September 19
@
12:05 PM
–
12:55 PM
The Mechanics Seminar Series is a weekly seminar given by campus and visiting speakers on topics across the spectrum of mechanics research (solids, fluids, and dynamics). Professor Matthew Brake is an Associate Professor at Rice University.
Presentation Title: Videographic Modal Analysis
Abstract: Experimental modal analysis is a classical tool that has formed the basis of modern vibration testing and qualification. Despite this, the process of experimental modal analysis is largely reliant upon techniques from over 60 years ago. Typical experiments of large structures can take months of planning and setup, and require thousands of channels of accelerometer data. To lower the cost and time associated with vibration testing, Videographic Modal Analysis (VMA) combines recent advances in data science, image processing, and traditional modal analysis concepts. First, natural frequencies are identified from videos of an experiment using a deep learning algorithm. Once these frequencies are identified, optical flow, phase-based motion magnification, and edge detection techniques are automatically applied to quantitatively characterize the mode shapes of the structure. Without instrumentation, we can now identify mode shapes and natural frequencies of arbitrary structures in real-world settings (not just beams vibrating in labs!). This talk will present the science behind VMA and demonstrate its applicability.
Bio: Prof. Brake started at Rice University in 2016 after working at Sandia National Laboratories for nine years. Prior to Sandia, Prof. Brake graduated from Carnegie Mellon University in 2007. Prof. Brake has been elected to several leadership positions, including as the director of the International Committee on Joint Mechanics, the chair of the Nonlinear Dynamics Technical Division of SEM, and the chair of the ASME Technical Committee on Vibration and Sound. He is a recipient of the 2012 Presidential Early Career Award for Scientists and Engineers, the 2018 C.D. Mote Jr Early Career Award, and the National Science Foundation Career Award. His primary research interests are in data-driven experimentation, multi-scale and multi-physics modeling, vibration, tribology, uncertainty propagation, structural health monitoring, and nonlinear dynamics.