Assessing the Viability of Visual Vibrometry for Use in Structural Engineering

Abstract

Visual vibrometry is a computer vision methodology that allows a user to perform modal analysis on an object using typical video cameras, for example a standard smartphone. There is significant potential in using visual vibrometry for structural monitoring and building modal analysis; it significantly reduces cost and set-up time as compared to traditional contact and non-contact based modal analysis techniques, and allows for the measurement of non-conventional structures which might be difficult to otherwise instrument. This paper describes an initial benchmarking of the effectiveness of the visual vibrometry technique in structural modal analysis, with testing conducted on a simply supported beam and a complex semi-rigid frame structure using both visual vibrometry and classical contact-based methods. Results indicated that visual vibrometry was effectively able to detect natural frequencies under 20 Hz for both structures, with reasonable correlation seen between visual and contact-based measurements, as well as analytical and numerical solutions. The poor detection of higher frequency modes was attributed to the receptance frequency response, generated when the collected visual displacement data is transformed to the frequency domain. Doing so highlights low frequency responses but suppresses intermediate and high frequency responses. Use of high-speed cameras may improve high-frequency detection, but further work is needed to determine practical limitations.KeywordsVisual vibrometryStructural dynamicsStructural health monitoringModal analysis