Practical Tools for Integration and Analysis in Materials Engineering: Session II
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS Structural Materials Division, TMS: Titanium Committee, TMS: Computational Materials Science and Engineering Committee, TMS: Integrated Computational Materials Engineering Committee
Program Organizers: Adam Pilchak, Pratt & Whitney; Michael Gram, Pratt & Whitney; William Joost; Raymundo Arroyave, Texas A&M University; Charles Ward, AFRL/RXM
Monday 2:00 PM
March 15, 2021
Room: RM 34
Location: TMS2021 Virtual
2:00 PM Invited
Foundations and Applications of DAMASK: Philip Eisenlohr1; Martin Diehl2; Pratheek Shanthraj3; Franz Roters4; Dierk Raabe4; 1Michigan State University; 2KU Leuven; 3The University of Manchester; 4Max-Planck-Institut für Eisenforschung
The Düsseldorf Advanced Material Simulation Kit (DAMASK) at its core provides a modular framework to simulate and analyze anisotropic elastoplasticity of (crystalline) materials at finite strains.The modular design enables the use of DAMASK within various boundary value solvers as well as allows to seamlessly integrate and extend the set of different descriptions of the material constitutive behavior.Recent enhancements facilitate the use of DAMASK in coupled multi-field simulations such as chemo-thermo-mechanics.This talk outlines the theoretical underpinnings of the framework design starting from the purely mechanical case and following the development towards coupled problems of crystal plasticity.The demonstrations of exemplary use cases include texture development of face- and body-centered cubic materials, thermo-mechanics during heating and cooling, and the influence of a nearby surface on crystal plasticity.
2:40 PM
Prisms-plasticity: An Open Source Crystal Plasticity Finite Element Software: Mohammadreza Yaghoobi1; Sriram Ganesan1; Aaditya Lakshmanan1; Srihari Sundar1; Duncan Greeley1; Shiva Rudraraju2; John E. Allison1; Veera Sundararaghavan1; 1University of Michigan, Ann Arbor; 2University of Michigan; University of Wisconsin-Madison
An open-source parallel 3-D crystal plasticity finite element (CPFE) software package PRISMS-Plasticity is presented here as a part of an overarching PRISMS Center integrated framework. Highly efficient rate-independent and rate-dependent crystal plasticity algorithms are implemented. Additionally, a new twinning-detwinning mechanism is incorporated into the framework based on an integration point sensitive scheme. The integration of the software as a part of the PRISMS Center framework is demonstrated. This integration includes well-defined pipelines for use of PRISMS-Plasticity software with experimental characterization techniques such as electron backscatter diffraction (EBSD), Digital Image Analysis (DIC), and high-energy synchrotron X-ray diffraction (HEDM), where appropriate these pipelines use popular open source software packages DREAM.3D and Neper. In addition, integration of the PRISMS-Plasticity results with the PRISMS Center information repository, the Materials Commons, will be presented. The parallel performance of the software demonstrates that it scales exceptionally well for large problems running on hundreds of processors.
3:00 PM
A Fast Fourier Transform Based Crystal Plasticity Constitutive Model for Predicting Creep and Rupture Lifetime in Metallic Systems: Nathan Beets1; Laurent Capolungo1; Arul Mariyappan1; Ricardo Lebensohn1; 1Los Alamos National Laboratory
Accurately predicting creep response and rupture lifetime of metallic components under high temperatures and multiaxial stresses is critical to the rapidly evolving energy industry. To this end, we present a mechanistic crystal plasticity-based constitutive model, used to derive engineering-scale creep rupture life criteria. A microstructure-sensitive dislocation kinetics law defines local plastic slip, a Coble creep law models vacancy-mediated plasticity, and latent hardening evolves local dislocation density. Void nucleation/growth are tracked via reaction- diffusion framework and coupled viscoelastic and diffusive dissipative processes. This physical-based framework is incorporated into a parallelizable code which uses fast Fourier transforms (FFTs) to predict the local and global stress response of the material. This framework is faster than FEM-based elasto-viscoplastic codes and therefore can be efficiently used in combination with data analytics and surrogate modeling techniques. In conjunction with a fitting procedure, this enables the determination of a rupture-lifetime criteria for 347H steel.
3:20 PM
PRISMS-PF: A High Performance Phase-field Modeling Framework to Simulate Microstructure Evolution: David Montiel1; Stephen DeWitt1; Yanjun Lyu1; Katsuyo Thornton1; John Allison1; 1University of Michigan
PRISMS-PF is an open-source framework for phase-field simulations of microstructure evolution developed with an emphasis on performance, flexibility, and ease-of-use. The framework’s competitive performance is enabled by its use of a matrix-free finite element method for explicit time integration, advanced adaptive meshing, and multi-level parallelization. We demonstrate the flexibility of PRISMS-PF and its adaptability to study a wide variety of phenomena, such as the interaction of precipitates in magnesium-rare earth alloys and the effects of microstructure in corrosion. We also present benchmark tests which show that the performance of PRISMS-PF either meets or exceeds that of other common open-source phase-field frameworks and largely exceeds that of finite difference codes. Finally, we discuss the integration of PRISMS-PF with other computational packages such as DREAM.3D, VisIt, and the Materials Commons information repository and collaboration platform from the PRISMS Center at the University of Michigan.
3:40 PM Invited
Tools for Microstructural Analysis Using Computer Vision and Machine Learning: Elizabeth Holm1; Bo Lei1; Andrew Kitahara1; Nan Gao1; Ryan Cohn1; 1Carnegie Mellon University
Microstructural science relies on quantitative characterization and analysis of microstructure, typically based on images obtained from microscopy, diffraction, or other experimental modalities. Recent progress in data science, including computer vision (CV) and machine learning (ML), offer new approaches to extracting information from microstructural images. However, because they are generally developed to analyze natural images, and because they are not part of the materials curriculum, applying these methods to materials problems is not always straightforward. This talk presents the basic steps for encoding and analyzing microstructural image data: image preprocessing and data augmentation; feature vector construction; and unsupervised and supervised machine learning. A Python code tutorial that applies these operations on an open access data set of steel defects is included. Case studies demonstrate practical aspects of developing CV/ML workflows, including dataset considerations, hyperparameter selection, ML technique, and potential pitfalls.
4:20 PM
AMPIS: Automated Materials Particle Instance Segmentation: Ryan Cohn1; Timothy Prost2; Iver Anderson2; Emma White2; Jordan Tiarks2; Elizabeth Holm1; 1Carnegie Mellon University; 2Ames Laboratory
Instance segmentation has been shown to be a powerful tool for image analysis but has not been adopted by the materials science community. Thus, we present AMPIS, an open-source framework for applying instance segmentation to materials image data. We provide a case study applying this tool to segment individual powder particles and satellites in images of additive manufacturing (AM) feedstock powders. Detecting, quantifying, and minimizing the presence of satellites is critical to opening up low-cost feedstock options to the AM community. Despite labeling a very small number of images to train the model, segmentation is successfully performed on a wide variety of images, including images captured with different magnifications and imaging modes. The results demonstrate the first ever direct measurements of satellite contents in powders. To demonstrate the flexibility of the technique we provide a second case study characterizing the spheroidite content in microscope images of steel.
4:40 PM
A Method to Reconstruct Prior Beta Grain Orientations from Measured Alpha-phase Electron Backscatter Diffraction Data: Adam Pilchak1; 1US Air Force Research Laboratory
Titanium alloys undergo an allotropic phase transformation during cooling from above the beta transus. Each beta grain decomposes into one or more alpha-phase orientations that share a special crystallographic orientation relationship (OR) with the beta-phase that was first reported by Burgers, viz (0001)alpha || {110}beta and <11-20>alpha || <111>beta. This OR can be exploited to reconstruct the high temperature parent beta-phase from alpha-phase orientations measured at room temperature. The author wrote a code to do this over 10 years ago while a PhD student at Ohio State. After performing countless reconstructions for colleagues and collaborators, the time has come to release the code into the wild! The purpose of this presentation is to describe the approach, the code and its pitfalls, and to provide a link to a persistent repository so that future graduate students and researchers need not spend their valuable time and financial resources developing this same capability.