle, we could also implement some simple time evolution laws like exponential decay or linear change if this makes sense. I would like to hear your opinion on that.

Best wishes
Bernd

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Recrystallisation

Post by Bernd » Mon May 07, 2018 5:33 pm

Dear Deepu,

Here is a typical nucleation condition for adding reX-energies during run-time:

...
# Input for seed type 3:
# ----------------------
# Type of 'position' of the seeds?
# Options: bulk region interface triple quadruple [restrictive]
interface
# Phase of new grains (integer) [unresolved|add_to_grain]?
1 add_to_grain
parent_grain
# Reference phase (integer) [min. and max. fraction (real)]?
1
# Substrate phase [2nd phase in interface]?
# (set to 1 to disable the effect of substrate curvature)
1
# maximum number of new nuclei 2?
-1
# Grain radius [micrometers]?
0.
# Choice of growth mode:
# Options: stabilisation analytical_curvature
stabilisation
# critical recrystallisation energy [J/cm**3 or MPa]?
-1000.
# Determination of nuclei orientations?
# Options: random fix fix_direction parent_relation
parent_relation
# Minimal value of rotation angle? [Degree]
-0.000
# Maximal value of rotation angle? [Degree]
+0.000
local
0. 1. 0.
# Shield effect:
# Shield time [s] [shield phase or group number] ?
0.
# Shield distance [micrometers] [ nucleation distance [micrometers] ]?
1.
no_categorize
# Input of minimal and maximal energy of the nuclei:
# Minimum of recrystallisation energy? [J/cm**3 or MPa]
5.0000
# Maximum of recrystallisation energy? [J/cm**3 or MPa]
50.0000
# Nucleation range
# min. nucleation temperature for seed type 2 [K]
0.
# max. nucleation temperature for seed type 2 [K]
3000.000
# Time between checks for nucleation? [s]
constant
0.1
# Shall random noise be applied?
# Options: nucleation_noise no_nucleation_noise
no_nucleation_noise
#
...



It is done in such a way that grain numbers and orientations are not changed, only reX energy. An interface seed type has been chosen because it is more economic. Please tell me if you have doubts.

Bernd

deepumaj1
Posts: 104
Joined: Fri Feb 13, 2015 12:00 pm
anti_bot: 333
Contact:

Re: Recrystallisation

Post by deepumaj1 » Tue May 08, 2018 2:58 am

Hi Bernd,
Thanks for the input. It worked.

I have another question. In case of recrystallization, lets say, I have multiple types of nucleation (say nucleation at the interface and bulk). Since the phase number is same everywhere, how can I ensure that the new 'interface' nucleation is not going to happen at the interface of the recrystallized grains present in bulk?

Deepu

Bernd
Posts: 1074
Joined: Mon Jun 23, 2008 9:29 pm

Re: Recrystallisation

Post by Bernd » Tue May 08, 2018 4:26 pm

Dear Deepu,

This what you describe will not happen by default. When nucleating a recrystallized seed at an interface between identical phases (of the same phase as the seeding phase), the smaller value of the recrystallisation energies of both grains is taken and compared with the threshold energy defined in the phase data. If it is lower (which is the case if one of the two adjacent grains is already recrystallized), no nucleation happens.
In fact, nucleating on an interface between deformed and recrystallized grains does not make sense in the frame of the standard reX-model. However, if you like, you can change this behaviour by adjusting the reX-energy of the recrystallized grains and/or the threshold value accordingly.

Bernd

Post Reply