Before the advent of computers, scientists used time domain periodogram analysis (TDPA) for analysis of phenomena described by time-dependent parameters. Famous examples of this approach include investigations of periodicities in Sun activity, analysis of annual rings of trees and deposits of clays in large lakes, both of which are indicators of climate variations. These periodicities found by TDPA were later confirmed using modern Fourier transform-based methods. Since the advent of the computer era and the fast Fourier transform, TDPA has been almost completely abandoned. In this work, we develop a computer-based TDPA algorithm and demonstrate its' utility for determining the existence (Level 1) of structural damage. First, the TDPA is compared to a frequency domain periodogram analysis for synthetic signals with well-defined statistical properties. This comparison elucidates the features of the TDPA algorithm. Next, a damage existence `feature' is identified from the TDPA and is shown to be a reliable indicator of the existence of structural damage. This is demonstrated using simulated and actual experimental data.
[1]
Truong Q. Nguyen,et al.
Wavelets and filter banks
,
1996
.
[2]
Charles R. Farrar,et al.
Applying the LANL Statistical Pattern Recognition Paradigm for Structural Health Monitoring to Data from a Surface-Effect Fast Patrol Boat
,
2001
.
[3]
Ronald W. Schafer,et al.
Digital Processing of Speech Signals
,
1978
.
[4]
P. Welch.
The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms
,
1967
.
[5]
Randy Ribler,et al.
Two performance tool design issues and CHITRA's solutions
,
1996,
SPDT '96.
[6]
Petre Stoica,et al.
Introduction to spectral analysis
,
1997
.
[7]
Alan V. Oppenheim,et al.
Discrete-Time Signal Pro-cessing
,
1989
.