Kai Zhou

Lamb wave-based structural health monitoring (SHM) technology provides an effective means for online damage monitoring of carbon fiber-reinforced polymer (CFRP) plates. However, affected by the complex multimodal and dispersive characteristics of Lamb waves, feature extraction from damage monitoring signals heavily relies on manual experience. To address this issue, this paper proposes an automatic Lamb wave damage localization method integrated with deep learning, where an end-to-end deep learning framework fusing the one-dimensional convolutional neural network and the Kolmogorov-Arnold Network (1DCNN-KAN) is developed. Taking raw damage monitoring signals as input, the 1D-CNN is adopted to extract local features of the signals, while the KAN decision layer explicitly models the mapping relationship between features and damage locations through learnable one-dimensional spline functions, thereby realizing structural damage localization without complex manual signal feature extraction. For damage location prediction, the damage monitoring area is gridded, and random damage is mapped to different grid nodes to achieve classification-based damage localization. To tackle the difficulty in acquiring monitoring signals of damage at arbitrary locations, a physics-informed data augmentation method is proposed, which generates Lamb wave monitoring signals covering random damage locations across the entire monitoring area based on the signals acquired at grid nodes. Finally, damage localization experiments are carried out on a CFRP plate. For damage of different sizes at 40 random locations (the minimum damage size is 10 mm × 5 mm × 20 mm), the 1DCNN-KAN model achieves grid-cell-level positioning for all test samples, and the predicted damage location and the actual location are both within the same grid cell (with a grid cell resolution of 20 mm × 20 mm). The classification accuracy of the damage location reaches 100%. Comparisons with different models demonstrate the superior performance of 1DCNN-KAN. This study provides an intelligent solution for damage localization of CFRP plates.

Lamb waves have been widely used in thin-walled structural health monitoring due to their long propagation distance, low attenuation, and high sensitivity to minor damage. However, traditional Lamb wave-based damage localization methods heavily rely on manual feature extraction and expert experience due to challenges such as multi-mode propagation, dispersion, boundary reflections, and environmental noise. This leads to complex signal processing and limited reliability, hindering practical engineering applications. To address these limitations, this paper proposes RCT-Net, a lightweight deep neural network with a serial hybrid architecture. The model integrates 1D-CNN with max pooling and residual blocks for local feature extraction and deep refinement, a CBAM attention mechanism to enhance damage-sensitive features, and a gated multi-head Transformer encoder that dynamically regulates attention head weights to overcome the equal-weight fusion limitation of standard Transformers. This design enables end-to-end damage localization directly from raw one-dimensional Lamb wave time series signals without manual feature engineering or prior wave propagation knowledge. Experimental results on aluminum plates demonstrate that RCT-Net achieves 100% classification accuracy under both noise-free and various noisy conditions, significantly outperforming CNN-LSTM and several advanced Transformer variants. In generalization tests on unseen damage locations, the model exhibits strong robustness, attaining 89.7% accuracy on Dataset 2 (58 previously unseen damage points) and 90.0% on Dataset 3 (300 samples generated through 10 repeated measurements per damage case). The proposed RCT-Net effectively balances accuracy, robustness, and computational efficiency, providing a practical and efficient solution for rapid damage localization in thin-walled structures under real engineering scenarios.

Water seepage in concrete can significantly degrade the durability of hydraulic concrete structures. Therefore, this paper introduces a new method that combines the percussion method with deep learning techniques to detect the depth of water seepage in concrete structures. Initially, percussion sound signals were collected for different water seepage depths. Then, the proposed one‐dimensional convolutional bidirectional gated recurrent unit (BiGRU) network with wide first‐layer kernel (1D‐WCBGRU) classifies the percussion sound signals for different water seepage depths. The 1D‐WCBGRU uses a wide first convolutional kernel to extract features directly from the original percussion signals without the need to extract features manually. Subsequently, the BiGRU is utilized to capture long short‐term information from the data, thereby enhancing feature separability and improving the classification accuracy and robustness of the model. Experiments confirm that the 1D‐WCBGRU exhibits excellent performance in the seepage depth detection task compared to traditional learning algorithms.

In this paper, the multidisciplinary coupled dynamics simulation method and reliability modeling method are studied for the problems of low accuracy, stuck and strength degradation of aero engines. Based on the ADAMS software platform, the parametric modelling is proposed for aero-engine guide vane adjustment mechanism of variable geometry turbine, and the parametric model is simulated. The effects of dimensional uncertainty, component deformation, load uncertainty, strength uncertainty and strength degradation on the reliability of the mechanism are considered. Through simulation analysis, the key factors affecting the reliability of the mechanism can be identified and guidance can be provided for the establishment of the reliability model. Finally, the reliability of the mechanism regarding motion accuracy failure, stuck failure and strength failure was calculated by Monte Carlo method.

The ultrasonic Lamb wave-based damage detection method is widely used in plate structures. Due to the multimode characteristic, the collected Lamb wave signals are very complicated, making it difficult to obtain damage information. This paper proposes a damage imaging algorithm based on the baseline construction of the Lamb wave for plate structures. A mobile transducer set including two transducers is designed for Lamb wave excitation and signal collection. To extract damage scattering signals, the response of the signal in healthy areas (the structural response signal) is measured as a temporarily constructed baseline signal. The structural response signal is obtained based on the identified healthy signals collected in different areas of the same structure. Damage scattering signals are extracted as the differences between the current signals and the structural response signal. Damage indexes are calculated based on the damage scattering signals to locate the damage with the imaging algorithm. A damage detection experiment is conducted on an aluminum plate to verify this proposed method. The imaging result is in good agreement with the damage location in the experiment. This paper provides a potential alternative for damage detection of plate structures in practice.

The baseline-free damage detection method of Lamb waves has the potential to obtain damage information efficiently in plate structures through damage scattering signals. However, the missing detection of damage occurs occasionally due to the angular scattering characteristic of Lamb waves. To solve this problem, a novel baseline-free damage detection approach based on path scanning at the detection region edges using mobile piezoelectric transducers is proposed herein. Several sensing points carrying separated damage scattering signals were picked out from the scanning paths. By removing the direct and boundary reflected signals, the damage signals were extracted and exported to a delay-and-sum imaging method to locate the damage. Two experiments with and without mobile transducers were conducted to validate the proposed method on an aluminum plate with artificially fabricated crack-like damage. The results show that the proposed baseline-free approach can locate the crack-like damage with high accuracy and efficiency and avoid potential loss of damage information. The proposed baseline-free method provides a novel and practical damage detection approach when considering the angular-dependent scattering characteristic of Lamb waves and can enhance the credibility of results in damage detection.

To active coherent jamming a new method is proposed by designing the orthogonal waveforms. Firstly, the design method of linear frequency modulated segmented recombination signal is given. By segmenting the linear frequency modulated (LFM) signal and then randomly recombining it,, the generated waveforms have good autocorrelation and crosscorrelation performance. Thus, the proposed method cannot only suppress false target jamming, but also resist frequency shift suppression jamming. Secondly, the signal characteristics is studied, and the correlation performance versus different segmentation number is analyzed. Thirdly, the anti-jamming analysis of the proposed method is presented when the designed waveform is applied. Finally, several simulations are conducted to demonstrate the effectiveness of the proposed method. The results show that LFM with random segmentation and recombination (RSLFM) is better than LFM signal in resisting false target jamming and suppressing jamming.

Aiming at the motion stagnation and local damage failure of a 2D needled Cf/SiC adjustable nozzle during operation, the motion performance of the nozzle and the contact stress at the key parts were studied. A dynamic simulation method of the nozzle with adjustable nozzle considering clearance fit was established, and the throat contraction process was analyzed by a simulation example. Firstly, the critical analysis positions of the nozzle were determined by trial calculation, and the model was simplified and the mesh was optimized. Then, the simulation was carried out under the boundary conditions involving multiple fields. The motion characteristics of the nozzle throat during the throat contraction process were analyzed due to the velocity results, and the stress state and failure state at the key position of the nozzle during the impact process were finally determined.

To assure the integrity and safety of truss structures, much work has been done for inspection in bars using ultrasonic guided waves methods. Guided waves have been widely used for Non-Destructive Testing (NDT) and Structural Health Monitoring (SHM) because of their long propagation distance and reliable responses. Though good for assessment of guided wave in the regular section structures such as cylinders, pipes and rails, the excitation of guided wave modes in irregular bars lack further research due to the arbitrary cross-section. This paper presents the excitation and propagation of guided waves in arbitrary cross-section structures by coupling the normal mode expansion (NME) method with the semi-analytical finite element (SAFE) method. Then, a Tee-bar example of finite element (FE) simulations is performed to validate the proposed SAFE-NME method. Moreover, the validated method is used to develop a double mode inspection strategy for multi-damages identification, showing the potential for NDT & SHM applications.