Acoustic emission detection and prediction of fatigue crack propagation in composite patch repairs using neural network

Abstract

This article presents the results of acoustic emission (AE) monitoring of crack propagation in 2024-T3 clad aluminum panels repaired with adhesively bonded octagonal and elliptical boron/epoxy composite patches using FM-73 adhesive under tension-tension fatigue loading. Two crack propagation gages and four broadband AE sensors were used to monitor crack initiation and propagation, respectively. The acquired AE signals were processed in time and frequency domain to identify sensor features correlated with fatigue cycle and crack propagation, which were used to train neural networks for predicting crack length. The results show that AE events are correlated with crack propagation, and crack propagation signals can be differentiated from signals due to matrix cracking, fiber breakage, and shear of the composite patch. Three backpropagation cascade feed-forward networks were trained to predict crack length using number of fatigue cycles, number of AE events, and number of fatigue cycles and number of AE events together as inputs, respectively. It was found that network with fatigue cycles as input gave good results, while the network with just AE events as input gave greater error. However, the network using both fatigue cycles and number of AE events as inputs to predict crack length gave much better results.

Using acoustic emission to understand fatigue crack growth within a single load cycle

“…AE is under active investigation as a means of locating damage and measuring damage size in composite or adhesively bonded materials, as e.g. recently reported in [16][17][18][19]. It should be stressed that this was not the purpose of the current research.…”

Analysis of the damage mechanisms in mixed-mode delamination of laminated composites using acoustic emission data clustering

“…Furthemore, the frequency ranges of pure epoxy resin and glass fiber bundle under tensile load were investigated, and the results indicated that the dominant frequency ranges of the matrix cracking, debonding, and fiber breakage are at 140-250 kHz, 250-350 kHz, and 350-450 kHz, respectively. 9,28,33,34 Therefore, the AE events and damage mechanisms could be related to each other using In the case of specimens U1, U2, and U3, fiber bridging event and resistance of the fibers against crack opening occurs after initiation of the delamination (see Figure 13). However, in specimen U4 and the woven specimens, fiber bridging event is not observed by visual inspection.…”

“…[3][4][5][6] Until now, some researchers discriminated the AE signatures of damage mechanisms in composite materials. [7][8][9][10][11][12] Their results showed that AE technique is capable of characterizing different damage mechanisms in thermoplastic and thermoset-laminated composite materials. Most of these studies exploited AE parameters such as amplitude, count, and energy for characterizing the damage mechanisms.…”

“…Additionally, the achieved results specified that higher frequencies and amplitudes emerging in the tests were related to fiber breakage, while lower ranges contributed to matrix cracking and fiber-matrix debonding damages. 2,8,9,10 In order to obtain more reliable and accurate information for identification of damage mechanisms, pattern recognition algorithms and time-frequency analyses were utilized as multivariable classification procedures of AE data. [13][14][15][16] These methods made it possible to study the delamination damage mechanisms by simultaneous consideration of various AE features.…”

Failure Mode Analysis of Carbon Fiber Composite Laminates by Acoustic Emission Signals

“…In addition, [16,17] analyzed the acoustic emission signal to locate the damage as well as the mechanical properties of fiber composite materials. References [18][19][20][21] used several pattern recognition algorithms to analyze the collected AE signals, such as principal component analysis, K-means genetic algorithm, and fuzzy C-means. All these references show that AE technology is suitable for real-time dynamic monitoring and detection of CFRP.…”