"Grain radius" in grain input (v 6.402)

ripening phenomena, dislocations, grainboundary topology
Post Reply
billyzhangubc
Posts: 25
Joined: Fri Feb 01, 2019 8:53 pm
anti_bot: 333
Location: University of British Columbia

"Grain radius" in grain input (v 6.402)

Post by billyzhangubc » Fri Mar 06, 2020 3:28 am

Hi all,
The "(minumum/maximum) grain radius" in the random option of grain input seems a bit misleading.

For example, I would like to generate a structure of 5 round grains with a mean diamater of 80 μm (with periodic boundary conditions). I did a quick calculation and a box of 160x160 μm will be able to accommodate them theoretically. However, when I put in 40.00 for "min and max radius" though the structure will have "liquid" and the .TabK file gives an average radius of around 20 (based on area fraction). When I put in 80.00 for "min and max radius" then the structure is perfect and the average radius is what I want. An example input file (v 6.402) is attached.

I suspect that the term "grain radius" is misleading and the input values are actually taken as the diameters of the grains.

I wonder if this issue has been taken care of in 7.000, because from the document online it seems that the input file prompt is still "grain radius". I have did quite a bit of my work with version 6.402 so I'm a bit reluctant to migrate to 7.000.

Thanks,
Billy
Attachments
trial_20.in
(17.06 KiB) Downloaded 266 times

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

Re: "Grain radius" in grain input (v 6.402)

Post by Bernd » Sun Mar 08, 2020 5:19 pm

Hi Billy,

I don't think that this is unreasonable. The min/max radius which you input means that MICRESS will randomly select center positions and set spherical particles with the given radius. However, the area of the grain will be less than PI x r2 if grains overlap with others.
The "average" radius in .TabK is determined from the area. Thus, if grains are overlapping, the radius given in .TabK will be lower than what you specified as radius range in the input file.

You can test that easily with your example. If you select a grain radius range (min/max) between 9 and 11, you get a pretty good value around 10 in the .TabK because there is no overlap. However, in your attached case with radius between 77 and 83, the area of the domain is fully covered. Thus, in .TabK you find r_avg=SQRT(160x160/5/PI)=40,37. That is what it should do, isn't it?

Best wishes
Bernd

Post Reply