How does 'cut' in Neper deals with grain being cut at the boundaries?

[hdt5kt]

Hi,

I am using Neper to generate a synthetic crystal structure to study some algorithms for stitching actual HEDM scans where multiple scans are required. To do so, I first generated a "true" Neper crystal structure and perform the cut. I have some questions regarding how Neper handle the grain the are being cut into two grains when cut is perform.

In particular,

1. I first generated a random crystal structure as this (I have some scripts to handle automatic parsing and crystal type generations so I apologize for the long command line:

neper -T -id 42 -dim 3 -domain cube(20.0,20.0,20.0):translate(-10.0,-10.0,-10.0) -oridescriptor euler-bunge -o cpfe_aps/hedm_study/test/voronoi -statcell id,vol,w,x,y,z,radeq -format tess,geo,ori -morpho gg(3.0) -morphooptistop iter=10 -n from_morpho

--> this gives me 400 cells after 10 iterations (I limited for demonstration)

2. Then to perform the cut (there can be arbitrary amounts of cut, but let's just cut in half for demonstration purposes

Cut 1:

neper -T -n 400 -id 42 -domain cube(20.0,20.0,20.0):translate(-10.0,-10.0,-10.0) -morphooptiini file(cpfe_aps/hedm_study/test/voronoi.tess) -oridescriptor euler-bunge -transform cut(hspace(0.0,0,0,1),hspace(10.002,0,0,-1)) -o cpfe_aps/hedm_study/test/hedm_scan/raw_files/scan_0 -statcell id,vol,w,x,y,z,radeq -format tess,geo,ori

--> this gives me 208 cells

Cut 2:

neper -T -n 400 -id 42 -domain cube(20.0,20.0,20.0):translate(-10.0,-10.0,-10.0) -morphooptiini file(cpfe_aps/hedm_study/test/voronoi.tess) -oridescriptor euler-bunge -transform cut(hspace(10.002,0,0,1),hspace(-0.0,0,0,-1)) -o cpfe_aps/hedm_study/test/hedm_scan/raw_files/scan_1 -statcell id,vol,w,x,y,z,radeq -format tess,geo,ori

--> this gives me 192 cells

As you can see, if I simply join these two files together, I get back 400 cells (of course, the centroid locations and radius of the cells at the boundary are not the same so indeed the cell centroids are being updated).

What I do not understand is why I still get back 400 cells. For me intuitively, if you cut the cells in half, you would generate one more cells. Therefore, if I naively sum up the two cells for these cuts, I should have more cells than the original pre-cut crystal structure.

Any insights on this matter is greatly appreciated.
Thank you!

[rquey]

That's a good observation. Since -transform "cut(...)" uses mirror seeds for performing the cut, it initially considers only the interior seeds. It will therefore generate cells only for those seeds. That's why you get only 400 cells in total.

For a "true" cut, use -transform "planecut(...)".

[hdt5kt]

Hi,

Thank you for your quick response. I have given planecut a try but I got error: "Error : You have discovered a bug in Neper! Please file an issue at https://github.com/rquey/neper/issues. Thank you.

Here is my command:, I have tried both

neper -T -loadtess cpfe_aps/hedm_study/test/voronoi.tess -domain cube(20.0,20.0,20.0):translate(-10.0,-10.0,-10.0) -oridescriptor euler-bunge -transform planecut(0.0,0,0,1) -o cpfe_aps/hedm_study/test/hedm_scan/raw_files/scan_0 -statcell id,vol,w,x,y,z,radeq -format tess,geo,ori

and

neper -T -n 400 -id 42 -domain cube(20.0,20.0,20.0):translate(-10.0,-10.0,-10.0) -morphooptiini file(cpfe_aps/hedm_study/test/voronoi.tess) -oridescriptor euler-bunge -transform planecut(0.0,0,0,1) -o cpfe_aps/hedm_study/test/hedm_scan/raw_files/scan_0 -statcell id,vol,w,x,y,z,radeq -format tess,geo,ori

I suppose I should file a "bug" report?

Thank you

[rquey]

You may have to re-run the full command including the cut, i.e. without using -loadtess. I'm not sure.

If it doesn't work, please provide a full bug report (files + terminal output) so that I can reproduce the bug.

[hdt5kt]

Yeah I have tried including the cut in the full command but still got the bug error. I will create a bug report of the minimal working example.

Thank you.