Thursday, 28 May 2020 Speaker: Prof. Samantha Daly Department of Mechanical Engineering, University of California at Santa Barbara, USA Machine Learning and Computer Vision For Structure-Property Analyses

Abstract

The influence of microstructure on deformation and failure mechanisms, such as twinning, dislocation slip, grain boundary sliding, and multi-crack systems, includes complex stochastic and deterministic factors whose interactions are under active debate. In this talk, the application of machine learning and computer vision to microscale displacement data for the high-throughput segmentation and identification of deformation mechanisms – in this example, twinning in magnesium – and their evolution under load across mm-scale fields of view will be discussed. A recently created experimental approach to obtain high spatial resolution, large field of view microscale deformation maps will be presented. The use of a new machine learning framework for the multi-modal alignment of deformation data with crystallographic information, and the rapid segmentation and identification of twin variants across thousands of grains, will be discussed. The analysis of deformation twinning in the Mg alloy WE43 over thousands of grains in each individual test, including an examination of the impact of microstructure on the relative activity of specific twin variants (automatically identified from microscale strain fields) and their evolution in time, will be presented.

Biography

Samantha Daly is an Associate Professor in the Department of Mechanical Engineering at the University of California at Santa Barbara. She received her Ph.D. from Caltech in 2007 and subsequently joined the University of Michigan, where she was on the faculty until 2016 prior to her move to UCSB. The Daly group investigates the mechanics of materials, with a focus on fatigue, fracture, creep, composites, multi-functional materials, and new experimental and data-driven approaches for the characterization of processing – structure – property relationships. Her recognitions include the Experimental Mechanics Best Paper of the Year Award, IJSS Best Paper of the Year Award, DOE Early Career Award, NSF CAREER Award, AFOSR-YIP Award, ASME Eshelby Mechanics Award, Journal of Strain Analysis Young Investigator Award, ASME Orr Award, and Caddell Award. She currently serves on the Executive Board of the Society of Experimental Mechanics, and as an Associate Editor of the journals Applied Mechanics Reviews, Experimental Mechanics, Strain, and Mechanics of Materials.

Notes

• Large Field-of-View in situ SEM-DIC maps (1500 grains, 100M data points) from Mg alloy WE43 tensile experiments were reconstructed and aligned with EBSD maps.


• Each grain data points (ε₁₁, ε₂₂, ε₁₂) are segmented following a k-means clustering method and each cluster is later identified either as twin or not based on the dissimilarity between experimental strains and theoretical twin strains.


• High Schmid Factor twinned grains had larger twin area fractions, however not all high Schmid Factor grains presented twins and some low Schmid Factor grains were twinned.

Other readings

• C. Kaira, X. Yang, V. De Andrade, F. De Carlo, W. Scullin, D. Gursoy, N. Chawla
  Automated correlative segmentation of large Transmission X-ray Microscopy (TXM) tomograms using deep learning
  Materials Characterization 142, 203-210 (2018)


• Y. Balit, E. Charkaluk, A. Constantinescu
  Digital image correlation for microstructural analysis of deformation pattern in additively manufactured 316L thin walls
  Additive Manufacturing 31, 100862 (2019)


• K. Kaufmann, C. Zhu, A. Rosengarten, D. Maryanovsky, T. Harrington, E. Marin, K. Vecchio
  Crystal symmetry determination in electron diffraction using machine learning
  Science 367, 564-568 (2020)

LMS seminars on related topics

• 20/05/2020, Dr. Eric Maire, In situ experiments in X ray tomography: Exploring the interior of ordinary materials to improve their mechanical analysis