Fibre optic distributed corrosion sensor

Corrosion monitoring is an important aspect of increasing the safety and reliability of large structures such as aircraft, ships, submarines and bridges. In general, current non-destructive corrosion detection techniques are only able to detect local changes after significant damage has occurred. Moreover, corrosion often localises in areas that are inaccessible for inspection, especially in large structures. This implies a need for sensitive detection systems with multiple sensors in order to monitor the initial stages of corrosion over large areas. Thus, the development of a wide-area detection system for continuously monitoring structural changes, together with a capability for non-destructive evaluation and early warning, is an important task. The aim of the present research is to evaluate the feasibility of a distributed optical fibre sensor for detecting corrosion at arbitrary points of interest in a large structure. A detection method based on the estimation of water ingress into conventional optical fibre during the corrosion process was tested. This method was based on changes in the water peak absorption in a sensor made from conventional optical fibre. This did not give conclusive results. This method may be attractive for sensor designs based on hygroscopic glasses or fibres with small cladding thickness. In the method chosen for further development, the fluorescence from the complex Tris-(8-hydroxyquinoline)aluminum (Alq3) produced by the reaction of Al3+ ions with the indicator 8-HQ was detected. A theoretical model was developed to evaluate the evanescent field interaction mechanism and critical parameters of the detection system. This suggested that fluorescent emissions generated by the evanescent field and coupled back into guided modes could be detected by means of photon counting technology. Moreover, the location of the fluorescence could be resolved by measuring the time-offlight of the fluorescent emission returned from a short excitation pulse. Two constructions of the optical fibre sensor, based on etched D-shaped fibre and declad plastic-clad fibre, were examined as sensing elements for distributed measurements. Evaluation of the results in terms of practical considerations such as cost, additional development requirements and the risks to achieving the required results, led to the choice of a plastic clad fibre that could be easily declad as a sensing element for the detection system. A system based on pulsed excitation and time-gated photon counting was used to detect a localized source of fluorescence at an arbitrary position along the fibre length. Environments examined were the complex Alq3 excited at 405 nm and the infrared dye IR-125, excited at 780 nm. A model to characterize the response of the sensor to a Gaussian input pulse was developed and tested. Analysis of the Alq3 fluorescence response confirmed the validity of the model and enabled the fluorescence lifetime to be determined. The decay rate was found to be proportional to concentration, which is indicative of collisional deactivation. The intrinsic lifetime obtained (18.2+/-0.9 ns) is in good agreement with published data. The minimum Alq3 concentration that could be detected is 5x10^-5 M (~20 ppm) over an exposed length of 0.2 m, whereas the detection limit for IR-125 is 2x10^-6 M (~1.2 ppm). The model allows the spatial resolution of the distributed sensing system to be improved in those cases where the fluorophore lifetime is longer than the resolution of the sensing system. The absorbance and fluorescence response of the multimode optical fibre evanescent wave sensor was analysed to investigate the influence of different conditions, such as environmental refractive index, excitation wavelength, and launching conditions. Theoretical models were developed to investigate and improve the efficiency of the optical fibre detection system. The potential for using the developed detection system in practical sensing applications was successfully demonstrated in a stimulated corrosion test.