Re: Discrepancy in Misorientation Distribution Function (MDF) from Real EBSD Data and Synthetic microstructures using Dream3D

[Michael Jackson]

Dear Aayush, 

   DREAM.3D will do its best to generate a microstructure that matches the input statistics. Sometimes though those input statistics prove to be impossible to recreate during the synthetic generation. While I agree that something does not look correct with the MDF, without the original input and and sufficient time to track the problem I’m not sure what to suggest. Review the data, review the MatLab script to ensure no errors. You might have found an inconsistent output from within the StatsGenerator algorithms.

--
Mike Jackson                    mike.j...@bluequartz.net
BlueQuartz Software         www.bluequartz.net

President/Owner               Dayton, Ohio

Principal DREAM.3D Developer

On Sep 12, 2023 at 00:50:00, Aayush Trivedi <aayusht...@gmail.com> wrote:

Ok, I will rephrase the question. So, I used StatGenerator filter -> MDF for misorientation and as given in MDF_plot.png, I gave this misorientation. But when I outputted the microstructure, I could not see this given misorientation there. I have used the prebuilt pipeline "Single Hexagonal Phase Equiaxed" which is already given in Dream3D. So, I wished to know if I am missing some appropriate filters? I have been exploring other filters based on responses here, but am unable to get the correct misorientation in the microstructures.
Any help is appreciated.

On Tuesday, 5 September, 2023 at 10:21:59 am UTC+5:30 Aayush Trivedi wrote:

Greetings,

I am currently engaged in generating a synthetic microstructure derived from an EBSD dataset, utilizing the "StatsGenerator" filter within Dream3D. The procedure involved the calibration of grain size through the size distribution function, followed by the implementation of ODF data, which was sourced from the original EBSD file. For subsequent analysis, I converted the Dream3D output into .ang format using a Python script and was successful in reconstructing the microstructure in MTEX. For visualization of these stages, kindly refer to figures: D3D_vtk_0_paraview.png and vtk_to_ang_mesh_MTEX.jpg.

A critical issue has emerged during the comparison of Misorientation Distribution Function (MDF) plots against frequency. Notably, the MDF obtained from the synthesized microstructure markedly contrasts with the MDF from the actual EBSD data, as evidenced in EBSD_Misoreintation_plot_frequency.jpg. I also attempted to employ the MDF function within StatsGenerator in Dream3D, as showcased in MDF_plot.png, yet the divergence remains evident. The series Sample_*.png encapsulates the varied outputs derived from Dream3D output files.

Please provide me with recommendations, or methodologies that could potentially address and resolve this observed discrepancy. Any insights or suggestions to address this inconsistency would be greatly appreciated.

Thank you for your assistance.

Best Regards,
Aayush

--
You received this message because you are subscribed to the Google Groups "dream3d-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dream3d-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dream3d-users/39c67d51-4f56-446f-a5c1-7d172d19864cn%40googlegroups.com.