Merging multiple observations of the same grain

[jadball]

Hello,

I have a box scan stitching algorithm that finds grains that have been observed multiple times across multiple 3DXRD scans. Do you recommend a way to merge the observations of multiple grains together into a single grain, using the grain data in a map file? At the moment, I am doing a volume-weighted average of the UBI matrices of the contributory grains, but the strain data of those grains looks untrustworthy. This is related to #78 too

Thanks in advance!

[jonwright]

Hell James,

Are these data a series of letterboxes as a function of height in the sample? You could compute the strain and orientation for each slice and then make weighted averages of those values instead. If there is an orientation gradient the I guess that could mess up the strain otherwise.

For the tomographic problem then you probably need to look into Axel's code...

Best

Jon

[jadball]

Hi Jon,

Yes, they're a series of letterboxes as a function of height. My classes for storing per-grain data use your grain class under the hood, which is nice because the strain etc are all calculated properties and can't be accidentally overwritten. I suppose I could average the strain and orientation of all the observations of a grain, then calculate the UBI from U and eps, then initialise the merged grain with the UBI?

Best,

James