Ultrasonic resonance evaluation method for deep interfacial debonding defects of multilayer adhesive bonded materials

Hou H., Li J., Xia S., Meng Y., Shen J.

Abstract In the field of non-destructive testing (NDT), The detection of bonding defects between ultra-thin metal and silica gel is a difficult problem. In this study, In this study, ultrasonic resonance method was used to evaluate the bonding strength of ultra-thin metal to silica gel bonding structure. The composite parts of ultra-thin nickel sheet and silicon sheet with three different bonding states were studied. The bonding state of nickel sheet and silica gel is different, and the absorption of ultrasound is different. Using the resonance generated by high-frequency ultrasound in ultra-thin nickel sheet, the acoustic attenuation of the combination of ultra-thin nickel sheet and silicon rubber sheet was analyzed by resonance signal, and the bonding state between ultra-thin nickel sheet and silicon rubber sheet was characterized by bonding coefficient. Through experimental comparison, the results showed that the attenuation of ultrasonic signal in the nickel sheet and silicon film with different adhesive states characterize the adhesive state of ultra-thin nickel sheet and silicon film by the bonding coefficient, the bonding coefficient of good parts, weak adhesive parts and debonded parts is reduced successively. By setting an appropriate determination threshold value, the bonding state between the ultra-thin nickel sheet and the silicon film can be accurately determined according to the bonding coefficient obtained by detection.

Zhang M., Chen R., Zheng L., Yao J., Liu F., Chen Y.

• The thickness influence of the bonding layer on the reflection signals is specially considered. • A joint adaptive noise reduction and reconstruction method is proposed. • Improved back propagation neural network for nonlinear and non-stationary signals. • Both the lower and upper adhesive layer interfaces debonding can be identified. This paper investigates an electromagnetic ultrasonic method for debonding detection of adhesive structures. Considering the non-stationary and nonlinear characteristics of the electromagnetic ultrasonic reflection signals, a joint adaptive noise reduction and reconstruction (JANRR) method, namely, complete ensemble empirical mode decomposition(CEEMDAN) and singular spectrum analysis(SSA) based on continuous mean square Error(CMSE) are proposed, in which, mirror extension is used to suppress the endpoint effect of decomposition, and the order of SSA noise reduction is determined based on singular entropy theory. For two side debonding detection of the adhesive layer, a back propagation (BP) neural network classification model is developed, and five characteristics such as signal skewness factor are taken as the network inputs based on Chi square test. Experiments were carried out for validation, which show that the proposed method can distinguish both signals of the upper and lower interfaces of the adhesive layer under different bonding states.

Wang X., Wang J., Shen G., Li X., Huang Z.

Due to the difficulty in testing and evaluation interface bonding quality, an ultrasonic oblique incidence method is proposed from analyzing a wave propagation mechanism in multilayer composites. A new mathematical model of ultrasonic testing with acoustics parameters is derived based on the spring model and the boundary conditions at two interfaces. The correlations among the parameters, including the incidence angle, the stiffness coefficient, the frequency shift of the transmission coefficient peak (peak frequency), and the transmission coefficient are analyzed in this mathematical model. The sensitivity of transmission coefficients to bonding properties is revealed, and these detection parameters are also optimized. The numerical solution results in the bonding interface of aluminum (Al) and polymethyl methacrylate (PMMA) suggest that at the same incidence angle, with the increasing stiffness coefficient, the frequency of transmission peak decreases to the lowest at first, then increases, and finally tends to be stable. The transmission coefficient increases to the constant as the increase of stiffness coefficient. The ultrasonic testing sample is the same as the simulated material. The experimental results are basically consistent with numerical solution results. This method provides a theoretical basis for ultrasonic testing the interface quality from bonding structures of composite materials.

Spytek J., Ambrozinski L., Pieczonka L.

• Local wavenumber estimation enables characterization of defects in multilayer plates. • The wavenumber sensitivity depends strongly on damage topology. • Sensitivity study enables selection of optimal inspection parameters. • The disbond depth can be determined based on the theoretical dispersion curves. • The approach was validated experimentally using multilayer bonded aluminum plate. Multilayer plates are widely used as structural components, especially in high-tech industries. However, compared to much simpler homogeneous plates, they pose unique challenges to nondestructive testing (NDT) techniques. The inspections of multilayer plates using traditional ultrasonic (US) testing modalities can be challenging and time-consuming. Faster inspection can be achieved with the use of full-field techniques based on guided elastic waves. In particular, the local wavenumber estimation (LWE) of Lamb waves has been proven effective for damage detection in various plate-like structures. However, the effectiveness of LWE depends on multiple parameters, such as the excitation frequency, the choice of a Lamb wave mode, or the inspection side, among others. This is especially significant in the case of multilayer plates, where different types of damage result in a nontrivial behavior of Lamb waves. In this work, we present a framework for the evaluation of disbonds in adhesively bonded multilayer plates through local wavenumber estimation. The efficacy of the proposed approach is demonstrated on a test sample made of three layers of aluminum with different thicknesses bonded together with an epoxy adhesive. The sample contains artificial defects of various sizes introduced at different adhesive interfaces. First, we present a theoretical analysis of the dispersion curves in the test sample to determine the optimal LWE inspection parameters for particular types of defects. Next, the approach is validated experimentally based on a set of full-wavefield datasets acquired with a scanning laser Doppler vibrometer (SLDV). We demonstrate that the proposed methodology allows obtaining clear images of defects and determining their through-thickness locations in multilayer structures. Based on the obtained results, we believe that the proposed framework can also be effectively applied to different structures and damage types.

Yilmaz B., Smagulova D., Jasiuniene E.

This work proposes a cost-effective technique to evaluate the reliability of adhesive bonding quality assessment with ultrasonic NDT. Detectability of debonding defects and weak bonds in aluminum-epoxy single lap joints was estimated by the model-assisted probability of detection curves. For those adhesive joints containing three different bonding qualities –debonding at the interface, weak bond with release agent contamination, and weak bond due to faulty curing– numerical models were built. Ultrasonic wave propagation was simulated by using the semi-analytical finite element method via CIVA and validated with experimental investigations. In order to create a parametric study, uncertain parameters and their distribution were determined based on experiments and expertise. Ultrasonic signal responses in the time and frequency domain were analyzed as opposed to defect characteristic values. According to reliability analysis, it is shown that the detection probability of debonding is mostly based on the gate selection and ultrasonic echo amplitude, while weak bond detection requires a more delicate assessment. Frequency-based calculations can improve the detection reliability of weak bonds due to release agent contamination. Weak bonds caused by faulty curing can be detected via frequency domain maximum amplitude or the attenuation of the signal response.

Haldren H., Yost W.T., Perey D., Elliott Cramer K., Gupta M.C.

A primary mechanism of adhesive bond failure is a degradation of the adherent/adhesive interfacial stiffness from unwanted contamination or exposure to those environmental factors, which reduce adhesion quality. Substantial research has been conducted on the assessment of adhesively bonded structures and the detection of "kissing" bonds. Advanced ultrasonic assessment methods to interrogate bonded joints and measure interfacial stiffness using a distributed spring interface model have been developed. Amplitude-based ultrasonic methods have traditionally been used in adhesive bond quality assessment, but recent advancements in ultrasonic phase measurements allow for high measurement resolution with low-uncertainty. In this work, an ultrasonic phase technique for the monitoring of adhesively-bonded interfaces is demonstrated. Constant frequency measurements are obtained from the ultrasonic phase of the reflection coefficient from the adhesive bond with a glass adherent, where the degree of cure is controlled by exposure to ultraviolet light. A peak in the phase of the reflection coefficient, as predicted by the interfacial spring model, is measured experimentally. It is shown that the peak phase predicts the interfacial stiffness when some frequency dependent threshold value is crossed. With knowledge of the acoustic impedances of both materials at the interface, the interfacial stiffness is determined by an inverse algorithm involving measurements of ultrasonic phase shifts of bonded joint reflections. By monitoring the interface of bonded structures and coatings, this method permits a nondestructive inspection of bond strength from structural construction through its service life.

Guo X., Yang Y., Han X.

Debonding problems along the propellant/liner/insulation interface are a critical factor affecting the integrity of solid rocket motors and one of the major causes of their structural failure. Due to the complexity of interface debonding detection and its low accuracy, a method of wavelet packet transform (WPT) combined with machine learning is proposed. In this research, multi-layer structure specimens were prepared to simulate the structure of a solid rocket motor. First, ultrasonic non-destructive testing technology was used to obtain defect data. Then, WPT algorithm was employed to extract characteristic signals of the defect data. Moreover, k-nearest neighbor model, Random Forest model and support vector machine model were applied to the classification. The results showed that the accuracies of the three models were 84.67%, 90.66% and 95.33%, respectively. Positive results indicate that WPT with machine learning model exhibited excellent classification performance. Therefore, WPT combined with machine learning can achieve a precise classification of debonding defects and has the potential to assist or even automate the debonding inspection process of solid rocket motors.

Zheng S., Zhang S., Luo Y., Xu B., Hao W.

Detecting the debonding defects in the composite/rubber/rubber bonding structure is very difficult because of the high attenuation and aliasing of the ultrasonic wave from the structure. In order t...

Simões Hoffmann L.F., Parquet Bizarria F.C., Parquet Bizarria J.W.

Debonding problems along the propellant/liner/insulation interface are a critical point to the integrity and one of the major causes of structural failures of solid rocket motors. Current solutions are typically restricted to methods for assessing the integrity of the rocket motors structure and visually inspecting their components. In this context, this paper presents an improved algorithm to detect liner surface defects that may compromise the bonding between the solid propellant and the insulation. The use of Local Binary Patterns (LBP) provides a structural and statistical approach to texture analysis of liner sample images. Along with color information extraction, these two methods allow the representation of image pixels by feature vectors that are further processed by a Multilayer Perceptron (MLP) neural network classifier. The MLP neural network analyzes liner sample images and classifies each pixel into one of three classes: non-defect , foreign object , and defect . Several tests were executed varying different parameters to find the optimal MLP configuration, and as a result, the best classification accuracy of 99.08%, 90.66%, and 99.48% was achieved for the corresponding classes. Moreover, the defect size estimate showed that the MLP classifier correctly identified defects less than 1 mm long, with a relatively small number of training examples. Positive results indicate that the algorithm can identify liner surface defects with a performance similar to human inspectors and has the potential to assist or even automate the liner inspection process of solid rocket motors. • An algorithm to detect defective areas on the liner surface of solid rocket motors is proposed. • A neural network model classifies image pixels based on color and texture features. • The algorithm can detect defects on the liner surface with a precision comparable to human inspectors. • The algorithm can prevent the late characterization of bonding defects.

Smagulova D., Jasiuniene E.

Adhesive joints of bonded metal to composite materials are a recent development and in a high interest in automotive and aerospace industry. Current industrial progress substitutes metal components by carbon fiber reinforced plastic (CFRP) in a load bearing structures due to reduction of weight and energy saving. However, not all metal components can be replaced by CFRP. In this case, combination of metal and CFRP gives improved enhancement of the structure. It is challenging to detect disbonding type defects in bonding area between dissimilar joints due to different properties of metal and composite materials and multilayered structure. Though extensive work was done to evaluate interface bonding in similar material joints yet not enough attention was paid to dissimilar joints. This work presents high quality nondestructive testing (NDT) process which includes quantitative evaluation, experimental study and model assisted probability of detection (MAPOD) evaluation. Ultrasonic pulse echo technique was used for inspection of adhesively bonded aluminum to CFRP in order to detect disbonds which occur between dissimilar layers. Computer modelling was performed in order to learn behavior of ultrasonic wave transmitted and reflected in the case of dissimilar joints. To verify modelling results experimental study was performed using 10MHz focused transducer in immersion pulse echo mode. Sensitivity analysis of parameters that can influence the measurement was performed as well as MAPOD assessment.

Loukkal A., Lematre M., Bavencoffe M., Lethiecq M.

The microelectronics industry is expressing an increased demand for the development of non-destructive tools and methods for health control and diagnostics in multilayered structures. The purpose of these tools is to detect problems such as delaminations, inclusions and microcracks. The aim of this paper is to study the effect of imperfect interfaces on the wave propagation in multilayered structures. This type of structure represents the typical architecture of many microelectronic components. This study will be based on the calculation of the reflection coefficient and the guided waves dispersion curves. The investigated structure is an isotropic trilayer where two metallic layers are bonded together by an adhesive layer made of an epoxy resin. Comparisons were performed in order to evaluate numerically the influence of several properties of the adhesive layer on the guided waves behavior. In addition, an imperfect viscoelastic interface layer model [1] has been implemented in order to simulate different adherence qualities between the metallic layers.

Sahoo S.K., Narasimha Rao R., Srinivas K., Buragohain M.K., Sri Chaitanya C.

Aerospace structural components are made by adhesively bonded layers of different materials for achieving specific properties for industrial applications. One such advanced structural material is made of glass fibre reinforced plastic (GFRP) composite. This is a layered engineered material with composite adhesively bonded to rubber to one of the surfaces. Such engineered materials can withstand structural, thermal and corrosive environments. GFRP composite have to be bonded with rubber using rubber solution or other adhesive to keep them structurally bonded. Any separation of these two layers will make the engineered structure redundant. In this paper, we report application of various NDE methods for evaluating bonded interfaces of engineered structural composite materials. Conventional methods such as X-ray radiography and ultrasonic (acousto-ultrasonic) methods were experimentally verified for identifying defects at the adhesively bonded interfaces. Disadvantages of these conventional methods were overcome by using advanced single sided nuclear magnetic resonance (NMR) technique. The use of low frequency single sided portable NMR for characterising defects at the adhesively bonded interfaces is discussed. In addition to the defect detection, application of NMR for measuring thickness of adhesive layers coated over rubber is also presented. Advantages of using NMR technique over other conventional methods is discussed.

Mondet B., Brunskog J., Jeong C., Rindel J.H.

In room acoustic simulations the surface materials are commonly represented with energy parameters, such as the absorption and scattering coefficients, which do not carry phase information. This paper presents a method to transform statistical absorption coefficients into complex surface impedances which are needed for phased or time-domain calculation methods. Two 5-parameter impedance models based on fractional calculus are suggested to achieve a general model for common acoustic materials, thereby ensuring that the impedance found has a physical meaning. The five parameters for the general models are determined by solving an inverse problem with an optimization method. Due to the non-uniqueness of retrieving complex-valued impedances from real-valued absorption coefficients, prior information about the absorber of interest can be used as constraints, which is shown to help determine the impedance more correctly. Known material models, such as Miki’s and Maa’s models, are taken as references to assess the validity of the suggested model. Further stability and sensitivity investigations indicate that the method presented constitutes an efficient solution to convert sound absorption coefficients back to their original complex surface impedances.

Jasiūnienė E., Mažeika L., Samaitis V., Cicėnas V., Mattsson D.

Ultrasonic inspection is widely used for non-destructive evaluation of composite adhesive joints. However, there are serious challenges in applying ultrasonic testing on metal to composite hybrid joints, because they are multi-layered, made out of dissimilar materials and relatively thin. The ultrasonic signals reflected by different layers are overlapped, scattered and attenuated. The aim of this research was to develop an ultrasonic inspection technique suitable for defect detection in hybrid metal to composite joints where the metal part has pin arrays, which entangle with the composite part. The immersion pulse echo technique was used to collect data. In order to overcome the problems related to the rough surface and non-parallel layers a novel signal post-processing algorithm for reconstruction of the joint area was developed and validated experimentally. It is shown that using the proposed technique the positions of different defects can be determined.

Yang H., Xiao Y., Zhao H., Zhong J., Wen J.

The underwater acoustic screens made of periodically perforated rubber layers with metal plates are often used to enhance reflections and reduce transmissions of sound in water, under high hydrostatic pressure. Previous studies have tended to focus on the acoustic response of finite-thickness structures of the screens, rather than on the wave propagation properties of the corresponding infinite crystal structures. In this work, a numerical method, which combines finite element method and the layer-iteration technique, is developed to study the complex band structure of the infinite crystals and the acoustic response of the finite-thickness structures for the screens. Numerical results for the screens with elastic and viscoelastic rubber materials are presented for analyzing the features of propagation modes, band gaps, and attenuation of the waves in the screens. Reflection, transmission and absorption spectra of the screens are compared with the corresponding complex band structure of the infinite crystals in detail for providing a comprehensive understanding of the wave propagation and attenuation properties of the screens.

Wang C., Wang X., Khumalo A., Jiang F., Lv J.

Machine vision was utilized in this study to accurately classify the low concentration slurry. Orthogonal experiment L9(34) indicated that the optimal coal slurry collection images were achieved with exposure value of 10, slurry layer thickness of 7 cm, and light intensity of 5 × 104 lux. Subsequently, a new low concentration classification model was systematically developed, encompassing aspects such as original image acquisition, data augmentation, dataset partitioning, classification algorithm design, and model evaluation. DCGAN was employed for image generation, achieving favorable outcomes with generator learning rate set at 5 × 10− 5, discriminator at 1 × 10− 6, and iteration number at 2000. At the point, the maximum SSIM similarity reached 0.9381, and the pHash similarity was 0.9375. Results from subsequent CNN model training, with 200 iterations, the accuracy on training and validation sets was demonstrated over 95% for coal slurry concentration prediction. Further evaluation using recall, precision, and F1-score revealed CNN network model metrics: maximum recall 1.000, minimum 0.800; maximum precision 1.000, minimum 0.700; and highest F1 score 1.000, lowest 0.778. Additionally, the accuracy of this model on the test set reached as high as 94%. The findings indicated the excellent performance in low concentration detection of coal slurry throughout this study.

Zhang Q., Guo C., Cheng G., Song S., Ding J.