This study explores a structural damage detection strategy employing the novel use of chaotic time series excitation. Chaotic time series have several useful properties such as determinism and controllable dimension. Therefore, these series are attractive candidates for probing a structure’s dynamics for the subtle changes that could occur because of damage. This approach is applied to a metal frame structure connected by bolted joints. Under loading, fatigue damage causes the pre-loads in the bolts to lessen, leading to an increasing inability of the joint to accommodate design loads. This project explores the use of propagating chaotic waveforms through the frame structure and to determine a diagnostic parameter that reflects the structural health of the bolted joints. Analysis of vibration response to chaotic input is used to detect the extent and location of pre-load loss. Using the cross-prediction error as a feature for damage detection was successful in identifying damage and, using excitation to response prediction errors, locating damage. However, in other cross comparisons of attractors, the prediction error was not able to locate the damage. In addition, the extent of damage was not correlated with the magnitude of prediction error. NOMENCLATURE N dimension of a system x, x(t) a time-dependent variable corresponding to a degree of freedom in the system x ) the N derivative of x T time delay for attractor reconstruction m embedding dimension of attractor x,y,z three coordinates describing the degrees of the system of the Lorenz attractor m global mean of the resampled data subsets Za/2 standard normal deviate associated with 95% confidence s global standard deviation of the resampled subsets n size of each resampled subset
[1]
Holger Kantz,et al.
Practical implementation of nonlinear time series methods: The TISEAN package.
,
1998,
Chaos.
[2]
E. Lorenz.
Dimension of weather and climate attractors
,
1991,
Nature.
[3]
Thomas Schreiber,et al.
Detecting and Analyzing Nonstationarity in a Time Series Using Nonlinear Cross Predictions
,
1997,
chao-dyn/9909044.
[4]
J M Nichols,et al.
Attractor reconstruction for non-linear systems: a methodological note.
,
2001,
Mathematical biosciences.
[5]
Michael D. Todd,et al.
Using state space predictive modeling with chaotic interrogation in detecting joint preload loss in a frame structure experiment
,
2003
.
[6]
F. Takens.
Detecting strange attractors in turbulence
,
1981
.
[7]
H. Kantz,et al.
Nonlinear time series analysis
,
1997
.
[8]
H. Abarbanel,et al.
Determining embedding dimension for phase-space reconstruction using a geometrical construction.
,
1992,
Physical review. A, Atomic, molecular, and optical physics.
[9]
S. Levin,et al.
FROM INDIVIDUALS TO POPULATION DENSITIES: SEARCHING FOR THE INTERMEDIATE SCALE OF NONTRIVIAL DETERMINISM
,
1999
.
[10]
J M Nichols,et al.
Use of chaotic excitation and attractor property analysis in structural health monitoring.
,
2003,
Physical review. E, Statistical, nonlinear, and soft matter physics.