Problems with Insert Precipitates filter

[latm...@gmail.com]

Hello,

I faced some problems with Insert Precipitates filter.

1. Limited volume fraction of precipitates that can be placed into the synthetic microstructure.

In the older versions of Dream.3D (4.x), I was able to place up to ~50-60% of spherical particles. At the same time in Dream.3D 6.x, the maximum volume fraction I have ever achieved was about 10% even though I used many different settings in StatsGenerator. To illustrate the problem, below is an image of particles obtained in Dream.3D ver. 4.x and in Dream.3D ver. 6.x with the same volume fraction, morphology and size descriptors defined in stats file. Perhaps the "problem" is caused by Radial Distribution Function introduced in StatsGenerator starting from ver. 6?

2. To overcome the problem, I tried using Already Have Precipitates function in the filter but it does not seem to work for me too.
For a simple test, I wanted to place a single precipitate at the center of the microstructure (of size 65^3) using a text file with the following contents:

1
2, 32.0, 32.0, 32.0, 4.0, 4.0, 4.0, 0.9, 0.0, 0.0, 0.0

but I did not get any second phase in the generated microstructure. Increasing the number of precipitates in this file did not help either. Could you please suggest what I am missing here?

I attach the pipeline, stats, and precipitates files which I used for microstructure generation with Already Have Precipitates.

Thanks,
Marat

[Sean Donegan]

Hi Marat,

During the transition to version 6 we had tweaked the algorithm for packing precipitates to make the particles "hard", such that they would not overlap.  This has the side effect of reducing the total possible volume fraction that could be occupied by precipitates.  It also ended up pushing a larger fraction of the precipitates out towards the edges of the packing volume as opposed to the center.  We plan on reverting this change soon due to these previous problems.  Unfortunately, until this fix gets in, the filter may have some fundamental limitations on the volume fraction of precipitates it can insert.

For the second issue, the problem was the file format used: the "Already Have Features" options in the "Pack Primary Phases" and "Insert Precipitates" filters need to be single-space delimited with no commas.  The documentation files for these filters was incorrect in this regard (they listed the formats as requiring commas).  I have fixed this error in the documentation.  If you remove the commas from your file, it should correctly insert a spherical precipitate at the center of the box (see attached image as an example I was able to run using your files).

hope that helps,

-Sean

[Sean Donegan]

Hi Marat,

As a quick update, I have reverted the change that made the "Insert Precipitates" filter insert hard particles.  The filter now allows particles to overlap, as in previous versions of DREAM.3D.  This should ensure hitting high volume fractions and also should avoid biases in particle placement.  The nightly builds will have this fix, and can be downloaded here:

dream3d.bluequartz.net/binaries/experimental

-Sean

On Thursday, December 3, 2015 at 9:02:00 AM UTC-5, latm...@gmail.com wrote:

[latm...@gmail.com]

Thank you, Sean!