Akademik Veri Yönetim Sistemi
PACE-2021 International Congress on the Phenomenological Aspects of Civil Engineering, Erzurum, Türkiye, 20 - 23 Haziran 2021, cilt.1, sa.1, ss.39-54
This
research reviews the development and application of time domain and frequency
domain transformations in the field of structural identification and
particularly in the structural health monitoring (SHM) practices in the last
half century. The challenges in conventional transformations and future trends
are discussed in the light of the development of Time-Frequency domain
techniques for the SHM of civil engineering structures. Fourier Transform,
short time Fourier transform, Laplace transform, Wavelet transforms, and
Hilbert transforms are exemplified as some currently used transformations in
data mining studies in engineering practice. Feature extraction is mostly based
on the conventional parametric modelling in identification studies in time and
frequency domains. The SHM operation covers the operational evaluation during
monitoring the structure, data acquisition, fusion, cleansing, feature
extraction, and decomposition. Time-Frequency analysis shows that the values of
the structural parameters are considerably changing under operational
conditions as observed from field measurements. Fluctuating values or stability
information extracted from SHM data continuously reflect the seasonal, temporal
or permanent state under different triggering sources under environmental
conditions and internally developing stresses and strains. In fact, the
recorded responses are indirect observations of unobservable effects underlying
several factors. One of them might be nonlinear
dynamic responses. Using Time-Frequency analysis techniques, system
identification on a real structure may give the possibility to trace the buried
signals to identify the structural behavior, damage or progressive
deterioration including the source and collapse mechanism. If the whole
structure is instrumented dense enough including every structural member, it
will be possible to monitor the health of the structure even at the sensitivity
of the capacity exceedance or the condition before the crack or flaw
initiations. At the member sensitivity level, structural performance will be
automatically assessed tracking the stress-strain variations of the structural
frame in near future. From announcing to repairing the vulnerability will be
the solution plan after information collection through the SHM network. For
such an automated structural control, direct practical results that can be
obtained from Time-Frequency techniques will be available and then can be used
effectively by a decision maker unit.
Time Frequency analysis may affect
and improve the building codes for a risk consistent structural design. The
future revisions of aseismic design and condition assessment methodologies will
use the time-frequency domain analysis to better understand the chaotic natural
forces and their responses. Time-Frequency analysis that promises new horizons
will affect and improve the current procedures, check points and limits to
artificially generate the seismic ground motions or evaluate the recorded strong
ground motions for the time domain analysis.