In this study, we focus on analyzing the acoustic emission waveforms of the fatigue crack growth despite the conventional statistics-based analysis of acoustic emission. The acoustic emission monitoring technique is a well-known approach in the non-destructive evaluation/structural health monitoring research field. The growth of the fatigue crack causes the acoustic emission in the material that propagates in the structure. The acoustic emission happens not only from the crack growth but also from the interaction of the crack tips during the fatigue loading in the structure. The acoustic emission waveforms are generated from the acoustic emission events; they propagate and create local vibration modes along the crack faces (crack resonance). In-situ fatigue and acoustic emission experiments were conducted to monitor the acoustic emission waveforms from the fatigue cracks. Several test specimens were used in the fatigue experiments, and corresponding acoustic emission waveforms were captured. The acoustic emission waveforms were analyzed and distinguished into three types based on the similar nature in both time and frequency domains. Three-dimensional harmonic finite element analyses were performed to identify the local vibration modes. The local crack resonance phenomenon has been observed from the finite element simulation that could potentially give the geometric information of the crack. The laser Doppler vibrometry experiment was performed to identify the crack resonance phenomenon, and the experimental results were used to verify the simulated results.

The acoustic emission (AE) waveforms from a fatigue crack advancing in a thin metallic plate possess diverse and complex spectral signatures. In this article, we analyze these waveform signatures in coordination with the load level during cyclic fatigue. The advancing fatigue crack may generate numerous AE hits while it grows under fatigue loading. We found that these AE hits can be sorted into various groups based on their AE waveform signatures. Each waveform group has a particular time-domain signal pattern and a specific frequency spectrum. This indicates that each group represents a certain AE event related to the fatigue crack growth behavior. In situ AE-fatigue experiments were conducted to monitor the fatigue crack growth with simultaneous measurement of AE signals, fatigue loading, and optical crack growth measurement. An in situ microscope was installed in the load-frame of the mechanical testing system (MTS) to optically monitor the fatigue crack growth and relate the AE signals with the crack growth measurement. We found the AE signal groups at higher load levels (75%–85% of maximum load) were different from the AE signal groups that happened at lower load levels (below 60% of load level). These AE waveform groups are highly related to the fatigue crack-related AE events. These AE signals mostly contain the higher frequency peaks (100 kHz, 230 kHz, 450 kHz, 550 kHz). Some AE signal groups happened as a clustered form that relates a sequence of small AE events within the fatigue crack. They happened at relatively lower load level (50%–60% of the maximum load). These AE signal groups may be related to crack friction and micro-fracture during the friction process. These AE signals mostly contain the lower frequency peaks (60 kHz, 100 kHz, 200 kHz). The AE waveform based analysis may give us comprehensive information of the metal fatigue.

In this article, ultrasonic guided wave propagation and interaction with the rivet hole cracks has been formulated using closed-form analytical solution while the local damage interaction, scattering, and mode conversion have been obtained from finite element analysis. The rivet hole cracks (damage) in the plate structure gives rise to the non-axisymmetric scattering of Lamb wave, as well as shear horizontal (SH) wave, although the incident Lamb wave source (primary source) is axisymmetric. The damage in the plate acts as a non-axisymmetric secondary source of Lamb wave and SH wave. The scattering of Lamb and SH waves are captured using wave damage interaction coefficient (WDIC). The scatter cubes of complex-valued WDIC are formed that can describe the 3D interaction (frequency, incident direction, and azimuth direction) of Lamb waves with the damage. The scatter cubes are fed into the exact analytical framework to produce the time domain signal. This analysis enables us to obtain the optimum design parameters for better detection of the cracks in a multiple-rivet-hole problem. The optimum parameters provide the guideline of the design of the sensor installation to obtain the most noticeable signals that represent the presence of cracks in the rivet hole.

This paper presents the interaction of Lamb waves with rivet hole cracks from multiple directions of incident using the finite element approach. Lamb waves undergo scattering and mode conversion after interacting with the damage. Shear horizontal waves appear in the scattered waves because of the mode conversion. Instead of analyzing the whole large structure, the local damage area is analyzed using finite element analyses and analytical formulation is used to analyze the whole structure. The scatter fields are described in terms of wave damage interaction coefficients that involve scattering and mode conversion of Lamb waves. Lamb wave mode (S0 and A0) hit the damage from multiple directions and corresponding wave damage interaction coefficients are obtained around the damage. Harmonic analysis has been performed over the fundamental frequency domain and “scatter cubes” of complex-valued wave damage interaction coefficients are formed. The scatter cube provides the information of relative amplitude and phase of scattered waves around the damage that can be used for designing the sensor installation. An application based on real time domain signal has been illustrated for the problem of multiple-rivet-hole cracks using the scatter cubes with the analytical framework.

Piezoelectric wafer active sensors (PWAS) are commonly used for detecting Lamb waves for structural health monitoring application. However, in most applications of active sensing, the signals are of high-amplitude and easy to detect. In this article, we have shown a new avenue of using the PWAS transducer for detecting the low-amplitude fatigue-crack related acoustic emission (AE) signals. Multiphysics finite element (FE) simulations were performed with two PWAS transducers bonded to the structure. Various configurations of the sensors were studied by using the simulations. One PWAS was placed near to the fatigue-crack and the other one was placed at a certain distance from the crack. The simulated AE event was generated at the crack tip. The simulation results showed that both PWAS transducers were capable of sensing the AE signals. To validate the multiphysics simulation results, an in-situ AE-fatigue experiment was performed. Two PWAS transducers were bonded to the thin aerospace test coupon. The fatigue crack was generated in the test coupon which had produced low-amplitude acoustic waves. The low-amplitude fatigue-crack related AE signals were successfully captured by the PWAS transducers. The distance effect on the captured AE signals was also studied. It has been shown that some high-frequency contents of the AE signal have developed as they travel away from the crack.