ME Faculty Candidate Brian Wisner Research Seminar
- Monday, February 26, 2018 at 3:10pm
- Roberts Hall, Room 307 - view map
Damage State Identification by Coupling Microstructure-Sensitive Nondestructive Evaluation with Machine Learning Tools
Abstract: Damage is a complex and stochastic material specific process bridging several time and length scales. This presentation discusses the initiation of damage and identifies damage precursors in a range of materials used in the Automotive and Aerospace industry using material and scale agnostic Nondestructive Evaluation (NDE) tools combined with mechanical testing and characterization methods applied at a scale where damage incubation and initiation is occurring. Specifically, a novel setup built inside a Scanning Electron Microscope (SEM) and retrofitted to be combined with characterization and NDE capabilities is leveraged with the goal to track the early stages of the damage process in Al and Mg alloys. The characterization capabilities include ex situ Electron Backscatter Diffraction (EBSD) and Energy Dispersive Spectroscopy (EDS), X-ray micro-computed tomography (μ-CT) and nanoindentation, in addition to in situ microscopy achieved by the Secondary Electron (SE) and Back Scatter Electron (BSE) detectors. The mechanical testing inside the SEM was achieved with the use of an appropriate stage that fitted within its chamber and is capable of applying both axial and bending monotonic and cyclic loads. The NDE capabilities, beyond the microscopy and μ-CT, include the methods of Acoustic Emission and Digital Image Correlation (DIC). This setup was used to identify damage precursors in this material system and their evolution over time and space. The experimental results were analyzed by a custom signal processing scheme that involves both feature-based analyses as well as a machine learning method to relate recorded microstructural data to damage in this material. Finally, damage initiation and evolution is presented in terms of a carbon fiber reinforced polymer composite demonstrating the applicability of the NDE tools to understand damage evolution across material systems as well as length and time scales. Extensions of the presented approach to include information from computational methods will also be discussed.
Bio: Brian received his Bachelor’s Degrees in Mechanical Engineering and Physics from Widener University in 2011 and his Master’s Degree in Mechanical Engineering and Mechanics from Drexel University in 2013. Brian spent the next year and a half working in industry at Piasecki Aircraft, where he was a structural design engineer and configuration manager for a number of projects. During this time, Brian was an adjunct professor at his alma mater teaching courses related to key concepts in structural design. Brian returned to Drexel in 2014 and completed his PhD in 2017 focusing on experimental mechanics and damage precursor identification. During the course of his degree, Brian was able to work at a number of national labs including Sandia National labs and NASA Langley Research Center. Since the completion of his degree, Brian has been working as a post-doctoral fellow in the Theoretical and Applied Mechanics group working on and leading a number of projects.
- Department of Mechanical & Industrial Engineering