Abstract
Composite repairs are often used for damaged structures in order to recover the mechanical properties of the original structure. During service, there is the possibility that damage will occur in the repaired region and hence it would be useful to be able to monitor such repairs. This research investigates the use of chirped fibre Bragg grating (CFBG) sensors to monitor the development of fatigue damage initiation and growth in the repaired region of three different repair systems, i.e. glass fibre reinforced polymer (GFRP)-to-GFRP scarf repair, GFRP-to-GFRP patch repair and a GFRP patch repair to an aluminium panel. All of these repairs were investigated using a combination of experimental testing and theoretical predictions using finite-element analysis and optical prediction software. For each repair system, the CFBG sensor was embedded in the bond-line during the repair fabrication and the transparency of the GFRP material enabled damage to be observed and recorded. The work began by fatigue loading the GFRP-to-GFRP scarf repair coupon under 4-point bending. The CFBG sensor was embedded in the tensile side of the repair. For this system, the growth of the bond-line cracks could be detected but an accurate determination of the extent of damage was not possible. This was mostly due to the geometry of the scarf repair which led to a high degree of complexity in the interpretation of the data. There was good agreement between the trend of the changes in the spectra in the comparisons of the experimental results and finite-element/optical modelling, although the experimental spectra showed smaller changes than were produced by the modelling. The second repair system investigated was the GFRP-to-GFRP patch repair which was tested in the same way as the scarf repair system. An asymmetric repair was fabricated with the patch being bonded on the tensile face of the coupon. Here, it can be said that growth of the bond-line cracks can be detected using CFBG sensors, and an accurate determination of the current length of the cracks from the spectra was achieved. A explanation of the shift of the low-wavelength end of the spectrum changes with increasing crack growth was provided in terms of strain field change caused by the bond-line cracks. There was good agreement in the comparisons of the experimental results and finite-element/optical modelling. Finally, the third repair system investigated was the patch repair of an aluminium panel which was fatigue loaded in tension. Prior to repair, the aluminium panel was notched at the centre in order to promote crack initiation. Again, an asymmetric repair was fabricated. It can be concluded from the results that the embedded sensor could clearly detect the approach of a fatigue crack and indicate when the crack had passed the location of the sensor. Again, there was good agreement in the comparisons of the experimental results and finite-element/optical modelling. The work has shown that CFBG sensors can be used to monitor damage development in various types of repairs and can give an indication of damage initiation for all of the cases investigated. However, where there is significant geometrical complexity to the repair, as in the case of a scarf repair, detailed interpretation of the spectra in order to extract information on damage growth is much more difficult.