Damage Severity Assessment of Multi-Layer Complex Structures Based on a Damage Information Extraction Method with Ladder Feature Mining.

Multi-layer complex structures are widely used in large-scale engineering structures because of their diverse combinations of properties and excellent overall performance. However, multi-layer complex structures are prone to interlaminar debonding damage during use. Therefore, it is necessary to monitor debonding damage in engineering applications to determine structural integrity. In this paper, a damage information extraction method with ladder feature mining for Lamb waves is proposed. The method is able to optimize and screen effective damage information through ladder-type damage extraction. It is suitable for evaluating the severity of debonding damage in aluminum-foamed silicone rubber, a novel multi-layer complex structure. The proposed method contains ladder feature mining stages of damage information selection and damage feature fusion, realizing a multi-level damage information extraction process from coarse to fine. The results show that the accuracy of damage severity assessment by the damage information extraction method with ladder feature mining is improved by more than 5% compared to other methods. The effectiveness and accuracy of the method in assessing the damage severity of multi-layer complex structures are demonstrated, providing a new perspective and solution for damage monitoring of multi-layer complex structures.

Impact monitoring of large size complex metal structures based on sparse sensor array and transfer learning.

During aircraft operations, the impact events experienced by the aircraft may cause damage to the structure, thus posing a safety hazard. Therefore, an accurate determination of where the impact occurred and the time history of the impact force can provide an important basis for assessing the condition of the aircraft. However, modern aircraft structures are often large and complex, and relying on dense arrays of sensors for monitoring adds additional weight to the aircraft and reduces the economics of aircraft operation. This paper proposes a region-to-point monitoring strategy. First, a Convolutional Neural Network (CNN) model with region localization capability is trained using the sparse sensor array acquisition data. Then, the weighted center algorithm is used to determine the specific location where the impact occurs, in which the added fuzzy genetic algorithm can provide the ability to adjust weights to improve localization accuracy adaptively. As for the impact force prediction, this paper adopts a model based on a Convolutional Neural Network-Gated Recurrent Unit combined with a Squeeze-Excitation attention mechanism (CNN-GRU-SE), which is capable of predicting the impact force occurring in the flat plate and reinforced structure region of the aircraft under different energy conditions. Finally, the impact of incorporating a transfer learning approach on model performance and training cost is investigated for fuselage regions with different structures.

Corrosion quantification of plate-type structures using Lamb wavefield and monogenic signal processing.

Corrosion is one of the most common damage types which seriously affects structural safety. In this paper, a Lamb wavefield-based monogenic signal processing algorithm is proposed to quantify the corrosion parameters, including location, area, shape and depth, in plate-type structures. The monogenic signal processing based on Riesz transform will cause a serious problem, that is, phase wrapping. To solve this problem, a robust fast phase unwrapping algorithm is developed. Then, the phase spatial distribution of the extracted Lamb wavefield can be extracted, which can be used to calculate the wavenumber distribution. The wavenumber distribution is related to the structure thickness or corrosion depth, which can be further used for corrosion imaging. Simulated Lamb wavefield signals calculated by finite element simulation are employed to evaluate the parameters of circular corrosion and complex umbrella corrosion. The results show that the proposed algorithm has a great advantage in corrosion identification accuracy and calculation time compared with the existing algorithms. A completely non-contact laser ultrasonic system is established for acquiring Lamb wavefield containing square corrosion, and it is proved that the proposed algorithm is able to quantify the corrosion location, area, shape and depth with good accuracy in the experiment.

Distributed Estimation With Cross-Verification Under False Data-Injection Attacks.

Under false data-injection (FDI) attacks, the data of some agents are tampered with by the FDI attackers, which causes that the distributed algorithm cannot estimate the ideal unknown parameter. Due to the concealment of the malicious data tampered with by the FDI attacks, many detection algorithms against FDI attacks often have poor detection results or low detection efficiencies. To solve these problems, a conveniently distributed diffusion least-mean-square (DLMS) algorithm with cross-verification (CV) is proposed against FDI attacks. The proposed DLMS with CV (DLMS-CV) algorithm is comprised of two subsystems: one subsystem provides a detection test of agents based on the CV mechanism, while the other provides a secure distribution estimation. In the CV mechanism, a smoothness strategy is introduced, which can improve the detection performance. The convergence performance of the proposed algorithm is analyzed, and then the design method of the adaptive threshold is also formulated. In particular, the probabilities of missing alarm and false alarm are examined, and they decay exponentially to zero under sufficiently small step size. Finally, simulation experiments are provided to illustrate the effectiveness and simplicity of the proposed DLMS-CV algorithm in comparison to other algorithms against FDI attacks.

A transformer model with enhanced feature learning and its application in rotating machinery diagnosis.

Deep learning has become the prevailing trend of intelligent fault diagnosis for rotating machines. Compared to early-stage methods, deep learning methods use automatic feature extraction instead of manual feature design. However, conventional intelligent diagnosis models are trapped by a dilemma that simple models are unable to tackle difficult cases, while complicated models are likely to over-parameterize. In this paper, a transformer-based model, Periodic Representations for Transformers (PRT) is proposed. PRT uses a dense-overlapping split strategy to enhance the feature learning inside sequence patches. Combined with the inherent capability of capturing long range dependencies of transformer, and the further information extraction of class-attention, PRT has excellent feature extraction abilities and could capture characteristic features directly from raw vibration signals. Moreover, PRT adopts a two-stage positional encoding method to encode position information both among and inside patches, which could adapt to different input lengths. A novel inference method to use larger inference sample sizes is further proposed to improve the performance of PRT. The effectiveness of PRT is verified on two datasets, where it achieves comparable and even better accuracies than the benchmark and state-of-the-art methods. PRT has the least FLOPs among the best performing models and could be further improved by the inference strategy, reaching an accuracy near 100%.

Qualitative Classification of Lubricating Oil Wear Particle Morphology Based on Coaxial Capacitive Sensing Network and SVM.

In addition to lubricating and cooling, aero-engine lubricating oil is also a transport medium for wear particles generated by mechanical wear. Online identification of the number and shape of wear particles is an important means to directly determine the wear state of rotating parts, but most of the existing research focuses on the identification and counting of wear particles. In this paper, a qualitative classification method of wear particle morphology based on support vector machine is proposed by using the wear particle capacitance signal obtained by the coaxial capacitive sensing network. Firstly, the coaxial capacitive sensing network simulation model is used to obtain the capacitance signals of different shapes of wear particles entering the detection space of different electrode plates. In addition, a variety of intelligent optimization algorithms are used to optimize the relevant parameters of the support vector machine (SVM) model in order to improve the classification accuracy. By using the processed data and optimized parameters, a SVM-based qualitative classification model for wear particles is established. Finally, the validity of the classification model is verified by real wear particles of different sizes. The simulation and experimental results show that the qualitative classification of different wear particle morphologies can be achieved by using the coaxial capacitive sensing network signal and the SVM model.

Lamb Wave-Based Damage Localization and Quantification in Composites Using Probabilistic Imaging Algorithm and Statistical Method.

Quantitatively and accurately monitoring the damage to composites is critical for estimating the remaining life of structures and determining whether maintenance is essential. This paper proposed an active sensing method for damage localization and quantification in composite plates. The probabilistic imaging algorithm and the statistical method were introduced to reduce the impact of composite anisotropy on the accuracy of damage detection. The matching pursuit decomposition (MPD) algorithm was utilized to extract the precise TOF for damage detection. The damage localization was realized by comprehensively evaluating the damage probability evaluation results of all sensing paths in the monitoring area. Meanwhile, the scattering source was recognized on the elliptical trajectory obtained through the TOF of each sensing path to estimate the damage size. Damage size was characterized by the Gaussian kernel probability density distribution of scattering sources. The algorithm was validated by through-thickness hole damages of various locations and sizes in composite plates. The experimental results demonstrated that the localization and quantification absolute error are within 11 mm and 2.2 mm, respectively, with a sensor spacing of 100 mm. The algorithm proposed in this paper can accurately locate and quantify damage in composite plate-like structures.

A New In Situ Coaxial Capacitive Sensor Network for Debris Monitoring of Lubricating Oil.

Wear debris monitoring of lubricant oil is an important method to determine the health and failure mode of key components such as bearings and gears in rotatory machines. The permittivity of lubricant oil can be changed when the wear debris enters the oil. Capacitive sensing methods showed potential in monitoring debris in lubricant due to the simple structure and good response. In order to improve the detection sensitivity and reliability, this study proposes a new coaxial capacitive sensor network featured with parallel curved electrodes and non-parallel plane electrodes. As a kind of through-flow sensor, the proposed capacitive sensor network can be in situ integrated into the oil pipeline. The theoretical models of sensing mechanisms were established to figure out the relationship between the two types of capacitive sensors in the sensor network. The intensity distributions of the electric field in the coaxial capacitive sensor network are simulated to verify the theoretical analysis, and the effects of different debris sizes and debris numbers on the capacitance values were also simulated. Finally, the theoretical model and simulation results were experimentally validated to verify the feasibility of the proposed sensor network.

Real-Time Life-Cycle Monitoring of Composite Structures Using Piezoelectric-Fiber Hybrid Sensor Network.

In this paper, an in situ piezoelectric-fiber hybrid sensor network was developed to monitor the life-cycle of carbon fiber-reinforced plastics (CFRPs), from the manufacturing phase to the life in service. The piezoelectric lead-zirconate titanate (PZT) sensors were inserted inside the composite structures during the manufacturing process to monitor important curing parameters, including the storage modulus of resin and the progress of the reaction (POR). The strain that is related to the storage modulus and the state of resin was measured by embedded fiber Bragg grating (FBG) sensors, and the gelation moment identified by the FBG sensors was very close to those determined by dynamic mechanical analysis (DMA) and POR. After curing, experiments were conducted on the fabricated CFRP specimen to investigate the damage identification capability of the embedded piezoelectric sensor network. Furthermore, a modified probability diagnostic imaging (PDI) algorithm with a dynamically adaptive shape factor and fusion frequency was proposed to indicate the damage location in the tested sample and to greatly improve the position precision. The experimental results demonstrated that the average relative distance error (RDE) of the modified PDI method was 68.48% and 46.97% lower than those of the conventional PDI method and the PDI method, respectively, with an averaged shape factor and fusion frequency, indicating the effectiveness and applicability of the proposed damage imaging method. It is obvious that the whole life-cycle of CFRPs can be effectively monitored by the piezoelectric-fiber hybrid sensor network.

A New Laser Ultrasonic Inspection Method for the Detection of Multiple Delamination Defects.

Delamination is one of the most common types of defects for carbon fiber reinforced plastic (CFRP) composites. The application of laser techniques to detect delamination faces difficulties with ultrasonic wave excitation because of its low thermal conductivity. Much of the research that can be found in the literature has only focused on the detection of a single delamination. In this study, aluminum foil was pasted onto the surface of the composite so that it was vulnerable to ablation and could acquire a usable signal. Using a fully noncontact system with laser excitation at a fixed point and a scanning laser sensor, the effects of different aluminum foil sizes and shapes on the wavefield were studied for the composites; we decided to use a rectangle with 3 mm length and 5 mm width for laser excitation experiments. Wavefield characteristics of the composite plates were analyzed with single- and multi-layered Teflon inserts. Taking the time window for standard ultrasonic testing as a reference, the algorithms for localized wave energy with appropriate time windows are presented for the detection of single and multiple defects. The appropriate time window is meaningful for identifying each delamination defect. The algorithm performs well in delamination detection of the composites with one or multiple Teflon inserts.

Multi-frequency localized wave energy for delamination identification using laser ultrasonic guided wave.

Delamination is the most common and dangerous failure mode for multilayered structures. Delamination defects of different shapes and sizes have different sensitivity to guided wave of different frequencies and modes. So that it is necessary to study the application of multi-frequency methods for achieving detection. In this study, the algorithm of multi-frequency localized wave energy is present using laser ultrasonic guided waves for delamination identification. Localized wave energy is acoustic energy in space under specific wavenumber. New wavenumber components occur in damaged composite plates and its localized wave energy can be used for delamination identification. The localized wave energy is not only related to mode conversion caused by the decrease of structural thickness above the delamination, but also the scattering waves in delamination region. The scattering waves make acoustic energy redistributed and it is enhanced at specific spatial position. The discovery has been verified in simulation and experiment. Multi-frequency experimental results show lower noises and more discernible profile of delamination region in two specimens, including medial and non-medial delamination. In the case of medial delamination, the actual dispersion curve is closer to the dispersion curve of upper laminate at high frequency; in the case of non-medial delamination, the actual dispersion curve is similar to the ideal situation ignoring the effect of epoxy resin. Based on the actual dispersion curves, two critical parameters of proper frequencies and filter threshold are selected for delamination identification using laser ultrasonic guided wave.

Cure monitoring and damage identification of CFRP using embedded piezoelectric sensors network.

Because of the advantages of high specific strength and high specific stiffness, carbon fiber reinforced plastics (CFRP) have been the most ideal materials in the field of civil aviation. Cure monitoring in manufacturing process and damage identification in service stage of CFRP are always hot topics. The Semi-Analytical Finite Element (SAFE) method and micromechanical model are employed to analyse the propagation characteristics of the Lamb-like waves in a continuous flat aluminium plate attached to a viscoelastic unidirectional CFRP in semi-infinite half-space. Then the vacuum bag moulding process of prepregs is monitored using Fiber Bragg Grating (FBG) and piezoelectric sensors encapsulated in Stanford Multiactuator-Receiver Transduction (SMART) Layer. The calculated energy velocities and attenuations of guided waves show the same trends with the numerical results while curing. After the CFRP is demoulded, the damage identification experiments are carried out. By continuing to use the sensor network embedded in the manufacturing phase, the artificial damages can be precisely located. The results demonstrated that the life-cycle monitoring of CFRP can be achieved effectively by the piezoelectric sensors network.

Embedded FBG Sensor Based Impact Identification of CFRP Using Ensemble Learning.

Impact brings great threat to the composite structures that are extensively used in an aircraft. Therefore, it is necessary to develop an accurate and reliable impact monitoring method. In this paper, fiber Bragg grating (FBG) sensors are embedded in unidirectional carbon fiber reinforced plastics (CFRPs) during the manufacturing process to monitor the strain that is related to the elastic modulus and the state of resin. After that, an advanced impact identification model is proposed. Support vector regression (SVR) and a back propagation (BP) neural network are combined appropriately in this stacking-based ensemble learning model. Then, the model is trained and tested through hundreds of impacts, and the corresponding strain responses are recorded by the embedded FBG sensors. Finally, the performances of different models are compared, and the influence of the time of arrival (ToA) on the neural network is also explored. The results show that compared with a single neural network, ensemble learning has a better capability in impact identification.

Quantitative defect inspection in the curved composite structure using the modified probabilistic tomography algorithm and fusion of damage index.

The curved composite structures are popularly used in the aerospace field for their superior properties. Complexity of structure and geometry generally limit the inspection or monitoring effect of different types of defects in the curved composite structure. A feasible damage probabilistic tomography algorithm combined with ultrasonic guided wave technology is necessary to be developed for the structural health monitoring of curved composite structures. In this paper, defect zones in a curved composite structure are characterized using the modified probabilistic tomography (MPT) method and fusion of damage index (DI). The MPT with the defect shape factor ßM at each damage-impaired path and hybrid DI schemes are proposed to indicate the location and propagation of defect zones in tested sample. The feasibility of proposed approaches is verified on the curved carbon/epoxy composite structure, experimentally. The results show that the MPT and fusion DI methods successfully represent the extension of defect zones in a quantitative manner. It is suggested that the accuracy and reliability of localization results of the MPT algorithm is better than those obtained by the probabilistic tomography (PT) algorithm with the averaged ß.

An Eddy Current-Based Structural Health Monitoring Technique for Tracking Bolt Cracking.

Bolted joints are the primary structures for the load transfer of large-scale structures. It is vital to monitor the process of bolt cracking for enduring structural safety. In this paper, a structural health monitoring technique based on the embedding eddy current sensing film has been proposed to quantify the crack parameters of bolt cracking. Two configurations of the sensing film containing one-dimensional circumferential coil array and two-dimensional coil array are designed and verified to have the ability to identify three crack parameters: the crack angle, the crack depth, and the crack location in the axial direction of the bolt. The finite element method has been employed not only to verify the capacity of the sensing film, but also to investigate the interaction between the crack and the eddy current/magnetic field. It has been demonstrated that as the crack propagates, the variations of the induced voltage of the sensing coils are influenced by both eddy current effect and magnetic flux leakage, which play different roles in the different periods of the crack propagation. Experiments have been performed to verify the effectiveness and feasibility of the sensing film to quantify three crack parameters in the process of the bolt cracking.

Electromechanical Impedance Model for Free 1D Thin-Walled Piezoelectric Ceramics with a Novel Derivation.

The electromechanical impedance model of the piezoelectric ceramics in a free state can be used for screening and quality control in the structural health monitoring community, but the derivation process of the existing model is usually complicated. This paper describes a novel theoretical derivation methodology based on the assumption of zero-stress on the free boundary of the one-dimensional transducer, which can simplify the derivation of the model to a large extent. To assess the accuracy of the model, a signal processing method based on frequency shifting transformation and the Pearson correlation coefficient is also proposed to calculate the similarity between theoretically predicted and experimentally measured data. Two different piezoelectric ceramics were used in experiments to verify the effectiveness of the model. Experimental results convincingly demonstrate that the assumption proposed in this paper possesses good feasibility for one-dimensional thin-walled piezoelectric ceramics and the model has excellent precision.

Hidden corrosion detection using laser ultrasonic guided waves with multi-frequency local wavenumber estimation.

Hidden corrosion damage of aircraft structures is a potential threat to the flight safety and a major maintenance factor for aircraft. In previous researches on corrosion detection, the local wavenumber estimation is obtained from a sliding wavenumber window that maximizes the wave energy at each grid point for one mode and frequency. This paper presents the methodology based on multi-frequency local wavenumber estimation for quantitative assessment of hidden corrosion in plates, which is obtained from a short-windowed Fourier transform in space domain of the wavefield at single mode and multiple frequencies. The methodology is applied to simulated wavefield to demonstrate the effect of frequency on corrosion depth image. The simulation results show that the calculated shape of the corrosion is closest to the actual value when the window size is comparable or less than half the length of the corrosion that the guided wave passes through. A fully noncontact experimental platform is developed and used to detect hidden corrosion with the methodology in aluminum plates. The methodology was verified on the aluminum plates of 1 mm thickness with the corrosion (30 mm × 30 mm × 0.5 mm and Φ30mm × 0.2 mm), and the relative errors between the estimated and actual value of the corrosion depth are not more than 6%.

The Design and Verification of an Active SAMSR Ultrasonic Guided Wave Monitoring System with Ultra-Low Crosstalk.

Due to long propagation distance and high sensitivity to a variety of damages, ultrasonic guided wave technologies have been widely applied in the damage detection or health monitoring of pipe networks and large plate-like structures. However, there are two important problems to be solved when applying this technology; namely, the large scanning time required for monitoring large-scaled structures and the serious crosstalk between the actuation and receiving signals, especially when monitoring hot-spot regions. Therefore, this study mainly designed key parts, such as the matrix switcher and attenuation circuit. The single-actuation and multiple-simultaneous-reception (SAMSR) mechanism based on an analog switching matrix and a low noise charge amplifier circuit was designed and integrated with the SPI control bus to shorten the scanning time. Moreover, a two-stage attenuation circuit with an interlocking isolation structure is presented to effectively isolate the receiving signals from the actuation signals to obtain ultra-low crosstalk even under a high voltage actuation source. In this study, the designed matrix switcher and other components were integrated into the developed ultrasonic guided wave monitoring system. Several experiments were conducted on a stiffened composite structure to illustrate the effectivity of the developed SAMSR ultrasonic guided wave system by comparing the signals collected with those from a commercial ultrasonic guided wave system.

Propagation characteristics of ultrasonic weld-guided waves in Friction stir welding joint of same material.

Friction stir welding (FSW) is an important technology for manufacturing large-scale aluminum alloy propellant tank. Due to stress corrosion and cyclic loads, the key FSW joints require online monitoring to ensure the structural integrity and service safety of long-term storage propellant tanks. FSW joints in the propellant tank are regarded as a type of circumferential or longitudinal long and narrow region. In order to detect the flaws with high efficiency and fewer sensors, the propagation characteristics of ultrasonic guided waves in the FSW joint of same material is investigated in this paper. The weld of a FSW joint is characterized by concave cross-sectional shape and different microstructure-mechanical parameters. The micro-structure, micro-hardness, and Young's modulus of a real FSW joint are analyzed, and a two-dimensional semi-analytical finite-element (SAFE) method is employed to study the effects of different parameters on the modal characteristics of weld-guided waves in the FSW joint. In the studied fundamental modes (symmetric (S0), anti-symmetric (A0), and shear-horizontal (SH0)), an almost non-leaky A0-like weld-guided wave at a certain frequency range from 100 kHz to 210 kHz is discovered in the welded zone of a specific FSW model and shows a potential for long-distance detection. Parametric simulation results show that A0-like, SH0-like and S0-like modes at 120 kHz always exist when the weld width is changed while the moduli of the welded zone and base metal zone are maintained the same. Additionally, the simulations demonstrate that some weld-guided waves only exist if the modulus value of the welded zone is lower than that of the base metal zone when the cross section is geometrically continuous (i.e. the shoulder plunge depth is zero). Comparing with weld-guided waves affected by weld width, the weld-guided waves affected by the modulus change shows less obvious energy leakage during propagation. The experiments are conducted to validate the existence of A0-like weld-guided mode with a primary energy trapping effect.

Assessment of low-velocity impact damage in composites by the measure of second-harmonic guided waves with the phase-reversal approach.

Structural strength and integrity of composites can be considerably affected by the low-velocity impact damage due to the unique characteristics of composites, such as layering bonded by adhesive and the weakness to impact. For such damage, there is an urgent need to develop advanced nondestructive testing approaches. Despite the fact that the second harmonics could provide information sensitive to the structural health condition, the diminutive amplitude of the measured second-order harmonic guided wave still limits the applications of the second-harmonic generation-based nonlinear guided wave approach. Herein, laminated composites suffered from low-velocity impact are characterized by use of nonlinear guided waves. An enhancement in the signal-to-noise ratio for the measure of second harmonics is achieved by a phase-reversal method. Results obtained indicate a monotonic correlation between the impact-induced damage in composites and the relative acoustic nonlinear indicator of guided waves. The experimental finding in this study shows that the measure of second-order harmonic guided waves with a phase-reversal method can be a promising indicator to impact damage rendering in an improved and reliable manner.