Enhancing the electromechanical properties of piezocomposite materials is crucial for the advancement of ultrasound transducers and medical imaging technology. To address the limited electromechanical coupling coefficient of traditional 2-2 piezocomposites utilized in phased array ultrasonic transducers, this work proposes a new 2-3 piezocomposite structure based on Pb(Zr1-xTix)O3 (PZT) ceramics. Both simulated and experimental results indicate that the designed 2-3 piezocomposite exhibits superior electromechanical coupling coefficient, with an electromechanical coupling coefficient (keff) of 0.643, exceeding the 2-2 piezocomposite (0.597) and matching the performance of 1-3 piezocomposites. This enhancement is attributed to the introduction of the longitudinal k33 vibration modes. Furthermore, a 3 MHz array transducer fabricated with the 2-3 piezocomposite achieves a -6 dB bandwidth of 105.4 % and an insertion loss of -21.57 dB. Those results not only outperform PZT-ceramic-based 2-2 array transducers but are also comparable to those of a single-crystal-based 2-2 counterpart. Additionally, this structure shows potential for extension to other piezoelectric ceramic and single-crystal systems. This design strategy offers a promising avenue for developing high-performance piezocomposites and high-resolution ultrasound transducers.

Cleaning is important in the beverage and food industries, where it is conducted in validated, predefined cycles using automated methods known as cleaning-in-place (CIP). Existing sensors monitor cleaning indirectly by measuring operating conditions or detecting deposits suspended in the cleaning fluid. Ultrasound (US) provides a direct approach by detecting deposits that remain on the surface during cleaning. While ultrasound has been used effectively for monitoring fouling beneath planar structures, its application in closed geometries such as pipes is challenging due to complex wave propagation. In this study, a clamp-on US system was developed to investigate the effect of organic deposit removal on propagating ultrasound waves in pipes. The US measurements during pipe cleaning showed that deposits reduced signal amplitude and delayed waveform arrival. Features including energy, time shift, and signal correlation were extracted from the US signals. Deposit removal was quantified by conducting cleaning trials at varying durations and subsequently determining the gravimetric mass of remaining deposits. The results showed that the features were significantly correlated with the amount of deposits left in the pipe. Building on this, an online detection algorithm was implemented to track the slope of the energy feature in real-time. In independent trials, the algorithm consistently monitored cleaning until the onset of a plateau in the energy feature, corresponding to an average residual deposit amount of 0.8 ± 0.5 mg/cm[2]. These findings demonstrate that ultrasound enables direct detection of deposits on pipe surfaces during cleaning and estimation of the near end point of cleaning.

Carbon fiber reinforced polymers (CFRP) exhibit exceptional properties such as low density, high strength, corrosion resistance, and design flexibility, making them widely employed in aerospace, automotive, and biomedical applications. However, various types of damage may occur during manufacturing, assembly, or service loading, potentially compromising structural integrity. Therefore, the development of efficient nondestructive testing (NDT) techniques is essential to ensure the reliability of CFRP-based systems. Ultrasonic NDT (UNDT) offers long-range detection, high sensitivity, and cost efficiency for damage evaluation, yet conventional UNDT methods rely heavily on expert interpretation and manual analysis, which limits their accuracy and efficiency. With the rapid progress of deep learning (DL), deep learning-based ultrasonic NDT (DL-UNDT) methods have emerged as a promising alternative, capable of automatically learning complex signal-damage relationships from ultrasound data. This paper provides a comprehensive review of recent advances in DL-UNDT for CFRP structures, summarizes key developments across damage detection, localization, and imaging tasks, and discusses prospective research directions for addressing current challenges.

Elastic parameter inversion methods based on Lamb waves have been widely used to characterize the mechanical properties of plate-like composite structures. However, similar Lamb wave responses from multiple elastic parameters in transversely isotropic composites lead to an ill-posed inverse problem with reduced parameter distinguishability. This study proposes a stepwise inversion framework that combines the S1 Zero-Group-Velocity (ZGV) resonance frequency with multi-directional Lamb wave dispersion data for identifying the stiffness coefficients of transversely isotropic plates. A spectral-collocation-based forward model is applied to calculate Lamb wave dispersion curves and ZGV resonance frequencies. The sensitivities of the stiffness coefficients are evaluated with respect to propagation direction, mode type, and frequency-thickness product, thereby guiding the assignment of parameter-specific priorities. The framework is validated on unidirectional carbon fiber reinforced polymer (CFRP) plates through numerical simulations and laser-ultrasonic experiments. The reconstructed stiffness coefficients show good agreement with the reference values and strain-gauge measurements, with deviations below 4% in both numerical and experimental validations, demonstrating the reliability of the proposed method. The proposed inversion framework provides a feasible approach for characterizing elastic parameters in transversely isotropic composite plates, particularly through-thickness parameters, and may be extended to parameter reconstruction in multilayer composite structures.

Thermal barrier coating (TBC) has long served in extreme thermal-mechanical coupled environments, rendering it highly susceptible to damage such as substrate cracking. Due to the multilayered, heterogeneous nature of TBC, conventional ultrasonic testing faces critical bottlenecks, including severe acoustic field scattering and difficulty in extracting weak damage features. To address these challenges, this paper proposes an intelligent laser ultrasonic imaging and classification method based on a physics-constrained dual-channel one-dimensional convolutional neural network (DC-1DCNN). First, a high-fidelity three-dimensional (3D) Multiphysics numerical model of TBC is constructed. A dual-driven dataset is generated through both numerical simulations and limited experimental data. Furthermore, a transformation matrix is introduced to establish a feature-interaction mapping mechanism between the virtual numerical space and the physical-entity space, thereby overcoming the bottleneck of poor generalization for damage samples in multi-layered heterogeneous structures. Meanwhile, a constraint matrix is constructed based on surface wave amplitudes, and an innovative threshold sparse self-attention mechanism (TSSAM) is proposed. This mechanism strongly constrains the depth of feature extraction in non-stationary ultrasonic signals, significantly enhancing the network's adaptive capability to capture weak crack features at heterogeneous interfaces. Experimental results demonstrate that the model achieves a test accuracy of 96.72 %, with the sparse self-attention mechanism improving it by 4.22 %. This method achieves an end-to-end mapping from single-point time-domain A-scan signals to high-performance C-scan imaging and enables precise crack classification, providing a new paradigm for the efficient and reliable non-destructive evaluation of multi-layered heterogeneous structures.

Photoacoustic imaging (PAI) enables visualization of optical absorption contrast through ultrasonic detection, and light emitting diode (LED) based PAI systems have attracted attention due to their high repetition rates, portability, and cost-effectiveness. However, these systems typically require high driving currents and rapid switching speeds to activate the LEDs, which generates significant electromagnetic interference (EMI). In this study, we identified that this EMI induces currents within the ultrasound probe circuitry, driving the probe to passively emit ultrasonic waves and resulting in ultrasonic noise. To address this issue, we analyzed the amplitude spectrum of the EMI signals and discovered that modulating the LED pulse width shifts its spectral nulls. A pulse width adjustment algorithm was proposed based on the ultrasound probe bandwidth. By aligning the spectral nulls with the center frequency of the probe, EMI coupling is minimized, thereby reducing ultrasonic noise. Additionally, we propose a residual reconstruction algorithm based on an optical shutter. Since the optical shutter modulates the generation of photoacoustic signals while leaving the ultrasonic noise unaffected, computing the residual between images reconstructed in the open-shutter and closed-shutter states allows for the elimination of ultrasonic noise. Validated on phantoms and in vivo human vessels, pulse width adjustment improved the signal-to-noise ratio by 3.4 dB, while residual reconstruction yielded an enhancement of 13.3 dB. This study provides an in-depth analysis of the noise issues in LED-based PAI systems, which might contribute to enhanced image quality and the development of low-cost photoacoustic systems, thereby facilitating the broader clinical application of LED-based photoacoustic imaging.

Ultrafast plane-wave imaging (PWI) trades image quality for frame rate under sparse acquisitions. Existing deep learning-based PWI reconstruction methods are limited by the lack of physics-guided degradation modeling, leading to distortions in anatomical structures and speckle statistics. In addition, an inherent low-frequency spectral bias attenuates high-frequency components, further exacerbating detail loss and amplifying reconstruction artifacts. In this study, we propose DiffPWI, a degradation-aware reconstruction framework. DiffPWI employs a latent diffusion model (LDM) to learn the distribution of physically informed degradations in a compressed latent space, providing a physics-informed prior for reconstruction. Furthermore, we introduce a frequency decoupling attention module (FDAM) that adaptively decomposes feature representations into hierarchical frequency sub-bands, enabling preservation of high-frequency anatomical details while suppressing acquisition artifacts. Extensive evaluations on phantom, multi-organ in vivo, and zero-shot datasets demonstrate that DiffPWI achieves improved spatial resolution, contrast recovery, and speckle fidelity compared with state-of-the-art methods, advancing high-fidelity ultrafast ultrasound imaging.

The de-icing technology based on the ultrasonic cavitation-coupled acoustic streaming effect is a novel non-contact de-icing method, but existing research is mostly limited to the single-mode planar ultrasound. Based on optical principles, a mathematical model for laser ablation of ice bodies was established, primarily analyzing the critical initial output light intensity of the laser for ice melting and the variation of the melting rate with changes in the initial temperature of the ice body. Using COMSOL Multiphysics, the crack phase field variation patterns of different gradually tapered conical hole structures were analyzed. Finally, the experiment measured and compared the acoustic streaming noise decibel values of the ice surface under both perforated and unperforated conditions, as well as the corresponding de-icing morphological characteristics under these conditions. Experimental results indicate that laser ablation holes form structural tips within the ice body, and the acoustic streaming creates directional flow along the hole cavity. Laser ablation holes enhanced the de-icing effect of ultrasonic cavitation coupled with acoustics streaming, achieving directional hole expansion within the ice body and providing stress concentration conditions for ice fracturing. Laser ablation can guide the path of the acoustic streaming generated by ultrasonic cavitation, making it a potentially valuable method for inducing crack propagation in ice.

Ultrasonic guided wave is recognized as a promising technique for damage detection of thin-walled metal structures. However, such structures are chronically subjected to dynamic environments, where the induced compound perturbations such as temperature variation and structural deformation will distort signal characteristics and reduce detection accuracy. To address this challenge, a multi-source domain dynamic adversarial adaptation network (MS-DDAAN) is proposed to enhance damage detection performance under compound perturbations. Firstly, single-frequency signals are obtained through a series of processes, including broadband chirp excitation, frequency deconvolution, filtering, and normalization. Meanwhile, signals acquired under different operating conditions are categorized into multiple source domains and a target domain. Afterward, a dual-layer feature extraction architecture is designed, integrating dynamic convolution and self-attention mechanism to adapt to signal distribution shifts and extract domain invariant damage features. Additionally, an adversarial adaptation network, which contains global domain discriminator and local domain discriminator, is introduced to achieve distribution alignment and feature alignment across different domains. Finally, a label classifier is employed to map damage features to state labels, and a dynamic harmonic factor is introduced to balance label alignment with domain alignment during the training process. The proposed method is validated on thin-walled metal plate under 12 composite conditions of temperature variation and structural deformation. Experimental results demonstrate that MS-DDAAN achieves an accuracy over 95% in damage identification across all test conditions and its performance is superior to the compared other transfer learning methods.

Pipeline failures in industrial and municipal systems pose significant risks of environmental contamination, economic losses, and public safety hazards. However, it is often challenging to implement conventional pipeline inspection methods under dynamic environmental/operational conditions. Specifically, some of the most popular extended-range inspection methods for pipelines are based on acoustic guided-wave defect detection techniques, but they often exhibit degraded performance due to variable liquid flow rates and temperature fluctuations, resulting in critical gaps in continuous and reliable infrastructure monitoring. To address this challenge, this work presents a novel, deep-learning-based (DL-based) acoustic inspection method with two specialized deep learning architectures: a 1D convolutional autoencoder (CAE) and a 1D convolutional variational autoencoder (CVAE). The two architectures differ in how the latent representation is formed: the CAE learns a deterministic latent features, which favors stable and fine-grained multi-task regression, whereas the CVAE learns a probabilistic latent distribution, which explicitly models the uncertainty introduced by environmental variability and regularizes the latent space. These DL architectures are designed for noise-robust feature extraction and the disentangling of defect characteristics from environmental interference. To evaluate this new method, experiments were conducted on a test pipe in a flow loop with variable water flow rate and temperature. The tests used a 6-foot [183 cm] galvanized steel pipe (2-inch diameter) equipped with a sparse sensor network of three broadband transducers, generating rich experimental datasets. DL-based signal processing performed on these datasets demonstrated robust performance, with over 97 % physically simulated defect detection accuracy under flow rates from 30 to 122 mL/s and temperatures from 15 to 50 °C. The DL models achieved defect localization error under 10 cm and a relative defect quantification error under 10 %. Additionally, temperature and flow rate predictions by the DL methods yielded mean absolute errors below 0.83 °C and 1.56 mL/s, respectively. By integrating environment-invariant feature learning through convolutional layers and probabilistic latent space modeling, the proposed approach enabled simultaneous structural integrity assessment and environmental/operational parameter monitoring. These advancements have the potential to address a critical need for reliable inspection technologies in oil, gas, and energy infrastructure, particularly in dynamic environments where existing methods do not provide satisfactory solutions.

Ensuring structural reliability in aerospace and advanced manufacturing demands non-destructive evaluation (NDE) techniques that are both accurate and adaptable. We present a flexible framework based on geometric phase sensing, which captures topological shifts in acoustic fields to characterize internal defects. By integrating broadband excitation with remote sensing methods such as laser Doppler vibrometry, we demonstrate defect localization and tracking using a geometric phase change-index (GPC-I) and extend the method to quantitative pore structure analysis through global geometric phase measurements. Applied to additively manufactured Inconel 718 samples, our approach enables both spatial mapping of defect distributions and quantitative correlation with pore size, pore volume fraction, and pore clustering, as validated by micro-X-ray computed tomography. The method eliminates the need for prior material calibration, enabling real-time, model-free detection and microstructure evaluation under practical manufacturing constraints, offering a path toward scalable, high-fidelity structural health monitoring and quality assurance in complex manufacturing environments.

Due to the high risk of neurological or vascular injuries related to cortical perforation during pedicle screw insertion, the endoscopic ultrasound imaging shows potential for radiation-free intraoperative navigation, capable of providing real-time guidance for the insertion. To ensure penetration depth, low-frequency ultrasound is often selected. However, the quality of trans-osseous ultrasound imaging is affected by the limited resolution as well as the bone-induced aberration. To improve the image quality, a virtual source synthetic aperture focusing and coherence factor weighting method (VSSAF-CFW) was adopted for imaging, and a hybrid simulation method was applied to correct phase aberration. This paper proposed a hybrid analytical-numerical simulation phase correction method based on intermediate virtual points (HAN-IVP). This method is based on a simplification of the conventional full-wave model. The agreement of the proposed HAN-IVP method and conventional method was assessed by 12 sets of comparative simulations under various conditions. The simulation results indicate that the mean error of Time-of-Flight (TOF) is -0.003 μs and the average Pearson correlation coefficient is 0.915. The performance of the proposed HAN-IVP-based VSSAF-CFW method was evaluated through in vitro experiments using both plate and ring bone specimens. In the plate bone experiment, the proposed method reduced the center positional shift by 0.385 mm, reduced the -6 dB lateral beam width by 1.13 mm, and increased the peak-to-sidelobe contrast ratio by a factor of 1.26. The ring bone experiment further validated the applicability of the proposed method under a curved bone geometry closer to the trans-osseous endoscopic imaging scenario. The HAN-IVP method achieved precise phase correction while reducing the computational cost. The combination of HAN-IVP and VSSAF-CFW effectively improved the quality of rotational scanned trans-osseous ultrasound imaging, demonstrating its potential for future clinical translation.

The volumetric grain size distribution (GSD) plays an important role in the mechanical response of polycrystalline materials and is often assumed to be lognormal in ultrasonic studies, although gamma- and Gaussian-type distributions have also been reported. However, the influence of GSD type on elastic-wave scattering in polycrystals remains insufficiently understood. In this work, we compare three commonly used GSD types, namely Gamma, lognormal, and truncated Gaussian, and quantify their effects on the two-point correlation function (TPCF) and ultrasonic attenuation. Synthetic polycrystals with identical mean grain size and coefficient of variation (CV) are generated for two representative materials, aluminum and α-iron. The TPCFs measured from the synthetic microstructures are accurately reproduced by a recent analytical model introduced by the authors. For identical low-order grain size statistics, the truncated Gaussian distribution yields the shortest correlation length, followed by the Gamma distribution, whereas the lognormal distribution gives the longest one. Analytical expressions for attenuation in the Rayleigh regime are also derived for different GSDs, and comparison with the semi-analytical Weaver's model confirms the robustness of the proposed framework. The results show that changing only the GSD type can produce significant differences in low-frequency attenuation, with maximum relative differences of about 28% and 58% for CVs of 0.35 and 0.5, between the truncated Gaussian and lognormal cases. An inversion study further shows that the relative errors for identified mean grain size can reach 90%. These findings highlight the importance of the GSD type in ultrasonic characterization.

Advancements in additive manufacturing (AM) have created a strong demand for the qualification of metal alloys used in critical structural components. Ultrasound offers a promising non-destructive technique for analyzing mechanical properties from the complex and diverse microstructural features of AM materials. Targeting the physical mechanism from ultrasound to mechanical properties, we introduce a quantitative measure of microstructural inhomogeneity derived from parameterizing microstructural features. This inhomogeneity by inducing strain localization and the subsequent accumulation of geometrically necessary dislocations (GNDs), governs the material deformation. We demonstrate that this same inhomogeneity parameter also indicates the degree of ultrasonic attenuation. This establishes a direct, physical link between the elastodynamic response (attenuation) and the core plastic mechanism (GND accumulation), which ultimately determines material strength. Furthermore, we show numerically that the same inhomogeneity descriptor governs ultrasonic attenuation across materials belonging to different crystal systems, and we validate experimentally its effectiveness in predicting the relative strength of LPBF-fabricated SS316L, SS17-4PH and pure titanium. The full mechanistic chain linking attenuation to GND accumulation is established experimentally for LPBF SS316L, while its extension to other crystal systems is supported by simulation and requires material-specific calibration. These findings reveal the physical mechanism underlying the experimentally observed correlation between ultrasonic attenuation and mechanical strength in AM materials, enabling new framework for ultrasonic applications and enhancing the safety and reliability of AM alloys in critical structural components.

We put forward the characteristic equation of Stoneley waves along wavy interfaces, and calculate the dispersion curves, wave structures and cutoff frequencies. After confirming the existence of Stoneley waves at flat interfaces, by incorporating wavy boundary conditions, we derive the governing equation and reveal unique propagation behaviors, specifically a velocity reduction followed by an increase at high frequencies. Furthermore, the cutoff frequency is directly proportional to the Stoneley wave velocity and inversely proportional to the wavy period. The results indicate that the stable Stoneley wave propagation is confined to frequencies below the cutoff limit and to interfaces with a relatively low steepness. Finite element simulations are performed, and four different experimental samples are employed to verify the existence and characteristics of Stoneley waves such as dispersion curves and cutoff frequencies. The findings significantly advance the understanding of Stoneley wave propagation characteristics in wavy interfaces and establish a robust theoretical foundation for nondestructive evaluation in structures with wavy interfaces.