CaltechAUTHORS
  A Caltech Library Service

Reference-free detection of minute, non-visible, damage using full-field, high-resolution mode shapes output-only identified from digital videos of structures

Yang, Yongchao and Dorn, Charles and Mancini, Tyler and Talken, Zachary and Theiler, James and Kenyon, Garrett and Farrar, Charles and Mascareñas, David (2018) Reference-free detection of minute, non-visible, damage using full-field, high-resolution mode shapes output-only identified from digital videos of structures. Structural Health Monitoring, 17 (3). pp. 514-531. ISSN 1475-9217. http://resolver.caltech.edu/CaltechAUTHORS:20180530-100353522

Full text is not posted in this repository. Consult Related URLs below.

Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechAUTHORS:20180530-100353522

Abstract

Detecting damage in structures based on the change in their dynamics or modal parameters (modal frequencies and mode shapes) has been extensively studied for three decades. The success of such a global, passive, vibration-based method in field applications, however, has long been hindered by the bottleneck of low spatial resolution vibration sensor measurements. The primary reason is that damage typically initiates and develops in local regions that need to be captured and characterized by very high spatial resolution vibration measurements and modal parameters (mode shapes), which are extremely difficult to obtain using traditional vibration measurement techniques. For example, accelerometers and strain-gauge sensors are typically placed at a limited number of discrete locations, providing low spatial resolution vibration measurements. Laser vibrometers provide high-resolution measurements, but are expensive and make sequential measurements that are time- and labor-consuming. Recently, digital video cameras—which are relatively low cost, agile, and able to provide high spatial resolution, simultaneous, pixel measurements—have emerged as a promising tool to achieve full-field, high spatial resolution vibration measurements. Combined with advanced vision processing and unsupervised machine algorithms, a new method has recently been developed to blindly and efficiently extract the full-field, high-resolution, dynamic parameters from the video measurements of an operating, output-only structure. This work studies the feasibility of performing damage detection using the full-field, very high spatial resolution mode shape (of the fundamental mode) blindly extracted from the video of the operating (output-only) structure without any knowledge of reference (healthy) structural information. A spatial fractal dimension analysis is applied on the full-field mode shape of the damaged structure to detect damage-induced irregularity. Additionally, the equivalence between the fractal dimension and the squared curvature (modal strain energy) of the mode shape curve, when of high spatial resolution, is mathematically derived. Laboratory experiments are conducted on bench-scale structures, including a building structure and a cantilever beam, to validate the approach. The results illustrate that using the full-field, very high-resolution mode shape enables detection of minute, non-visible, damage in a global, completely passive sensing manner, which was previously not possible to achieve.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.1177/1475921717704385DOIArticle
Additional Information:© 2018 by SAGE Publications. Article first published online: May 3, 2017.
Subject Keywords:Structural dynamics, damage detection, modal analysis, full-field vibration, video processing, fractal dimension
Record Number:CaltechAUTHORS:20180530-100353522
Persistent URL:http://resolver.caltech.edu/CaltechAUTHORS:20180530-100353522
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:86706
Collection:CaltechAUTHORS
Deposited By: Tony Diaz
Deposited On:30 May 2018 17:45
Last Modified:30 May 2018 17:45

Repository Staff Only: item control page