PSI - Issue 52

Hongmin Zhu et al. / Procedia Structural Integrity 52 (2024) 679–689 Zhu et al./ Structural Integrity Procedia 00 (2019) 000 – 000

686

8

that effectiveness and localisation accuracy of the IB technique highly depend on the sensor positionings. As discussed earlier, the IB technique could fail to well localise damages due to the lack of sufficient equivalent paths, and such failure leads to much larger localisation errors compared to the other techniques. Furthermore, when temperature increases from 20℃ to 40℃ the localisation errors of the reciprocity-based method are within approximately 2 while the VTR technique seems slightly sensitive to temperature variations and such sensitivity could be due to the temperature-dependent behaviour of the time-reversibility. Simultaneously, the VTR and the reciprocity-based techniques have consistence and similar standard deviations (approximately 4 ) of under different temperature levels, which validates the robustness of such baseline-free techniques under varying temperatures.

Fig. 5. Comparisons between the IB, Reciprocity and VTR on localisation errors of single damages under 20℃ to 40℃ with a 5℃ gradient. The IB, reciprocity principle and the VTR techniques are further validated on the 12-sensors plate with simulated dual damages. 10 dual-damage scenarios are investigated where damages with the same radius as 15 are located at various locations, and the results of four typical scenarios are illustrated in Fig. 6. More specifically, it is shown that under scenarios 1,3 and 4 , where two adjacent damages are situated at off-centre locations, the reciprocity-based method is competent to identify and localise the dual damages at the same time despite that under scenario 4 estimated damage areas may overlap with each other, and such overlapping consequently complicates well separation of the damages. In contrast, the IB and the VTR techniques are able to detect increasing pixel values around the true damage locations but may fail to concurrently localise the dual damages under scenario 1 and 3 where one putty damage is masked or overlapped by another in the imaging results. Meanwhile, when adjacent damages are located near the centre of circular transducer array, i.e., scenario 2, such dual damages can be detected by the VTR and the reciprocity based approach but coming with the risk of overlapping pixel values in the middle of two damages. Under the same scenario the IB technique can only localise single damage while ignoring another left near-centre damage, which could be due to the lack of sufficient equivalent through-centre paths for comparisons. 5. Conclusions In this paper, several baseline-free methods in the field of GWSHM, particularly the VTR, IB and reciprocity-based techniques, are examined for the damage localisation purposes. Further investigations focus on the robustness and effectiveness of these techniques when localising damages under temperature variations and dual damage scenarios. In this regard, related experiments are carried out where the simulated damages are implemented on an anisotropic CFRP plate, while the localisation error is then deployed to compare the localisation accuracy of the baseline-free techniques. It is shown that the compared baseline-free techniques are able to localise single damages under temperature variations of 20℃ . Meanwhile, the baseline-free techniques also have the capability to localise dual damages despite the overlapping pixels between the damages. Future investigations are needed to improve the performance of the baseline free techniques when localising the dual damages.