MICRESS Forum

Hi,

Why MICRESS does not ask for critical radius when defining grain inputs using random option.

BR
Chamara

Good question!
We still have to change it. When we first implemented random grain positioning, the critical radius did not yet exist. Later, when we introduced the critical radius, we couldn't agree how to define it in this case and postponed the problem. At present, the critical radius is still simply set equal to the initial radius for each indivually set grain. How would you like to specify the critical radius for your application?
Regards,
Janin

When I use grain input as deterministic, there I can defined a critical radius apart from the grain radius. I was bit surprise why I was not able to do it. I m dealing with a simulation where I need to have 5nm grid spacing. with the interracial energy that I use, undercooling become so large. so I need to have a critical radius. otherwise I have to artificially reduce the interfacial energy. Would be nice If I can define a critical radius when I use the grain input as "random"

Hi Chamara,

The only solution at present is to use nucleation during run-time (e.g. using the seed density model) instead of initial grain setting...

Bernd

PS: ... apart from "dirty tricks" like changing resolution after initial grain setting by using restart from time t=0...

defining a grain radius larger than grid spacing together with "analytical curvature" seems to work. Also I had to change the initial starting temperature so that i get enough underccoling intitally

Yes, you are right, I remember! If you use "analytical_curvature", you always get a "small" grain, and the radius is interpreted as critical radius. That means, the implementation is consistent

Hi Bernd,

What does this option "steps between random values in degrees" do?

Hi Chamara,

this option does the same as "categorize" for nucleation. The random orientations are reduced to a finite number of explicit values which are defined with this "step" value. The number of these explicit values ("categories") is given by the orientation range (e.g. 0-360°) divided by the step value.
Defining the step instead of the number of categories is preferred because the allowed orientation range may also depend on symmetry and therefore is not always transparent for the user.

Bernd

Thank Bernd,

Also is there a way to extract the coordinates of the grains that created when we use the option "random" for "type of grain positioning"

In addition to that, if you are reading grain properties from a file, how should this file looks like? (I am using angle_2d for orientation)

Hi Chamara,

If you use the optional parameter "deterministic_infile", each grain which has been set will appear in the .in file as if you would have placed it deterministically. So you can know the position of each grain, although you need some text script for extracting the data to a table form.
If you read grains from file, you can specify all grain-specific data (like phase number, orientation, reX-energy) in an extra text file. In this file, each property has an own line (3D-orientations have 3 or 4 blank separated numbers in one line), and you may use comments to keep order:

...
# Grain 45:
# ---------
# Phase:
5
# Orientation:
123.56
# ReX energy:
2.357
#
# Grain 46:
# ---------
# Phase:
1
# Orientation:
0.00
# ReX energy:
1.38
...

Bernd