Damage characterization of adhesively-bonded Bi-material joints using acoustic emission

Abstract

The aim of the present study is to characterize the damage in bi-material steel-to-composite double-lap adhesively-bonded joints using Acoustic Emission (AE). Two different structural adhesives, a ductile (Methacrylatebased) and brittle (Epoxy-based), were used to bond CFRP skins to a steel core. The fabricated joints were loaded in tension while damage evolution was monitored by AE. Due to the difference in the fracture nature of the adhesives "ductile vs. brittle", different damage mechanisms were observed; including cohesive failure within the adhesive layer, steel deformation, failure at the adhesive/adherends interface (adhesive failure) and delamination in the CFRP skin. To classify these damages by AE, the AE features of each damage mechanism were first obtained by conducting standard tests on the individual constituents. Then, these AE reference patterns were used to train an ensemble decision tree classifier. The best parameters of the ensemble model were obtained by Bayesian optimization, and the confusion matrix showed that the model was sufficiently trained with the accuracy of 99.5% and 99.8% for Methacrylate-based and Epoxy-based specimens respectively. Afterwards, the trained model was used to classify the AE signals of the double-lap specimens. The AE demonstrated that the dominant damage mechanisms in the case of the Methacrylate-based were cohesive and adhesive failures while in the case of the Epoxy-based they were CFRP skin failure and adhesive failure. These results were consistent with the Digital Image Correlation, Fiber Optic Sensor and camera results. This study demonstrates the potential of AE technique for damage characterization of adhesively-bonded bi-material joints.

Deformation and damage evolution of a full-scale adhesive joint between a steel bracket and a sandwich panel for naval application

“…The selection of adhesive was based on its favourable mechanical properties like high toughness and flexibility, shock resistance, excellent impact and peel strength, as these are considered dominant characteristics of an adhesive joint used in shipbuilding industries. 31,32 The nominal mechanical properties of steel AH36 33 and adhesive MMA 34 are summarised in Table 1.…”

Dynamic feature evaluation on streaming acoustic emission data for adhesively bonded joints for composite wind turbine blade

“…12,13,15,24,26 Consequently, it is significant to evaluate the similarities or the distances between AE features, which requires a large amount of AE signals. The most commonly used metric for the similarity measurement of AE signals is the Euclidean distance, 3,6,12,26,31 that is, the AE signals appear as a hypersphere in the manifold space. Other four non-Euclidean distances have also been evaluated, 32 that is, the Manhattan distance, the Minkowski distance, the Chebychev distance, and the Mahalanobis distance.…”

“…The label can be damage sources or damage locations, dependent on the purpose of prediction. Typical algorithms include classification and regression trees, 31,33 Multilayer Perception, 33 least-squares support vector machines, 34 support vector regression, 16 and the artificial neural network. 35 However, it is time-consuming and labor-intensive to construct a training library which contains a lot of labeled data.…”

Damage assessment of a titanium skin adhesively bonded to carbon fiber–reinforced plastic omega stringers using acoustic emission

“…Structural health monitoring (SHM) proposes continuous monitoring of the integrity of the structure by employing different techniques that have the capability of in situ monitoring, such as acoustic emission (AE) and fiber optic sensor. [23][24][25][26][27][28][29] Da´vila and Bisagni 30 performed a multi-instrumented compression fatigue test on the single-stringer-stiffened carbon fiber-reinforced plastic (CFRP) panels. They used UT, passive thermography, high-speed camera, and digital image correlation (DIC) to detect the propagation of the artificial debonding and to track the sequence of damage mechanisms up to the final fracture.…”