Vibration-based Damage Detection with Compressive Sensing and System Identification

This research is driven by the high interest of industry to develop novel nondestructive testing (NDT) methods to detect defects in complex composite parts in a limited inspection time. Classical NDT methods often rely on thermal imaging techniques and on ultrasound. Within this PhD project, vibro-acoustic signals will be exploited to detect and identify damage in composites. The goal is the development of a technique, which is very complementary to the high frequency methods and allows global and cheap inspection. The main hypothesis made is that the redundant information of vibration and acoustics combined with an efficient and accurate numerical model allows to identify defects. The PhD will focus on the development of two different vibro-acoustic methods for damage detection in composites: (i) exploiting guided wave modes in extended structures and exploiting the dispersion relation of waves in a solid material, (ii) relying on Time Reversal Methods, exploiting the principle that, as long as the medium is non-dissipative, the response of an excitation measured at a certain point in the medium can be time-reversed (the measured signal is emitted backwards) and it will converge at the source location. In this PhD, numerical models of increasing complexity will be developed and supported by experimental campaigns.