Despite the promising application of Distributed Optical Fiber Sensors (DOFS) in monitoring damage in composite structures, their implementation outside academia is still unsatisfactory due to the lack of a systematic approach to assessing their damage detection performance. The existing tool developed for non-destructive evaluation, Probability of Detection (POD) curves, needs to be adapted for structural health monitoring applications to account for spatial and temporal dependency. Damage detection performance with DOFS is deeply related to the inherent variability sources of the system, the strain transfer properties of the optical fiber, and the loading conditions, which determine the damage-induced strain on the structure. This paper establishes a systematic approach based on the Length at Detection (LaD) method to qualify DOFS for damage detection in composites under different scenarios. Specifically, this study considers two DOFS with different strain transfer properties for monitoring delamination in carbon fiber reinforced polymers double-cantilever beam specimens under mode I quasi-static and fatigue loading. The POD curves derived from the LaD method confirm that this methodology can quantify the change in the detection performance due to the DOFS type and the loading conditions. The study also proposes a practical solution to compare POD curves obtained with different sample sizes, by introducing the concept of virtual specimens to simulate the lower confidence bound convergence.

Structural Health Monitoring (SHM) aims to shift aircraft maintenance from a time-based to a condition-based approach. Within all the SHM techniques, Acoustic Emission (AE) allows for the monitoring of large areas by analyzing Lamb waves propagating in plate like structures. In this study, the authors proposed a Time Reversal (TR) methodology with the aim of reconstructing an original and unaltered signal from an AE event. Although the TR method has been applied in Narrow-Band (NwB) signal reconstruction, it fails when a Broad-Band (BdB) signal, such as a real AE event, is present. Therefore, a novel methodology based on the use of a Frequencies Compensation Transfer Function (FCTF), which is capable of reconstructing both NwB and real BdB signals, is presented. The study was carried out experimentally using several sensor layouts and materials with two different AE sources: (i) a Numerically Built Broadband (NBB) signal, (ii) a Pencil Lead Break (PLB). The results were validated numerically using Abaqus/CAE^TM with the implementation of absorbing boundaries to minimize edge reflections.

The effect of adhesive thickness on fatigue crack growth in an epoxy film adhesive (FM94) was investigated, using a combination of experiments and numerical modelling. For the range of thicknesses investigated an increased thickness led to an increased crack growth rate. It was found that the energy required per unit of crack growth did not depend on the adhesive thickness. In contrast, the energy available for crack growth does depend on the adhesive thickness. The numerical analysis confirms that the energy required per unit crack growth is not sensitive to the adhesive thickness, but that the plastic energy dissipation increases with the thickness. The experimental results imply that this increase of plasticity has an anti-shielding effect, as the crack growth rate is increased.

In the Structural Health Monitoring (SHM) field, Acoustic Emissions (AE) is the process by which acoustic signals generated during the formation of damage are captured by sensors, analyzed and used for localization within the structure. In plate like structures, these signals lead to the formation of Lamb Waves (LW), which are broadband in nature. These LW are generally captured by Piezoelectric Titanum Zirconate (PZT) sensors. As such, the captured broadband signals are of difficult interpretation in part due to several phenomena such as dispersion or attenuation suffered by the waves during their propagation. In this study, we hypothesize that the nature of the emitted signal contains information on the damage type, as if the features of the emitted signal were a 'fingerprint' of the damage. Wing or fuselage panels are some of the aeronautical structures were LW can develop during the emission of an acoustic signal. In operational service environments, the damage type and size may lead to the generation of different signal sources. This study aims at the development, through experimental techniques, of a classification algorithm based on Artificial Intelligence (AI) for determining the source of the emission in addition to their location within a structure. It is envisioned that the AI algorithms will be capable of identifying specific features within the emitted signals and thus correlate them to a database of known signals and their corresponding associated damage types. In order to create an AE signal damage database, the captured signal cannot be used since it has been affected by its propagation through the structure. As such, a Time Reversal process will be implemented in order to reconstruct the original signal. This original signal will be the one utilized by the AI algorithm in order to identify its corresponding damage source.

This study focuses on the development of a source identification algorithm inspired by the SHAZAM music app. The algorithm makes use of a spectrogram analysis technique for distinguishing different Acoustic Emission (AE) events. The peaks of the spectrogram are used to obtain a constellation map generating a "fingerprint" like pattern for each acoustic emission source. The fingerprints are then used within an artificial intelligence algorithm as part of a Knowledge Discovery database. The database is then able to link the AE signal to a specific source type. An experimental program was developed to test the methodology. The results of this study demonstrate that signal sources can be classified and linked to specific emission types with a high level of accuracy.