6th International Congress on 3D Materials Science (3DMS 2022): 3D Data Processing I: Workflows
Program Organizers: Dorte Juul Jensen, Technical University of Denmark; Marie Charpagne, University of Illinois; Keith Knipling, Naval Research Laboratory; Klaus-Dieter Liss, University of Wollongong; Matthew Miller, Cornell University; David Rowenhorst, Naval Research Laboratory
Monday 9:30 AM
June 27, 2022
Room: Columbia A&B
Location: Hyatt Regency Washington on Capitol Hill
Session Chair: Andrew Polonsky, Sandia National Laboratories
9:30 AM Invited
Galaxy Framework for Next-generation Data Reduction Strategies at CHESS: Kelly Nygren1; Matthew Miller2; 1Cornell High Energy Synchrotron Source; 2Cornell University
High-energy synchrotron X-ray techniques allow us to probe polycrystals in 3D non-destructively on timescales that enable mapping and/or tracking evolutions of micromechanical and microstructural states in situ. As we move into the next-generation of instrumentation and analysis and continue to push temporal and spatial limits, datasets are becoming massive (>TBs) and data reduction strategies increasingly complex, often integrating modeling efforts into the data reduction itself. This makes datasets difficult to move, requires special computational needs, and researchers to have requisite knowledge of computing strategies and architectures before meaningful data is extracted. CHESS has adopted the Galaxy Science Gateway framework to abstract away these barriers so scientists can focus on what the software is doing, rather than how it is doing it. This framework lowers the barrier to entry for data reduction and changes the nature of interactions between scientists sharing their data, data reduction pipelines, and data reduction code.
10:00 AM
Dual Energy XCT through Optimised Projection Fusion: Oliver Helps1; Philip Withers2; Alan Clarke3; Timothy Burnett2; Ian Nicholson3; 1University of Manchester & TWI; 2University of Manchester; 3TWI
The increasing use of composite materials as primary structures provides ongoing challenge for non-destructive evaluation (NDE) techniques. For example X-ray computed tomography (XCT) is increasingly being used for industrial inspection. However, the inspection of multi-material wide aspect ratio components using XCT often results in compromises being made in the resultant imaging quality. Typically, as a consequence of the scanning parameters chosen, high density regions can be underexposed or lower density regions overexposed due to the highly contrasting attenuation behaviour of each multi-material layer. This is compounded by large changes in attenuation path length as component is rotated during the XCT scan. This paper presents a novel method for combining projections from multiple scans in order to facilitate improved defect detection and feature recognition within an industrial setting.
10:20 AM
Alignment and Fusion of Multi-sensor Serial Section Data
: David Rowenhorst1; 1Naval Research Laboratory
It has become increasingly common to collect 3D data volumes using different modalities and techniques, which can elucidate unique microstructural features and behaviors, but it also requires the fusing the separate datasets into a single combined reference frame. In this presentation, we will use a set of serial sections from additively manufactured 316L stainless steel and show how serial-sections collected with EBSD, which clearly differentiates the grain structure, are combined with serial sections of backscattered electron images which clearly differentiate the pores within the structure. This is done by using a highly parallel computational method that filters out long-range drift in the serial section alignment parameters and then is adapted to align the two data-stacks while removing high frequency jitter. Only through this aligned dataset are we able to reveal the unique grain structures found near the lack-of-fusion pores within the additively manufactured structure.
10:40 AM Break
11:10 AM Cancelled
A Comparison of 3D Digital Microstructure Generation Techniques Applied to Biological Materials: Carli Marsico1; Donna Guillen2; Dwayne Arola1; 1University of Washington; 2Idaho National Laboratory
Mineralized biological materials are of great interest for next generation materials design due to their ability to exhibit a set of properties difficult to replicate with traditional materials. However, in many of these systems, the underlying microstructure-property-performance relationships are not well understood. Most biological materials are hierarchical requiring multiple imaging modalities to capture the structure. At the extreme ends of the scale, imaging remains expensive and difficult, limiting the ability to generate statistically relevant data sets. The hierarchical nature also introduces a level of complexity to the structure that traditional methods for analyzing microstructures are unable to capture. Recent work from a several groups have investigated various approaches to digitally generate statistically equivalent 3D volumes from 2D exemplars. Here, several such algorithms were adapted to biological materials and compared quantitatively as a first step towards robust characterization of biological microstructures in 3D.
11:30 AM
Development of 3D EBSD Serial Sectioning Workflows for Enhanced Local Orientation Analysis: Gregory Sparks1; Paul Shade2; Michael Uchic2; Stephen Niezgoda1; Simon Mason1; Michael Mills1; Mark Obstalecki2; 1The Ohio State University; 2Air Force Research Laboratory
In this talk, we discuss recent efforts to improve the fidelity of 3D EBSD data collected by serial sectioning experiments. Of specific interest is the incorporation of high-orientation-precision EBSD analysis. Spherical harmonic transform indexing can provide orientation precision of 0.06° or better, but is limited in accuracy by how precisely the diffraction geometry can be specified. On this latter point, we include a discussion of recent developments in pattern-center fitting for improved absolute accuracy of the local crystal orientation. This talk will also cover additional enhancements to the serial sectioning characterization workflow to include high-precision, non-contact surface height measurement by white-light scanning interferometry, potential techniques for estimating the bulk sample misorientation on a section-by-section basis, and automation strategies for improving repeatability of sample data acquisition conditions. Preliminary 3D EBSD experimental results that demonstrate some of these techniques discussed are presented.