PSI - Issue 64

Paul Winkler et al. / Procedia Structural Integrity 64 (2024) 1264–1270 Winkler et al. / Structural Integrity Procedia 00 (2019) 000 – 000

1269

6

Generally, one can observe that, in accordance with increasing number of visible cracks in the higher loading stages, the detected events also spread more over the individual zones. For 60 kN, they mainly restrict to the central zones, where the load was applied and the first cracks formed. In contrast for 120 kN one of the upper zones shows an event concentration, which is most likely connected to the big crack that lead to the failure of the structure. The higher number of total events for the earlier stage can be connected to overall higher signal amplitudes and additional signals,

which are not caused by cracking of the structure. 5. Linking modal analysis and acoustic emission

Both the trend in reduction of eigenfrequencies as well as the spatial distribution of AE source locations have shown plausible results in correspondence to the experimental observations. If both methods provide valid damage indications, the results should show some correlation. To reinforce this conclusion, we show in Fig. 5 on the one hand the total eigenfrequency shift averaged over all completely identified eigenmodes and on the other hand the cumulative count of detected events at the individual sensors as well as cumulative number of events matched with the TDoA method. All quantities are shown over the different loading stages and normalized with respect to the final value at 160 kN. We see the number of matched signals shows a more constant increase over all load steps, while the main frequency change happens in the final loading steps. This might be caused by the fact, that in the higher loading stages, although the event number increases, the average amplitudes decreases, which reduces accuracy in arrival time estimation. Nevertheless, the correlation with the individual sensor counts is good, especially for sensor 2, which is closer to the big crack of the 120 kN stage. Another useful quantity could be the cumulative event energy. However, this would require at least a third sensor for unique source localization to use the source distance for the computation of comparative amplitudes.

Fig. 5: Changes in eigenfrequencies vs AE-events

6. Results and discussion Using the example of laboratory tests, the change in identified modal parameters and the detection of crack formation by structure-borne sound measurements were analysed for their suitability as indicators for damage detection at concrete component that was increasingly damaged by a static load. While the modal identification concerns the global structural behaviour, the measurement of structure-borne sound signals as a result of progressive crack formation can be classified as a local method. The results obtained with both methods were found to be comparable. However, it also became clear that damage detection based on modal parameter changes can be