Page 17 - ASME SMASIS 2017 Program

Symposia
SYMPOSIUM 5
WAVELET TRANSFORM BASED METHODS FOR STRUCTURAL
HEALTH MONITORING
Anne Kiremidjian
Professor
Department of Civil & Environmental Engineering
Stanford University
Stanford, CA
Abstract
Over the past two decades numerous methods for structural health
monitoring have been developed that use statistical pattern recognition
and machine learning techniques. In this paper, methods utilizing the
continuous wavelet transform that have been developed by the research
team of the author will be presented. In particular, the wavelet energy
methods proposed by Nair and Kiremidjian (2009) and by Noh et al.
(2011,2012) will be summarized. These methods use the wavelet energy at
scales close to the predominant frequency of the structure as the damage
sensitive feature (DSF). Noh et al. (2012) showed analytically that the
wavelet energies are particularly sensitive to changes in structural
parameters, such as stiffness, and thus are good indicators of the
occurrence of damage. Most recently, the wavelet energies DSF were
shown to correlate well to increases in ice mass due to icing on cable
structures (Andre et al. 2016). Based on laboratory tests it was
demonstrated that the wavelet energies could be used as predictors of
the amount of ice formation. Example applications for earthquake damage
prediction and for ice formation using wavelet based DSF’s will be
discussed in the presentations as illustrations of the methods.
Biography
Professor Anne S. Kiremidjian is a Professor in the Department of Civil and currently undergoing this process for steel structures. Reading the fatigue
Environmental Engineering at Stanford University. From 1987 to 2002 she
also served as the Co-Director and Director of the John A. Blume
Earthquake Engineering Center at Stanford University. Dr Kiremidjian
received her B.S. degree from Columbia University in Civil Engineering
and her M.S. and Ph. D. degrees from Stanford University in Structural
Engineering. As faculty at Stanford she teaches courses in structural
analysis, earthquake hazard and risk analysis, structural reliability and
decision theory. Her research over the years has focused on all aspects
of earthquake hazard and loss estimation, regional risk assessment, risk
analysis of transportation systems, damage detection algorithms, wireless
sensor development and structural sensing system design.
Professor Kiremidjian is the recipient of several awards and honors. Most
notable are the Stanford University School of Engineering Distinguished
Advisor Award, National Science Foundation Faculty Award for Women,
Society of Women Engineers Distinguished Educator Award, Extraordinary
Achievement Award from Applied Technology Council, and C. Martin Duke
Award for Excellence in Lifeline Earthquake Engineering Research by the
American Society of Civil Engineers (ASCE). In 2014 she was elected
Distinguished Member of the American Socitye of Civil Engineers, the
highest honor given by the society.
Dr. Kiremidjian has more than 350 publications that include journals
papers, book chapters, technical reports, and refereed conference
proceeding papers. She has been an invited, keynote and distinguished
lecturer at major national and international meetings and makes frequent
presentations at seminars and other professional gatherings. She holds
patents related wireless structural monitoring and advanced damage
detection algorithms.
A PIEZOELECTRIC FATIGUE FUSE BASED WIRELESS SHM SYSTEM
Christopher S. Lynch 17
Professor and Chair
Department of Mechanical & Aerospace Engineering
University of California, Los Angeles
Los Angeles, CA
Abstract
A fatigue fuse is a small strip of metal with a notch to create a stress
concentration. It is bonded to a structure with the intent that a crack
nucleate and propagate across the fuse after a certain number of loading
cycles. Fatigue fuses offer an elegantly simple method of counting
loading cycles on metallic structures and are commercially available from
Metal Fatigue Solutions (MFS). Each fuse is designed to fail after a given
number of cycles at a pre-selected loading ratio R and stress amplitude A.
The values of R and A are determined from knowledge of the anticipated
loading spectrum. Fatigue fuse design is based on crack nucleation and
fatigue crack growth properties of the fuse material. This is followed by
fabrication and extensive verification under cyclic loading. Over the past
year, fatigue fuses have undergone qualification testing to bring them to a
technology readiness level of 6 (TRL-6) for aluminum structures and are
fuses requires a visual inspection through a viewing port through a cover
that protects the fuse from environmental assisted crack growth (corrosion
assisted cracking). The need for visual inspection can be eliminated using
a sensor network.
A piezoelectric based pitch-catch system is under development (NSF
funded STTR) to remotely read the status of the fatigue fuse in cases
where the fuses are mounted in locations that are difficult to access for
inspection. The pitch piezoelectric excites the fundamental bending
resonant mode in the fatigue fuse and the catch piezoelectric responds in
sensor mode. When the crack propagates across the fuse the signal is
disrupted by shifts in the amplitude and resonant frequency. Because the
fatigue fuse is mechanical and the status only needs to be interrogated at
intermittent intervals, the electronics for remote sensing can remain in a
low power quiescent mode except when reading the fuse status and
sending the status across a wireless network. This enables a fatigue
monitoring system that can detect the rate of accumulation of cyclic
fatigue in highway structures and that requires little or no maintenance.