Characterization: Structural Descriptors, Data-Intensive Techniques, and Uncertainty Quantification: Structural Descriptors Enabling PSP Linkages
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS Extraction and Processing Division, TMS: Computational Materials Science and Engineering Committee, TMS: Materials Characterization Committee
Program Organizers: Shawn Coleman, DEVCOM Army Research Laboratory; Tomoko Sano, U.S. Army Research Laboratory; James Hogan, University of Alberta; Srikanth Patala, SASA Institute; Oliver Johnson, Brigham Young University; Francesca Tavazza, National Institute of Standards and Technology
Monday 8:00 AM
February 24, 2020
Room: Theater A-3
Location: San Diego Convention Ctr
Session Chair: Srikanth Patala, North Carolina State University; Shawn Coleman, CCDC Army Research Laboratory
8:00 AM Introductory Comments
8:05 AM Invited
Feature Engineering of Material Structure for Extracting Process-structure-property Linkages: Surya Kalidindi1; 1Georgia Institute of Technology
This paper will review our recently developed framework for a rigorous statistical quantification of the hierarchical material internal structure spanning multiple material structure/length scales, and its efficient representation using various orthogonal representations including Fourier basis and principal component analyses (PCA). This new framework offers a consistent and systematic quantification of the material structure in low dimensional forms that are ideally suited for the extraction of low computational cost process-structure-property (PSP) linkages needed to efficiently explore the vast materials design space. The versatility and broad applicability of this framework will be demonstrated through case studies selected in different materials classes and different material length scales.
8:35 AM Invited
Integrated Structural Methods Addressing Aviation Challenges in Composites: Andrew Makeev1; Sarvenaz Ghaffari1; 1University of Texas Arlington
This presentation will showcase experimental and theoretical approaches being developed at the University of Texas Arlington Advanced Materials and Structures Lab (AMSL) for microstructural characterization needed for design, qualification, production, and sustainment of advanced composite materials and structures. These innovations are based on the development, and most importantly effective integration, of the following elements. First, high-resolution nondestructive inspection and structural diagnostics are enabling objective 3D measurement of material structure including location and size of manufacturing irregularities. with automated transition to structural and microstructural models. Second, effective methods are capturing essential in-situ material constitutive properties to complete the input data requirements for comprehensive strength/fatigue models. These methods capture physics of the investigated phenomena including stress-strain response, strength/fatigue and toughness material characteristics. Third, the physical observations and material properties are combined in a computational analysis to capture multiple damage modes and their interaction and predict material performance and structural integrity.
9:05 AM Invited
Investigations of Microstructural Effects on Porosity Evolution: Nathan Barton1; 1Lawrence Livermore National Laboratory
We present results from an investigation of microstructural effects on failure of ductile metals. The computational model makes use of a crystal mechanics based constitutive model that includes porosity evolution. The formulation includes nucleation behavior, enhancing capabilities for modeling small length scales at which nucleation site potency and volume fraction are more variable. Anisotropic crystal response and interactions among the crystals produces heterogeneities that influence spall response and spatial resolution of the polycrystal allows for investigation of various types of nucleation site distributions. By focusing validation efforts on models that connect to experimentally measurable microstructural features, we can then build confidence in use of the models for components prepared under different processing routes, with different chemical compositions and attendant impurity distributions, or subjected to different loading conditions. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 (LLNL-ABS-664056).
9:35 AM Break
9:55 AM Invited
Large Scale Microstructure Synthesis Using LEGOMAT: Application to Additive Manufacturing: Veera Sundararaghavan1; Iman Javaheri1; 1University of Michigan
We present a software LEGOMAT applicable to large scale synthesis of microstructural models (with millions of grains) embedded into engineering CAD models. We demonstrate the use of LEGOMAT for embedding microstructures for an additively manufactured (laser engineered net shaping) part. The approach embeds microstructures along variety of user specified laser paths accounting for hatch widths, layer thicknesses and scan directions. The input microstructures are obtained from location specific electron back scatter diffraction (EBSD) maps which are converted to 3D microstructures using Markov random fields. Crystal plasticity simulations of large scale microstructures are carried out using our open source PRISMS Plasticity code. The results are useful for data science studies to understand process-structure-property relationships and to quantify the errors and uncertainties in homogenization methods used in materials design.
10:25 AM Invited
Advancement of Data Intensive Approaches in Materials Discovery and Design: David Elbert1; Brian Schuster2; Nick Carey1; Connor Krill1; Ali Rachidi1; William Phelan1; Tyrel McQueen1; 1Johns Hopkins University; 2ARL
Recent advances in data collection and application provide the promise of more efficient and innovative materials discovery to accelerate creation and use of novel materials to serve society. To realize the opportunities of such data requires a materials data infrastructure (MDI) including scalable visualization and analytics as well as open and integrated ways to understand and share information across the materials domain.Our collaborative MDI leverages SciServer Big Data architecture, streaming data, and unified semantics to support data sharing and analysis. We're particularly focused on accelerating work with experimental data through automated analysis and real time applications. As an example, rapid image segmentation from deep learning models is transformative in a range of applications from synthesis to characterization. Ultimately the revolution in data-driven materials science depends on connecting such visions of MDI throughout the community. Functional methods and commitment to open data will be the fuel for AI/ML innovation.