Data-Enabled Quantification of Aluminum Microstructural Damage Under Tensile Loading
Image credit: Unsplash
Abstract
The study of material failure with digital analytics is in its infancy and offers a new perspective to advance our understanding of damage initiation and evolution in metals. In this article, we study the failure of aluminum using data-enabled methods, statistics and data mining. Through the use of tension tests, we establish a multivariate acoustic-data matrix of random damage events, which typically are not visible and are very difficult to measure due to their variability, diversity and interactivity during damage processes. Aluminium alloy 6061-T651 and single crystal aluminium with a (111) orientation were evaluated by comparing the collection of acoustic signals from damage events caused primarily by slip in the single crystal and multimode fracture of the alloy. We found the resulting acoustic damage-event data to be large semi-structured volumes of Big Data with the potential to be mined for information that describes the materials damage state under strain. Our data-enabled analyses has allowed us to determine statistical distributions of multiscale random damage that provide a means to quantify the material damage state.
Supplementary notes can be added here, including code and math.
Liang Zhang
PhD Student of Computer Engineering
My research interests include educational data mining and analytics, learning analytics and programmable matter.