Session: 12-16-03: Drucker Medal Symposium

Paper Number: 95322

95322 - Data-Driven Approaches for High-Throughput Materials Characterization 

Structural materials for aerospace applications, such as polycrystalline metallic alloys and composites, have hierarchical and heterogeneous structures that drive their deformation and failure mechanisms. The relationship between material structure and behavior -- such as the impact of the microstructure of a polycrystalline metal on dislocation slip, grain boundary sliding, and multi-crack systems, or the impact of the constituent layup of a composite on the accumulation of damage such as matrix cracking and fiber breaking, leads to eventual and sometimes catastrophic failure of the material component. The impact of structure on material behavior includes complex stochastic and deterministic factors whose interactions are under active debate.

In this talk, the application of data-driven and machine learning approaches to the mechanics of structural aerospace materials under a range of applications will be discussed, largely in the context of enabling high throughput experimentation and analysis, and in enabling new modes of structural health monitoring. Examples that will be discussed include large mm-scale mapping and evaluation of long range slip in Ti6242 under dwell fatigue with respect to the microstructure, and its impact on predictive lifing of these alloys with a focus on mixtrotextured region designation. This will include the discussion of a new experimental approach to capture high spatial resolution, large field of view microscale deformation maps, which is then used to capture slip across thousands of grains, and the subsequent analysis of long range slip and microstructural interactions.  Another example that will be discussed is machine learning applied to ceramic matrix composites to enable high throughput identification, or ‘fingerprinting’, of damage mechanisms in the bulk of elastically similar composites for the first time. This fingerprinting is shown to be sensor and sensor location independent and has direct application to the structural health monitoring of advanced composites. Many of the newly developed experimental and analytical approaches that will be presented in this talk are length scale independent and material agnostic, and can be modified to identify a range of deformation and failure mechanisms. Using these and other examples, such as superresolution of electron backscatter diffraction data using generative adversarial networks, this talk is meant to approach the intersection of machine learning and mechanics as a whole. There will be discussion of the overall opportunities and challenges in the application of data driven approaches to structural aerospace materials in these different contexts, with an integrated discussion of the pros and cons of various approaches across this representative set of applications.

Presenting Author: Samantha Daly UCSB

Presenting Author Biography: Sam Daly is a Professor in the Department of Mechanical Engineering. 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. Her research interests lie at the intersection of experimental mechanics, materials science, and data science. Currently, the group is engaged in the development of new methods for multi-scale material characterization and application of machine learning to understand the deformation and failure of advanced structural materials. Prof. Daly is a Fellow of The American Society of Mechanical Engineers (ASME) and serves as an Associate Editor of Applied Mechanics Reviews, Experimental Mechanics, and Strain.

Authors:

Samantha Daly UCSB