importing the initial microstructure

[Mehnoush]

Hi Bernd,

I use MICRESS 6.3 and try to simulate transformation of a domain with ferrit and cementite grains to a domain mainly composed of austenite grains. i have succesfully done this simulation for a artificial domain (manually drawn). In the next step i want to us REM images from the cross-section of the initial state to run the simulation. i coud import the geometery which i also attached here below but ican not assign the properties of the phases to the designated area (cementite and ferrite).
what i have done:
1)use REM images
2)fiter the REM images to determine ferrite from perrite
3)cut them and skale them using Gimb and imprting as *tiff
4)using Imagej sofware to convert them from *tiff to text data
can you please reffer me to an example were a real rem image is used?

[Bernd]

Dear Mehnoush,

Looking at your imported microstructure in the attached Word document I have the impression that your import worked fine: You have assigned different regions of the structure with different phase numbers. What do you mean by "ican not assign the properties of the phases to the designated area"? When you have assigned the phase number, all the phase data specified in the input file are automatically assigned. Is your problem perhaps how to distinguish different grains?

PS: You should not use phase number 0 for ferrite because phase 0 has a specific meaning of an isotropic background phase (liquid). You won't be able to define different grains with different orientations.

Bernd

[Mehnoush]

Hi Bernd,
thanks alot for your answer.
i didn't know how i can define different phases. in better words how to tell wich grain is for example ferrit and wich one is perrite .
so i made the second *txt file to define the grain properties (for example: phase number, direction, etc ). but it does not work properly my problem is concentartion coupled and is transformation of ferrit and austenite to perrit. i attach the file that shows the phases after defining the grain propeties (second *txt file). i don't know what is wrong in the second *txt file that changes the geometery
thanks alot.
Mehnoush

[Bernd]

Dear Mehnoush,

Can you please show us your second text file, and also the input file (at least the part where you define what you read from file). Otherwise we cannot know what you are doing...

Bernd

[ralph]

Hi Mehnoush,

see the grain growth example 'T020_Grain_Growth_InitialFromFile':

T020_Grain_Growth_Microstructure.txt
# Treatment of data?
# (n: none, 1: 1D, f: flip (bottom<->top), t: transpose,
# or p: 'phase to grains transformation')
fp
# CellsX for initial microstructure?
500
# CellsZ for initial microstructure?
400
# Number of grains at the beginning?
# (Set to less than 1 for the number of grain to be read from the input data,
# with optionally a minimal size, in cells)
-1
# value 765: grains 1 ... 53
# Number of grains found in input data: 53
# Read grain properties from a file?
# Options: input from_file identical blocks


If you use the phase to grains transformation, you get some usefull information about the found phases/values in your grain input file:
# value 765: grains 1 ... 53
# Number of grains found in input data: 53

You can use this information to assign properties blockwise, i.e. here per phase:
block
<starting grain id for block 1>
phase
orientation
etc.
<starting grain id for block 2>
...

Hope that helps a little.

Best,
Ralph

[Mehnoush]

Bernd and Ralph Thanks alot i will try it and inform you if i had worked oe not

[Mehnoush]

hi ,
i could solve the problem with help from a colleague of me. the point was that i should gotten rid of the noises in the picture. the noises are the small pixells that may have ligh cream or grau color in for example the with area (see the pic). every small pixel with the grow color is considered as a grain in the MICRESS. and the other mistake of me was that the cementite grains were black at supposedly MICRESS considered it as grain boundary.

[ralph]

That's good news.
You are right, every distinct value will be read as a new phase (phase to grains), resp. grain number. There is no segmentation according colors done.
If I remember correct, the minimal value will always be seen as interface, e.g. in the grain growth example which reads from an initial microstructure file (phase to grains) the interface cells have the value -1.

Best,
Ralph

[Mehnoush]

hi
concerning the same problem. as you can see in the attached file. there are lots of small grains in the perrite lammela. is there any option that i can merge this small grains? or can you give me a advice to solve this problem?

BR,
Mehnoush

[ralph]

As far as I know, there is no way to merge such grains afterwards.
There might be some tricks while preparing the input, i.e. to draw a line to connect these grains. However, that's only possible if they have same input values at all and belong to the same phase.

Best,
Ralph