Model-Based Detection of Structural Damage by Means of Vibration Measurement Data

The paper deals with health monitoring and damage detection of structural systems by means of model based approaches. Starting with the general sensitivity approach it is shown how the localization and quantification of structural damage determination from changes of measured vibration characteristics can be performed. Further, a concept is presented to generate local vibrations in a certain region of the structure while suppressing the motion of the structure by active vibration control at the other locations. When the local deformation coincides with the damage location, this results in a significant deviation of the measured system response and the damage can be diagnosed. The theory is applied to a composite beam with distributed piezoelectric elements. By means of its sensing and actuating capabilities as well as its "computational intelligence", it is shown that this smart system is able to perform a selfdiagnosis.