This paper presents an ultrasonic detection method tailored for evaluating impact damage in carbon fiber reinforced polymer (CFRP) plates, which fundamentally relies on the interaction between a bounded ultrasonic beam and a CFRP plate immersed in fluid. Initially, the non-specular reflection sound field expression for a bounded ultrasonic beam obliquely incident on the CFRP plate is derived. Subsequently, finite element analysis reveals the correlation between the signals detected by the specular receiver and the impact damage status under different parameter combinations. It is found that under the optimal parameter combination and the A0 critical angle, even a slight degradation in the elastic constants of the CFRP plate induces a monotonic and significant increase in the change rate of the specular reflection coefficient (SRC). In the experimental phase, tests conducted on the water-CFRP plate-water structure validate the aforementioned finite element simulation results. The experimental data consistently demonstrate that at the A0 critical angle, the change rate of SRC increases monotonically with increasing impact energy—confirming the effectiveness of the bounded ultrasonic beam-based non-specular reflection method at the A0 critical angle for assessing impact damage in CFRP plates.

Anisotropic materials are of significant interest for major structural applications, where fast and accurate acoustic source localization (ASL) of damages is crucial for structural assessment and maintenance. Existing ASL methods for highly anisotropic plates often require assumptions of straight‑line wave propagation, solution of nonlinear equations, prior knowledge of wave velocity, or acoustic event duration. To address these limitations, this paper proposed an ASL method using three L-shaped sensor clusters (TLSSC), combining the advantages of linear and nonlinear Lamb waves with the LSSC, for highly anisotropic plates having rhombus-shaped wavefronts. Feasibility of this approach was verified through localization experiments on a [0/90]2s composite plate using linear ultrasonics for passive acoustic emission detection at various acoustic source locations. Additionally, Lamb wave propagation in an anisotropic silicon wafer was simulated for active microcrack detection via nonlinear ultrasonics, considering different material orientations, microcrack locations, and cluster arrangements. Both experimental and simulated results confirm the effectiveness of the proposed method: compared to the LSSC-based ASL, localization accuracy is significantly improved, with absolute errors below 6 mm for a plate of dimension 500 mm × 500 mm and a wafer of 8 in. considered in this study and relative error under 5 % in nearly all cases. This method is not only effective for acoustic emission detection, but can also be extended to active detection of microcrack through appropriate mode pair selection and the pulse inversion technique.

Closed cracks in in-service structural materials generate extremely weak ultrasonic reflection/scattering signals, making them difficult to detect effectively using conventional linear ultrasonic testing methods. This paper develops a nonlinear ultrasonic total focusing method integrated with phase coherence to achieve high-sensitivity detection and precise localization of closed cracks. The approach enhances incident wave energy through parallel excitation and extracts high signal-to-noise ratio (SNR) nonlinear harmonic signals using pulse inversion. On this basis, a total focusing algorithm based on nonlinear responses is constructed, and phase coherence weighting is introduced to significantly suppress noise and clutter, thereby improving the imaging SNR and defect localization accuracy. To validate the effectiveness of the method, systematic experimental studies were conducted on aluminum specimens containing fatigue-induced closed cracks (including two cases of single and double cracks). The results show that the proposed method can clearly image closed crack structures and accurately locate their propagation tips. The study also reveals that closure stress affects the intensity of the nonlinear response of closed cracks, and the ultrasonic energy at the crack is a key factor determining the localization accuracy of crack tips. By integrating nonlinear ultrasound with phase-coherent total focusing imaging, this method effectively achieves high-sensitivity, visual detection and quantitative evaluation of small closed cracks inside materials.

Guided ultrasonic waves are widely used to characterize layered elastic structures; however, accurately predicting wave dispersion in multilayers composed of coupled solid, fluid, and porous media remains challenging. These difficulties arise from strong material heterogeneity and the complex physics governing multiphase interfaces. This work introduces a unified semi-analytical finite element framework combined with perfectly matched layers for modelling guided wave dispersion in solid-fluid-porous layered waveguides. A key novelty lies in a porosity-weighted interfacial formulation that rigorously enforces traction continuity, solid displacement compatibility, pore pressure equilibrium, and fluid flux conservation, thereby enabling physically consistent coupling across multiphase interfaces. In addition, a unified mathematical formulation is shown to handle both finite and infinite porous domains without reformulation, allowing seamless treatment of bounded porous layers and unbounded fluid-saturated media. By reducing the full three-dimensional waveguide problem to a two-dimensional cross-sectional eigenvalue formulation, the proposed approach achieves high computational efficiency while retaining geometric generality. The framework is validated through comparisons with analytical solutions and experimental measurements on representative porous structures, including battery anodes and saturated sintered materials. The results demonstrate accurate prediction of dispersion characteristics and reveal complex wave-structure interactions in multilayered porous systems. This unified framework provides a versatile and reliable tool for ultrasonic characterization and the analysis of heterogeneous material systems.

Signal modulation of higher harmonics can reduce the complexity of signal processing in nonlinear ultrasonic testing (NUT), which are effectively achieved by utilizing the bandgap characteristics of phononic crystals. However, proposing effective configuration of phononic crystals for NUT could be a significant challenge based on the complex conventional design process. Therefore, machine learning technique provides a feasible approach for inverse design of phononic crystal structures. This paper establishes an inverse design framework for phononic crystal configurations based on multi-branch convolutional neural networks (CNN) and generative adversarial networks (GAN). Then, the phononic crystal filter (PCF) with a truncated conical configuration is proposed. Meanwhile, numerical simulations are employed to verify the filtering ability of the designed filter, and the influences of the spatial arrangement and geometric parameters of PCF's unit-cell are also numerically investigated. Finally, the 3D-printed truncated conical PCF is manufactured to experimentally validate the efficient filtering ability for higher harmonics modulation. This study provides numerical and experimental basis for the application of machine learning-based inverse design of PCF in non-destructive testing.

Structural scattering, a key acoustic feature of shear wave angle beam transducers, generates echoes when elastic waves propagate and reflect within the transducer's wedge structure. This study adopted structural scattering to analyze the transducer's structural parameters and established a numerical model via COMSOL to characterize its dynamic acoustic field. Primary structural scattering is defined as Longitudinal wave Structure Scattering I (LSSI), while secondary scattering covers L-wave Structure Scattering II (LSSII) and S-wave Structure Scattering I (SSSI). The effects of critical parameters including wedge back angle, front angle, front wedge surface shape, and matching backing layer on structural scattering were systematically explored. Results reveal the primary scattering echo amplitude peaks at the wedge back angle of the first critical angle, with two weak-scattering intervals (29 degrees-34 degrees and 38 degrees-41 degrees) at larger angles. The maximum primary scattering amplitude occurs at a 50 degrees front angle, followed by two weak-scattering ranges (60 degrees-80 degrees and 100 degrees-115 degrees). Optimal low scattering is achieved with 3/4 lambda-deep sawtooth grooves and uniform backing layers. Prototyped transducer reduces structural scattering by over 90% versus commercial counterparts, delivering a 9.5 dB signal-to-noise ratio for Phi 1 weld defects. This work provides a foundation for optimized shear wave angle beam transducer design.

Debonding defects in the bonding layers of stiffened composite panels pose significant safety risks. However, they are challenging to detect due to their deep location and complex geometry, where conventional linear ultrasonic methods often fail. This study presents a broadband nonlinear ultrasonic guided wave approach for accurate debonding imaging. Chirp guided wave mixing (CGWM) is employed to excite broadband signals and effectively eliminate time-consuming signal processing. Defects are detected by analyzing three types of nonlinear components: second-harmonic generation (SHG), sum-frequency harmonic (SFH), and difference-frequency harmonic (DFH) responses. Results demonstrate that while linear ultrasonic detection struggles with deep debonding, all nonlinear ultrasonic components enable successful defect identification and imaging. Crucially, SFH and DFH allow efficient frequency-domain filtering, significantly reducing computational time compared to time-frequency domain processing required for SHG. Furthermore, the CGWM technique enhances nonlinear source behavior at defects while minimizing spurious nonlinearities from instruments.

Nonlinear ultrasonic techniques offer a viable approach for stress assessment. Nevertheless, the practical adoption of the second harmonic generation method is constrained by two principal factors: pronounced signal attenuation and spurious harmonics generated by the measuring apparatus. To overcome the limitations, the study interrogates the generation mechanism of the quasi-static component (QSC), which arises from the interaction of longitudinal critically refracted (LCR) waves with stress, for its application in nondestructive stress evaluation. We systematically assess the feasibility of using the QSC to quantify tensile stress in aluminum specimens through integrated numerical simulations and experiments. The results demonstrate that a normalized acoustic nonlinearity parameter (ANP) derived from the QSC increases monotonically with applied tensile stress from 0.0 to 120.0 MPa. The QSC-based ANP exhibits a strong linear correlation with stress, and validation on pre-stressed specimens yields prediction errors below 5.6%. The findings of this study establish the QSC-based ANP method as an effective and reliable factor for measuring elastic tensile stress. Compared to other nonlinear ultrasonic techniques, the proposed approach offers a highly sensitive method for stress evaluation in load-critical aluminum structures.

Welded-plate structures containing welded joints and parent plates are widely used in the engineering field. This paper presents a new method for assessing damage in welded joints by leveraging the reflection effect of a bounded ultrasonic guided wave (UGW) beam propagating in the parent plate. The reflection characteristics of the bounded UGW beam when it travels from the parent plate to the welded joint were initially analyzed theoretically. Subsequently, finite element (FE) simulations were conducted to assess the feasibility of using the reflection characteristics of the bounded UGW beam for material damage evaluation in the welded joint. The influence of different factors on the reflected field, including the incident angle and the extent of material damage (quantified by the reduced Young's modulus), was examined. It was observed that the defined damage index exhibited a significant correlation with the degree of material damage. To address practical engineering applications, a simulation of surface damage assessment in welded joints was conducted using a bounded UGW beam. The damage index showed a notable increase as the Young's modulus scaling factor decreased at a specific incident angle. Finally, the experimental results demonstrate that the polyvinylidene fluoride (PVDF) comb transducer can effectively excite and capture the bounded UGW beam in the parent plate, enabling the effective evaluation of surface corrosion damage in the welded joint. The significant increasing trend of the maximum amplitude and damage index with the increase of the corrosion level proves that the reflection effect of a bounded UGW beam incident from the parent plate is more sensitive and reliable for evaluating surface corrosion damage in the welded joint. This study provides a theoretical basis, FE simulation, and experimental validation of a bounded UGW beam using PVDF comb transducers for damage evaluation in welded joints, showcasing its tremendous potential for practical engineering applications.

Electron transport layer (ETL) plays a pivotal role in determining the interfacial integrity and operational robustness of n-i-p structured perovskite solar cells (PSCs). Conventional tin oxide based inorganic ETLs are often plagued by inherent point defects, while organic small-molecule ones frequently suffer from limited device efficiency and durability. In this study, we present an innovative molecular design strategy via developing thermo-crosslinking organic ETLs to overcome these persistent interfacial challenges. Novel organic electron transport materials (ETMs) have been successfully designed by strategically incorporating heat-inducible cross-linking triallyl or oxetane functional groups into naphthalene diimide-based conjugation scaffold, respectively. Such cross-linkable ETMs exhibit exceptional electronic properties, facile heat-induced film-forming capability, and enhanced charge transport. Specifically, featuring optimized energy level alignment and superior surface wettability, oxetane-functionalized ETL endowed n-i-p structured PSCs with a champion power conversion efficiency of 25.23%, among the highest values reported for organic ETL-based devices. Non-destructive ultrasonic testing and accelerated aging assessments have been explored for the first time to decode the substantial improvements in interfacial robustness and operational stability under thermal (85°C) and humid conditions (65% relative humidity). This work establishes a versatile material design paradigm for developing robust organic ETLs, paving the way for high-performance and durable perovskite photovoltaics.

Nonlinear guided waves (NGWs), owing to their exceptional sensitivity to microstructural variations in materials, have emerged as a powerful approach for detecting early-stage damage in structural health monitoring (SHM). However, such microstructural changes are typically very subtle, and the resulting nonlinear wave components are easily masked or distorted by other non-damage-related nonlinear sources within the system. Moreover, since the nonlinear energy is much weaker than that of the fundamental wave, existing approaches generally rely on frequency domain analysis to extract nonlinear components. This makes signal processing highly complex and imposes stringent requirements on the signal-to-noise ratio (SNR), thereby limiting the feasibility of real-time and reliable monitoring. To address these challenges, we propose a guided-wave purification strategy based on metamaterial filters (MFs). Bandgap-type MFs are attached to both the excitation and reception ends of the structure under inspection, enabling purification at both ends. These filters operate through local resonances induced by the coupling between patches and the host structure, thereby generating bandgaps within the target frequency range and effectively blocking the propagation of specific frequency components. The excitation-side filter suppresses parasitic second harmonic components generated by the excitation system, thereby eliminating unwanted nonlinear interference. Meanwhile, the reception-side filter enables selective extraction of damage-induced second harmonic components, allowing real-time identification of the structural health condition through variations in the second harmonic time domain signal. Finite element simulations and laboratory experiments conducted on aluminum plates demonstrate that the proposed bidirectional purification strategy can effectively suppress nonlinear interference and extract damage-related nonlinear responses, thereby simplifying signal processing procedures and exhibiting promising potential for practical engineering applications.

Coated metal thin-walled plates are extensively applied in aerospace and petrochemical industries for their excellent resistance to oil and chemical corrosion. To ensure structural integrity and long-term safe service, developing reliable ultrasonic nondestructive testing techniques is essential. However, ultrasonic techniques applied to coated plates present considerable challenges. On the one hand, high-frequency ultrasound-though featuring high defect resolution-undergoes rapid attenuation when propagating in highly attenuative coated materials. On the other hand, while low-frequency ultrasound exhibits low attenuation, it suffers from poor directivity and beam divergence, leading to insufficient spatial resolution and inaccurate defect localization. To address these issues, this study proposes a probabilistic damage identification method based on zero-frequency waves (ZFW) generated by primary high-frequency Lamb wave propagation, aiming to achieve high-resolution defect imaging in highly attenuative coated plates. Firstly, theoretical analysis and numerical simulation were conducted to investigate the propagation characteristics of high-frequency Lamb waves and the resultant ZFW in free plates and coated plates, followed by experimental validation. Furthermore, ZFW induced by high-frequency Lamb wave propagation were employed for probabilistic damage identification of highly attenuative coated plates, with the results compared against damage reconstruction outcomes from linear Lamb waves. Results demonstrate that high-frequency Lamb waves decay rapidly in highly attenuative coated plates, and their imaging results fail to capture defect location information. In contrast, the probabilistic reconstruction algorithm based on ZFW of high-frequency Lamb waves can accurately identify defects at various positions, and provides higher resolution and accuracy than low-frequency linear Lamb waves.

Multiple signal classification (MUSIC) algorithm provides a high-resolution solution for guided-wave-based damage localization, yet its application to composite structures is limited by two issues: the dependence on pristine baseline signals for extracting damage-scattered waves and the requirement for narrowband inputs with properly selected frequencies. This study proposes a baseline-free impact damage localization method that converts broadband guided-wave responses into fixed-frequency MUSIC inputs. Two tailored chirp guided waves are actuated into carbon fiber reinforced polymer (CFRP) composites containing barely visible impact damage. Owing to contact acoustic nonlinearity at the damaged interfaces, the interaction of the two broadband waves generates nonlinear acoustic responses. By using oppositely swept chirp waves, the sum-frequency harmonic remains at a prescribed fixed frequency. By using chirp waves with the same sweeping direction and a constant frequency offset, the difference-frequency harmonic is also converted into a fixed-frequency component. These fixed-frequency nonlinear responses can be isolated using simple frequency-domain filtering and then used as narrowband inputs for the near-field MUSIC algorithm. Finite element simulations and experiments on CFRP laminates are conducted to validate the proposed method. The results show that the impact damage can be localized without using baseline signals or complicated time-frequency decomposition. The proposed approach provides an efficient way to transform broadband guided-wave information into MUSIC-compatible narrowband inputs for baseline-free localization of impact damage in composite structures.

Weld regions in regenerative cooling thrust chambers are highly susceptible to cracking under thermo-mechanical cyclic loading, posing a risk of catastrophic rocket engine failure. While the inspection of these thin-walled, circumferentially periodic structures is critical, conventional non-destructive evaluation (NDE) methods are ill-suited for rapid, remote screening. This study introduces an axial-guided-wave-based technique for rapid detection and localisation of weld defects in complex cylindrical thin-walled structures. The core of this approach is a novel, unit-cell-based modelling strategy developed within the Semi-Analytical Finite Element (SAFE) framework, which incorporates pointwise constraints to accurately capture the modal behaviours and wave structures of existing guided wave modes in circumferentially periodic waveguides of arbitrary cross-section. From this model, we establish a systematic, mode-selection workflow for selecting guided wave modes with enhanced detection sensitivity to localised weld defects. This workflow, encompassing dispersion analysis, energy localisation, mode-shape evaluation, and excitability assessment, provides a physics-based alternative to traditional, ad-hoc mode selection. Through comprehensive three-dimensional finite element (FE) simulations and experimental validation, a specific longitudinal type L(0,1)-like guided wave mode is identified as optimal, exhibiting concentrated wave energy at the vicinity of the weld seam, minimal dispersion, and high excitability. Furthermore, we provide a direct basis for localising defects via time-of-flight and assessing their significance via reflection-coefficient analysis. The proposed methodology establishes a versatile new framework for applying axial guided waves, paving the way for quantitative integrity monitoring of complex, circumferentially periodic engineering components.

Non-destructive evaluation of grain size is crucial for characterizing the mechanical properties of polycrystalline materials. We present a novel ultrasonic methodology for evaluating mean grain size based on the interference effect of bounded ultrasonic beams at the Rayleigh critical angle. We analyzed and overviewed the reflected field of a bounded ultrasonic beam at a liquid-solid interface and the relationship between single-crystal and polycrystal elastic constants. Numerical simulations and experimental measurements were conducted to validate the effectiveness of the proposed method. Results show that when a bounded ultrasonic beam is obliquely incident on the liquid-solid interface at the Rayleigh critical angle, the acoustic pressure amplitude captured by the specular receiver exhibits a monotonic relationship with, and high sensitivity to, different mean grain sizes of polycrystalline materials. In contrast, only minor variations are observed at either the longitudinal or transverse wave critical angle. The smaller the deviation of the incident angle from the Rayleigh critical angle, the higher the sensitivity of the proposed method to variations in grain size. Furthermore, the results indicate that by optimizing the bandwidth of the excitation signal while keeping other parameters constant, maximum sensitivity to variations in grain size can be achieved. These findings demonstrate that the proposed method achieves high sensitivity in evaluating the mean grain size of polycrystalline materials, further enriching the ultrasonic measurement techniques for the mean grain size of polycrystalline materials.

A bounded ultrasonic beam scattering method is proposed for early damage detection of trigonometric-function-defined curvature-gradient thin-walled structures to meet industrial ultrasonic inspection requirements. The scattering behavior of bounded ultrasonic beams on complex curved surfaces is first approximated as the superposition of reflections from a series of variable-thickness plates, and the existence of a critical scattering condition is identified. Both finite element simulations and experimental results demonstrate that the proposed ultrasonic critical scattering method enables effective detection of early-stage material damage. Simulations verify its effectiveness in identifying Young's modulus degradation, microcracks, and surface roughness; experimental tests further confirm its capacity for early corrosion detection.

The nonlinear ultrasonic technique is an effective method for characterizing early-stage damage, such as initial corrosion, plastic deformation, and creep. However, the nonlinear response signal induced by micro-damage is typically one or two orders of magnitude weaker than the fundamental wave. This low signal-to-noise ratio is one of the main factors limiting the broader application of nonlinear ultrasonic technology. To address this issue, this study systematically investigates the quasi-static component pulse (QSCP) generated during the propagation of the longitudinal critically refracted (LCR) wave. The feasibility of assessing early corrosion damage in 7075 aluminum alloy was evaluated using the QSCP, a signal feature resulting from the interaction between corrosion-induced microcrack and LCR wave. Both numerical simulation and experimental results demonstrate a positive correlation between the QSCP-based acoustic nonlinearity parameter (ANP) and the extent of corrosion-induced microcrack, which is attributed to the increase in microcrack. Simulation results show that the ANP increases monotonically with the number of microcrack. This trend is experimentally validated, with the ANP showing a significant increase of approximately 54.8% by the fourth stage of corrosion compared to the baseline. These findings confirm the effectiveness and feasibility of the QSCP-based ANP method for detecting early-stage corrosion damage, offering a promising nondestructive approach with high sensitivity for assessing incipient corrosion in critical metallic structures.

Ultrasonic guided waves combined with the reconstruction algorithm for probabilistic inspection of damage (RAPID) can be used to produce clear visualizations of damage regions. However, conventional RAPID methods generally rely on comparing signals against undamaged baseline and exhibit limited sensitivity to capture the subtle nonlinear acoustic responses. This paper introduces a baseline-free framework via internal path self-referencing for imaging localized thermal damage in metal plates. We integrate a virtual time reversal (VTR) procedure with a nonlinear ultrasonic technique, which effectively enhances the visibility of subtle nonlinear responses induced by damage. This method constructs the damage image using the VTR-enhanced amplitudes of second harmonics generated along propagation paths. To eliminate the influence of the accumulation effect of second harmonic generation, a path-length stratification strategy is adopted to define the damage index. We validate this framework through finite element simulations and flame-damaged experiments on an aluminum plate. Compared with conventional RAPID, the proposed method reduces imaging artifacts, removes reliance on undamaged baselines, and improves sensitivity to microstructural changes. This approach shows strong potential for in-service inspection of thermally loaded components.

Microcrack induced acoustic nonlinear responses are generally weak, which compromises the reliability of nonlinear guided wave-based nondestructive testing (NDT) for microcrack. Therefore, physically enhancing such nonlinear responses is critical for reliable microcrack detection. Recently, metasurface-assisted damage detection has emerged as a promising approach to overcoming the limitations of conventional NDT methods. However, existing elastic metasurfaces are constrained by destructive designs, narrow bandwidths, and limited angular adaptability, while their potential for amplifying microcrack related nonlinear responses remains largely untapped. Herein, we propose a nondestructive elastic metasurface integrating moderately broadband and wide-field focusing capabilities. The proposed functionality is achieved through a new design strategy exploiting the synergistic interplay between compensated phase and angle-dependent incident phase. Building upon this strategy, a metasurface-assisted nonlinear guided wave framework is developed for selective enhancement of local microcrack-induced nonlinear responses. Comprehensive numerical and experimental results demonstrate efficient zero-order antisymmetric Lamb wave (A0 wave) focusing over 100-150 kHz and across a wide incident field ranging from −20° to 20°. Moreover, the proposed metasurface enables spatially selective enhancement of wave–microcrack interactions over a moderately broad frequency range. This enhancement results in up to a 235% increase in nonlinear second harmonic amplitude, thereby substantially improving microcrack detection sensitivity. This work provides a new design strategy for moderately broadband and wide-field focusing metasurface. It also paves the way for integrating elastic metasurface with nonlinear ultrasonic techniques, offering significant potential for elastic wave manipulation and structural health monitoring.

Advances in modern manufacturing have expanded the application of components with curvature surfaces in engineering, thereby increasing the demand for ultrasonic techniques capable of detecting near-surface damage in curved structures. This paper proposes a damage detection method based on the critical scattering fields generated by the interaction between ultrasonic waves and curvature surfaces. In this paper, we analyzed stainless steel structure comprised by two quarter-cylindrical segments and joined adhesively. Under fluid-solid coupling conditions, an obliquely incident acoustic wave interacts with a curved structure, splitting into a directly reflected wave and leaky Rayleigh waves. Our analysis shows that under specific conditions, re-radiated waves induced by leaky Rayleigh waves interfere with the direct reflected wave, forming a critical constructive interference field. We specifically investigate the origins of constructive interference angles in the studied structure and demonstrate that early-stage surface damage can be effectively identified using precisely configured transducer pairs under these constructive interference angle conditions. Conducted at three inspection points on specimen, Finite element simulations confirm that the proposed method can detect damage within a near-surface region extending up to one wavelength of the circumferential Rayleigh waves. Experimental results further validate the method's ability to reliably identify corrosion damage.