As an enabling technology, Structural Health Monitoring (SHM) assesses the state of mechanical structures in real time in order to prevent accidents or disasters during the structures’ operational lifetime. According to basic principles, several methods in SHM techniques can be defined e.g. mechanical, electric, electro-mechanical and electro-magnetic methods. The focus of this work is on the Electro-Mechanical Impedance method (EMI) and Cross Transfer Function method.
In this work, the EMI method is further developed for three different fields of application: self-diagnosis of bonded piezoelements, their sensing ability and their use for damage detection in fatigue experiments. Self-diagnosis is used to monitor whether the bonded piezoelectric elements either used as actuators or as sensors can continue to measure or function properly in SHM, which is decided by the change of the EM impedance spectrum. In the investigation of the sensing ability of piezoelectric elements, it is shown how the damage positions and the different frequency ranges of the input signals influence the element measurements using EMI spectrum. This is investigated experimentally in combination with the Spectral Element Method (SEM). In the last part of the work, EMI method is used to monitor the start of the crack in a vibrating aluminium plate in a fatigue experiment.
Another focus of this work lies on a better localization of damage using phased array techniques. Such arrays consist of several piezoelectric elements arranged in a certain geometric way. By controlling the phase shift of the input signal between each piezoelectric element and the superposition of all outputs, a wave propagation with a desired direction is obtained, where the direction of propagation is determined by the phase shift. The proposed technique works in the frequency domain and is based on the Cross Transfer Function method which uses the transfer functions between different actuatorsensors permutations of the array. In the frequency domain, the phase shifts are implemented numerically after the experimental determination of the transfer functions using a computer algorithm. This can be seen as a big advantage of the Cross Transfer Function method which makes it more flexible and efficient. The results show that the damage indicators are especially large in those directions where the damage is located.