17 An Experimental Investigation of Feature Availability in Nominally Identical Structures for Population-Based SHM 191 common patterns across structures in order to monitor new untested structures without the need of a full modelling or a complete knowledge of their characteristics. In order to do this, two wingtail sections of a Piper PA-28 ‘Cherokee’ and of one of its variants, the PA-28 ‘Arrow’ were acquired, and subsequently separated in half, creating thus two pairs of nominally identical structures. The ‘Arrow’ sections are subsets of the ‘Cherokee’ sections with shorter length and smaller mass. The natural frequencies and the mode shapes of the structures had already been compared in the previous study, but were briefly shown here as well to deviate, mainly due to manufacturing uncertainties, and potentially because of measurement errors. However, there were common patterns present. In order to explore those patterns in an SHM scheme, an added mass was used during testing in all of the structures and by using its effect on the FRFs, potential features were selected. Outlier analysis was the chosen method for ‘damage’ (added mass) detection and novelty detectors were built for the structures where the features were originally selected, but then subsequently tested with all the rest. It was shown that by choosing arbitrarily an FRF from a structure, without using the rest of the measurements from the other structures, then features can be used to identify the presence of mass successfully in all of them. Future work will address several of the testing inconsistencies with the help of a scanning laser vibrometer and also alter the structures more by actually damaging them or modify certain parts, e.g. stiffeners or replace rivets, to further explore the use of features in a population-based scheme. Acknowledgements The support of the UK Engineering and Physical Sciences Research Council through grant reference EP/J016942/1 is greatly acknowledged. The authors would also like to thank Mr Jamie Park for his help during the experimental tests. References 1. Lombardo JS, Buckeridge DL (2007) Disease surveillance, a public health informatics approach. Wiley, New York 2. Deering S, Manson G, Worden K, Allen DW, Farrar CR, Lombardo JS (2008) Syndromic surveillance as a paradigm for shm data fusion. In: Proceedings of the 4th European workshop on structural health monitoring, Crakow, Poland, 4–6 July 2008 3. Papatheou E, Rahman TAZ, Barthorpe RJ, Park J, Worden K (2014) An experimental investigation of feature complexity and diversity in nominally similar test structures. In: Proceedings of ISMA2014, Leuven, Belgium 4. Farrar CR, Worden K (2013) Structural health monitoring, a machine learning perspective. Wiley, New York 5. Ewins DJ (2000) Modal testing: theory, practice and application, 2nd edn. Wiley, New York 6. Peeters B, Van Der Auweraer H, Guillaume P, Leuridan J (2004) The polymax frequency-domain method: a new standard for modal parameter estimation? Shock Vib 11(3–4):395–409 7. Maia NMM, Silva JMM (1998) Theoretical and experimental modal analysis. Research Studies Press LTD, Baldock, Hertfordshire, England 8. Papatheou E, Manson G, Barthorpe RJ, Worden K (2010) The use of pseudo-faults for novelty detection in shm. J Sound Vib 329(12): 2349–2366 9. Papatheou E, Manson G, Barthorpe RJ, Worden E (2014) The use of pseudo-faults for damage location in shm: an experimental investigation on a piper tomahawk aircraft wing. J Sound Vib 333(3):971–990 10. Worden K, Manson G, Fieller NRJ (2000) Damage detection using outlier analysis. J Sound Vib 229(3):647–667
RkJQdWJsaXNoZXIy MTMzNzEzMQ==