Development of Virtual Visual Sensor Applications for Wood Structural Health Monitoring

Abstract

Abstract Inspection techniques for wood structures are typically based on visual observations of degradation, and often limited to localized observable surface damage on the structure, rather than holistic non-destructive evaluations. Hidden deterioration and overall impacts of distributed damage may not be included in the assessment. This research evaluates the implementation of Eulerian-based virtual visual sensors (VVS), applying commercially available digital video cameras, to characterize dynamic structural response of wood structures. Natural vibration frequencies are determined by monitoring the intensity value of a single fixed pixel coordinate over a few seconds of a video of structural vibration and then applying a fast Fourier transform to estimate signal frequencies. Changes in stiffness and mass of materials and structural systems that may relate to deterioration are reflected in the natural frequencies. The end goal is development and application of VVS to wood structures to obtain information relevant to objective structural health monitoring (SHM). In this development phase, the effects of moisture content and simulated damage on natural frequencies are observed on simply supported beams of dimensional lumber. Initial applications are also made to an in-place U.S. Forest Service pedestrian bridge. Results show comparable accuracy in determining vibrational frequencies with VVS and a commercially available transverse vibration measurement system, successful observation of vibrational frequencies in a timber bridge, and beneficial use of naturally occurring color gradients in wood structures in laboratory and field tests. Moisture content and simulated damage have measurable effects on natural frequencies. Eulerian-based VVS show potential as a tool for cost-effective SHM of wood structures, especially for quick, global screening of structures, with subsequent visual inspection and other means of evaluation.