Damage Bounding Structural Health Monitoring
The accuracy of locating damage using vibration measurements depends on the number and locations of sensors on the structure. In practical applications, only a limited number of on-board sensors will be available. The limited number of sensors causes spatial aliasing when solving an underdetermined...
Gespeichert in:
Veröffentlicht in: | Journal of intelligent material systems and structures 2006-07, Vol.17 (7), p.629-648 |
---|---|
Hauptverfasser: | , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The accuracy of locating damage using vibration measurements depends on the number and locations of sensors on the structure. In practical applications, only a limited number of on-board sensors will be available. The limited number of sensors causes spatial aliasing when solving an underdetermined inverse problem to locate damage. In this study, instead of trying to exactly locate damage, a method is developed that bounds the damage in-between the closest sensors on bar and beam structures. The method is based on current and predamage measurement of all the degrees of freedom (dofs) of vibration at a small number of sensor locations on the structure, and the excitation must be applied at the points where the response is measured. No model of the structure is needed. The technique bounds the damage location between the closest sensors, and it can compensate for a uniform change in temperature that would otherwise be interpreted as damage. The damage force is also approximated. A limitation of the technique is that rotational vibration must be measured if bending occurs. A finite-element method (FEM) model of an axial bar was used to demonstrate the technique using full and partial measurements. A simulation verified that the damage is always bounded between the closest sensors. Experiments were then performed in which reversible damage was put in an axial bar. The damage was detectable, but the damage location was found to be sensitive to errors in the experimental measurements. Suggestions for improving the experimental technique and for extending the method to plate and frame structures are given. The damage bounding method may also be useful as an FEM model verification tool. In this case, the technique might tell in what region of a structure there is a modeling problem. Existing model updating methods, or model-based damage detection methods, could then be used with more confidence to correct the model. |
---|---|
ISSN: | 1045-389X 1530-8138 |
DOI: | 10.1177/1045389X06060148 |