Categorisation

[chika]

Dear all,

I have an additional issue regarding the categorization of nucleated phases.

I defined the nucleation conditions using the code below, and the phase of newly nucleated grains (no.1) is defined as "categorize" in the section of "Phase properties".
I expected all nucleated grains to be categorized after nucleation, but some of them are not.
The proportion of uncategorized grains is a few percent.

If anyone has any ideas about the cause of this issue, I would appreciate your suggestions.


# Nucleation
# ==========
# Enable further nucleation?
# Options: nucleation nucleation_symm no_nucleation [verbose|no_verbose]
nucleation
# Additional output for nucleation?
# Options: out_nucleation no_out_nucleation
no_out_nucleation
#
# Number of types of seeds?
1
#
# Input for seed type 1:
# ----------------------
# Type of 'position' of the seeds?
# Options: bulk region interface triple quadruple front [restrictive]
triple
# Phase of new grains (integer) [unresolved|add_to_grain|split_from_grain]?
1
# Reference phase (integer) [min. and max. fraction (real)]?
3
# Substrate phase [2nd substrate phase]?
2
# Maximum number of new nuclei of seed type 1?
# (set negative for unlimited number)
300
# Grain radius [micrometers]?
5.000000E-02
# Choice of growth mode:
# Options: stabilisation analytical_curvature
stabilisation
# min. undercooling [K] (>0)?
1.000
# Shield effect:
# Shield time [s] [shield phase or group number] ?
0.0001
# Shield distance [micrometers] [ nucleation distance [micrometers] ] ?
20.0
# Nucleation range
# min. nucleation temperature for seed type 1 [K]
1000.000000000
# max. nucleation temperature for seed type 1 [K]
1200.000000000
# Time between checks for nucleation? [s]
# Options: constant from_file
constant
# Time interval [s]
0.00005
# Shall random noise be applied?
# Options: nucleation_noise no_nucleation_noise
no_nucleation_noise
#
# Max. number of simultaneous nucleations?
# ----------------------------------------
# (set to 0 for automatic)
1
#
# Shall metastable small seeds be killed?
# Options: kill_metastable no_kill_metastable
kill_metastable

Best regards,
Chika

[Bernd]

Dear Chika,

The "categorize" option was never meant to completely join all grains with identical properties, but only to considerably reduce the number of grains for performance reasons. The reason is that joining existing grains in some cases is not trivial, especially if they already have formed interfaces among each other.

If you want to systematically create several instances of the same grain, you may rather choose the "add_to_grain" option with suboption "new_set" in the nucleation input.

Bernd

[chika]

Dear Bernd,

Thank you so much for your suggestion.
I am going to try the method which you suggested.

Best regards,
Chika

[chika]

Dear Bernd,

I have tried the calculation with choosing the "add_to_grain" option with suboption "new_set" in the nucleation input.
As a result, The following error happened and the calculation stopped at the initial time-step of nucleation.
I would appreciate it if you have any ideas about the cause of this error.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Warning in spfAddPhInkr, t=0.5074300E-01
Wrong fractions at nTupelp= 4063
sum = 0.000000
x,y,z = 180 1 716
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Best regards,
Chika

[Bernd]

Dear Chika,

This looks bad. Such errors indicate that the interface list structure has broken. This may happen sometimes if numerical parameters are very odd, producing huge phase-field increments. Another typical reason is if initial microstructures are read from a restart file in an inconsistent way (e.g. using periodic boundary conditions on a non-periodic structure). In connection with nucleation, I could also imagine that multiple nucleation at the same point (in the same time step) under certain conditions could lead to such a problem.

If you are not able to resolve the problem by yourself, please send me your complete set of input files (.dri-file, .ges5, etc.) per mail or as attachment, so that I can try to reproduce and solve the issue.

Bernd

[chika]

Dear Bernd,

Thank you so much for your comment.
I have send you the related files by private message.

Best regards,
Chika

[Bernd]

Dear Chika,

I cannot test your case because the relevant input files are missing, but I guess that your problem is linked with the big grain radii which you use for nucleation. Given the small grid size of 0.01 µm, a grain radius for nucleation of new seeds of 0.05 is huge. We typically use a radius of 0 ("small grain") or a value below the grid spacing size in order to start with the correct composition of the new phase. The only exception perhaps is same-phase nucleation in recrystallisation where the composition is the same or not taken into account.

In cases of "add_to_grain" with suboption "new_set" like yours, I would recommend to use a value between >0 and Δx (e.g. 0.005 µm) to help initial growth, because here the identity of individual seeds is lost and the "small-grain" stabilisation model does not work as good. Then the initial seed consist of a single cell with a fraction of the new phase between >0 and <1.

Bernd

[chika]

Dear Bernd,

Thank you so much for important suggestion!
Calculation run successfully by setting nuclei-radius to 0.005um.

Best regards,
Chika