-
Microstructure Generation via Generative Adversarial Network for Heterogeneous, Topologically Complex 3D Materials
Authors:
Tim Hsu,
William K. Epting,
Hokon Kim,
Harry W. Abernathy,
Gregory A. Hackett,
Anthony D. Rollett,
Paul A. Salvador,
Elizabeth A. Holm
Abstract:
Using a large-scale, experimentally captured 3D microstructure dataset, we implement the generative adversarial network (GAN) framework to learn and generate 3D microstructures of solid oxide fuel cell electrodes. The generated microstructures are visually, statistically, and topologically realistic, with distributions of microstructural parameters, including volume fraction, particle size, surfac…
▽ More
Using a large-scale, experimentally captured 3D microstructure dataset, we implement the generative adversarial network (GAN) framework to learn and generate 3D microstructures of solid oxide fuel cell electrodes. The generated microstructures are visually, statistically, and topologically realistic, with distributions of microstructural parameters, including volume fraction, particle size, surface area, tortuosity, and triple phase boundary density, being highly similar to those of the original microstructure. These results are compared and contrasted with those from an established, grain-based generation algorithm (DREAM.3D). Importantly, simulations of electrochemical performance, using a locally resolved finite element model, demonstrate that the GAN generated microstructures closely match the performance distribution of the original, while DREAM.3D leads to significant differences. The ability of the generative machine learning model to recreate microstructures with high fidelity suggests that the essence of complex microstructures may be captured and represented in a compact and manipulatable form.
△ Less
Submitted 22 June, 2020;
originally announced June 2020.
-
Overview: Computer vision and machine learning for microstructural characterization and analysis
Authors:
Elizabeth A. Holm,
Ryan Cohn,
Nan Gao,
Andrew R. Kitahara,
Thomas P. Matson,
Bo Lei,
Srujana Rao Yarasi
Abstract:
The characterization and analysis of microstructure is the foundation of microstructural science, connecting the materials structure to its composition, process history, and properties. Microstructural quantification traditionally involves a human deciding a priori what to measure and then devising a purpose-built method for doing so. However, recent advances in data science, including computer vi…
▽ More
The characterization and analysis of microstructure is the foundation of microstructural science, connecting the materials structure to its composition, process history, and properties. Microstructural quantification traditionally involves a human deciding a priori what to measure and then devising a purpose-built method for doing so. However, recent advances in data science, including computer vision (CV) and machine learning (ML) offer new approaches to extracting information from microstructural images. This overview surveys CV approaches to numerically encode the visual information contained in a microstructural image, which then provides input to supervised or unsupervised ML algorithms that find associations and trends in the high-dimensional image representation. CV/ML systems for microstructural characterization and analysis span the taxonomy of image analysis tasks, including image classification, semantic segmentation, object detection, and instance segmentation. These tools enable new approaches to microstructural analysis, including the development of new, rich visual metrics and the discovery of processing-microstructure-property relationships.
△ Less
Submitted 28 May, 2020;
originally announced May 2020.
-
High throughput quantitative metallography for complex microstructures using deep learning: A case study in ultrahigh carbon steel
Authors:
Brian L. DeCost,
Bo Lei,
Toby Francis,
Elizabeth A. Holm
Abstract:
We apply a deep convolutional neural network segmentation model to enable novel automated microstructure segmentation applications for complex microstructures typically evaluated manually and subjectively. We explore two microstructure segmentation tasks in an openly-available ultrahigh carbon steel microstructure dataset: segmenting cementite particles in the spheroidized matrix, and segmenting l…
▽ More
We apply a deep convolutional neural network segmentation model to enable novel automated microstructure segmentation applications for complex microstructures typically evaluated manually and subjectively. We explore two microstructure segmentation tasks in an openly-available ultrahigh carbon steel microstructure dataset: segmenting cementite particles in the spheroidized matrix, and segmenting larger fields of view featuring grain boundary carbide, spheroidized particle matrix, particle-free grain boundary denuded zone, and Widmanstätten cementite. We also demonstrate how to combine these data-driven microstructure segmentation models to obtain empirical cementite particle size and denuded zone width distributions from more complex micrographs containing multiple microconstituents. The full annotated dataset is available on materialsdata.nist.gov (https://materialsdata.nist.gov/handle/11256/964).
△ Less
Submitted 4 February, 2019; v1 submitted 4 May, 2018;
originally announced May 2018.
-
A comparative study of feature selection methods for stress hotspot classification in materials
Authors:
Ankita Mangal,
Elizabeth A. Holm
Abstract:
The first step in constructing a machine learning model is defining the features of the data set that can be used for optimal learning. In this work we discuss feature selection methods, which can be used to build better models, as well as achieve model interpretability. We applied these methods in the context of stress hotspot classification problem, to determine what microstructural characterist…
▽ More
The first step in constructing a machine learning model is defining the features of the data set that can be used for optimal learning. In this work we discuss feature selection methods, which can be used to build better models, as well as achieve model interpretability. We applied these methods in the context of stress hotspot classification problem, to determine what microstructural characteristics can cause stress to build up in certain grains during uniaxial tensile deformation. The results show how some feature selection techniques are biased and demonstrate a preferred technique to get feature rankings for physical interpretations.
△ Less
Submitted 19 April, 2018;
originally announced April 2018.
-
Applied Machine Learning to Predict Stress Hotspots II: Hexagonal close packed materials
Authors:
Ankita Mangal,
Elizabeth A. Holm
Abstract:
Stress hotspots are regions of stress concentrations that form under deformation in polycrystalline materials. We use a machine learning approach to study the effect of preferred slip systems and microstructural features that reflect local crystallography, geometry, and connectivity on stress hotspot formation in hexagonal close packed materials under uniaxial tensile stress. We consider two cases…
▽ More
Stress hotspots are regions of stress concentrations that form under deformation in polycrystalline materials. We use a machine learning approach to study the effect of preferred slip systems and microstructural features that reflect local crystallography, geometry, and connectivity on stress hotspot formation in hexagonal close packed materials under uniaxial tensile stress. We consider two cases: one without any preferred slip systems with a critically resolved shear stress (CRSS) ratio of 1:1:1, and a second with CRSS ratio 0.1:1:3 for basal: prismatic: pyramidal slip systems. Random forest based machine learning models predict hotspot formation with an AUC (area under curve) score of 0.82 for the Equal CRSS and 0.81 for the Unequal CRSS cases. The results show how data driven techniques can be utilized to predict hotspots as well as pinpoint the microstructural features causing stress hotspot formation in polycrystalline microstructures
△ Less
Submitted 16 April, 2018;
originally announced April 2018.
-
Building Data-driven Models with Microstructural Images: Generalization and Interpretability
Authors:
Julia Ling,
Maxwell Hutchinson,
Erin Antono,
Brian DeCost,
Elizabeth A. Holm,
Bryce Meredig
Abstract:
As data-driven methods rise in popularity in materials science applications, a key question is how these machine learning models can be used to understand microstructure. Given the importance of process-structure-property relations throughout materials science, it seems logical that models that can leverage microstructural data would be more capable of predicting property information. While there…
▽ More
As data-driven methods rise in popularity in materials science applications, a key question is how these machine learning models can be used to understand microstructure. Given the importance of process-structure-property relations throughout materials science, it seems logical that models that can leverage microstructural data would be more capable of predicting property information. While there have been some recent attempts to use convolutional neural networks to understand microstructural images, these early studies have focused only on which featurizations yield the highest machine learning model accuracy for a single data set. This paper explores the use of convolutional neural networks for classifying microstructure with a more holistic set of objectives in mind: generalization between data sets, number of features required, and interpretability.
△ Less
Submitted 1 November, 2017;
originally announced November 2017.
-
Applied Machine Learning to Predict Stress Hotspots I: Face Centered Cubic Materials
Authors:
Ankita Mangal,
Elizabeth A. Holm
Abstract:
We investigate the formation of stress hotspots in polycrystalline materials under uniaxial tensile deformation by integrating full field crystal plasticity based deformation models and machine learning techniques to gain data driven insights about microstructural properties. Synthetic 3D microstructures are created representing single phase equiaxed microstructures for generic copper alloys. Unia…
▽ More
We investigate the formation of stress hotspots in polycrystalline materials under uniaxial tensile deformation by integrating full field crystal plasticity based deformation models and machine learning techniques to gain data driven insights about microstructural properties. Synthetic 3D microstructures are created representing single phase equiaxed microstructures for generic copper alloys. Uniaxial tensile deformation is simulated using a 3-D full-field, image-based Fast Fourier Transform (FFT) technique with rate-sensitive crystal plasticity, to get local micro- mechanical fields (stress and strain rates). Stress hotspots are defined as the grains having stress values above the 90th percentile of the stress distribution. Hotspot neighborhoods are then characterized using metrics that reflect local crystallography, geometry, and connectivity. This data is used to create input feature vectors to train a random forest learning algorithm, which predicts the grains that will become stress hotspots. We are able to achieve an area under the receiving operating characteristic curve (ROC-AUC) of 0.74 for face centered cubic materials modeled on generic copper alloys. The results show the power and the limitations of the machine learning approach applied to the polycrystalline grain networks.
△ Less
Submitted 13 June, 2018; v1 submitted 31 October, 2017;
originally announced November 2017.
-
The structure and motion of incoherent Σ3 grain boundaries in FCC metals
Authors:
Jonathan Humberson,
Elizabeth A. Holm
Abstract:
Synthetic driving force molecular dynamics simulations were utilized to survey grain boundary mobility in three classes of incoherent Σ3 twin boundaries: <112>, <110>, and <111> tilt boundaries. These boundaries are faceted on low energy planes, and step flow boundary motion occurs by glide of the triplets of partial dislocations that comprise the mobile facets. Systematic trends with inclination…
▽ More
Synthetic driving force molecular dynamics simulations were utilized to survey grain boundary mobility in three classes of incoherent Σ3 twin boundaries: <112>, <110>, and <111> tilt boundaries. These boundaries are faceted on low energy planes, and step flow boundary motion occurs by glide of the triplets of partial dislocations that comprise the mobile facets. Systematic trends with inclination angle are identified and characterized. Observations of thermally activated, anti-thermal, and athermal motion are explained in terms of the orientation of the Shockley partial dislocations along close-packed and non-close-packed directions. Thermally activated boundaries follow a compensation effect associated with a facet roughening transition. As for all faceting boundaries, system size and driving force must be chosen with care to prevent simulation artifacts.
△ Less
Submitted 10 April, 2017;
originally announced April 2017.
-
Exploring the microstructure manifold: image texture representations applied to ultrahigh carbon steel microstructures
Authors:
Brian L. DeCost,
Toby Francis,
Elizabeth A. Holm
Abstract:
We introduce a microstructure informatics dataset focusing on complex, hierarchical structures found in a single Ultrahigh carbon steel under a range of heat treatments. Applying image representations from contemporary computer vision research to these microstructures, we discuss how both supervised and unsupervised machine learning techniques can be used to yield insight into microstructural tren…
▽ More
We introduce a microstructure informatics dataset focusing on complex, hierarchical structures found in a single Ultrahigh carbon steel under a range of heat treatments. Applying image representations from contemporary computer vision research to these microstructures, we discuss how both supervised and unsupervised machine learning techniques can be used to yield insight into microstructural trends and their relationship to processing conditions. We evaluate and compare keypoint-based and convolutional neural network representations by classifying microstructures according to their primary microconstituent, and by classifying a subset of the microstructures according to the annealing conditions that generated them. Using t-SNE, a nonlinear dimensionality reduction and visualization technique, we demonstrate graphical methods of exploring microstructure and processing datasets, and for understanding and interpreting high-dimensional microstructure representations.
△ Less
Submitted 9 February, 2017; v1 submitted 3 February, 2017;
originally announced February 2017.
