Russian Journal of Nondestructive Testing最新文献

Defect inspection is critical for ensuring the safe and reliable operation of railways transportation systems. This paper presents a novel defect inspection system that combines the attention U-Net network, a type of neural network architecture, and a kind of active contour algorithm based on morphological operators to improve the accuracy of defect segmentation. The attention U-Net Network is used to generate an initial segmentation mask of the railway image with attention mechanisms that enable the network to focus on the most relevant features in the image. The active contour algorithm based on morphological operators is then applied to refine the segmentation mask. The system was tested on a dataset of railway images with various defects, and the results showed that the proposed system achieved higher accuracy in defect segmentation compared to traditional segmentation methods. The proposed system has the potential to improve the efficiency and reliability of railway defect inspection, leading to safer and more reliable railway transportation.

Ultrasonic non-destructive and imaging testing using the line array method is conducted on titanium alloy bars and the sound field distribution of linear array transducer probes is simulated and analyzed. The simulation of the sound field distribution of titanium alloy bars with different diameters is performed by CIVA software. CIVA is used for sensitivity and parametric studies and it can quantify the response to expected defects and can perform different inspection strategies. In this study, the different diameter bars and various defects are simulated by considering different frequencies, element sizes and numbers, and the test platform is used to verify the standard bar. Finally, the bar with natural defects is tested and a detection scheme based on the ultrasonic phased array method is proposed. By the simulation of focusing beam distribution, the distribution of the beam focus in the flat bottom holes can be analyzed, this can help the selection of focusing mode in the detection process and enhance the signal-to-noise ratio of defect detection. The test results show that utilization of the line array transducer to scan the titanium alloy bar can effectively solve the problem of defect detection in the titanium alloy bars.

Corrosion is one of the most common types of bridge cable damage, which can further develop into cable breakage. In this work, a novel corrosion degradation assessment method based on multiply scattered guided waves is proposed for directing the management and maintenance of bridge cables. The results reveal that simulated corrosion surfaces based on fractal theory can accurately describe steel wire surfaces with different degrees of corrosion. As the corrosion degree increases, multiple scattering echoes form multiple energy gathering points in the time–frequency domain, and the time–frequency energy of the signal gradually transfers from an end echo to multiple scattering echoes. With an increase of the total mass-loss rate, our proposed evaluation indicator rises monotonously. We show that the evaluation indicator, constructed by multiple scattering echoes, can effectively characterize the corrosion degree and is sensitive to identifying early signs of corrosion. Compared with other indicators, such as the phase velocity, the group velocity and the attenuation rate, the proposed indicators show the superior ability to characterize corrosion degradation. High-frequency guided waves have high resolution and sensitivity for characterizing degradation caused by corrosion. However, the optimal detection frequency still needs to be determined based on the attenuation of guided waves, because high-frequency guided waves attenuate more rapidly than lower-frequency waves. The interaction between wires in bridge cables has little effect on corrosion degradation assessment. With an increase of the technical grade of bridge cables, the amount of corrosion and the evaluation indicator rise gradually. The variation law of the measured evaluation indicator as a function of corrosion degree is in good agreement with our finite element analysis results, and the corrosion degradation assessment results of bridge cables using guided waves agrees well with our results based on visual inspection. Our study indicates that the corrosion degradation assessment of wires inside steel cables can be realized without damaging the outer sheath when using our assessment indicator based on the multiply scattered guided waves.

The electrical properties of a material representing a composite of epoxy resin, magnetic fluid, and carbon nanotubes are investigated. It is shown that in composites dried in the presence of a magnetic field, elongated conductive structures consisting of carbon nanotubes and magnetic fluid are formed. Their presence causes the appearance of anisotropy of the electrical properties of such composites. The anisotropy of the properties is studied by microwave waveguide methods, according to the frequency dependence of the reflection coefficient of microwave radiation from a periodic structure in which the composite under study was used as a damaged layer. It has been found that the electrical properties of the composite depend on the magnitude and direction of the magnetic field induction, as well as on changes in the concentration of components in the composite. Numerical modeling was performed and the importance of taking into account the anisotropy of the electrical properties of the formed structures when calculating the integral parameters of the composite is shown.

An ultrasonic reconfigurable reflector technique was introduced in the waveguide sensor to monitor the temperature of the pipe surface. Researchers used mostly cylindrical wire waveguides to measure fluid level, rheology, and temperature. However, the strip waveguides have flat surfaces that can show a better coupling effect with ultrasonic (transducer) sources. Also, the strip sensor can lie/lay easily on the measurement region and have more surface contact. This sensor development considered the S₀ wave mode in the thin strip as a pulse-echo approach using a single transducer at 0° orientation with the waveguide axis. We considered echogenic features (clamp-reflectors) to develop the distributed temperature sensors in the ultrasonic strip waveguide. Here, the reflector types “screw & clamp” were introduced to obtain the desired strength of amplitudes from each reflector which can be helpful in the signal’s peak tracking. We compress the waveguide at appropriate locations using a screw-clamp setup to obtain a suitable ultrasonic reflection without removing the material (notch reflector). We obtained change in time of flight (δTOF) between consequent reflectors at various temperatures and compared it with the co-located conventional thermocouple for calibrating the waveguide sensor. Then, we used the calibrated single-strip waveguide with a reconfigurable reflector to measure temperatures at multiple locations on the pipe surface. We performed multiple experimental trials to check for the sensors’ repeatability. The single-strip waveguide sensor developed comprises non-destructive reflectors that are easy to use, reconfigurable, durable, and cost-effective. Measuring the in-situ properties of any structure at various locations could be highly feasible.

In the aerospace industry and nuclear power plants, bolt preload is a crucial metric for assessing the quality of installation and the state of service of the flange connection parts of pressure vessels. The primary technique for determining the preload force is the ultrasonic stress measurement method based on pulse time-of-flight. However, the current ultrasonic preload measuring approach is no longer applicable for unique bolted structures like pointed bolts and high acoustic attenuation bolts, as it is unable to receive the bolt end face echo. This work aims to address this issue by proposing a detection system based on interlayer ultrasonic energy transfer-based bolt preload measurement method. The ultrasonic energy transmission model of the sandwich’s rough interface is built using both classical contact theory and ultrasonic propagation theory. When paired with a finite element simulation, this model shows a linear relationship between the sandwich’s ultrasonic transmission signal energy and the bolt preload force. The relationship between the ultrasonic transmission energy and the bolt preload force under the ultrasonic frequency contact surface roughness and other factors was systematically investigated. A testing platform was set up to conduct the ultrasonic energy transmission test on the preload force of the bolt interlayer. According to the test results, there is a good linear relationship between the ultrasonic transmission signal energy and bolt preload obtained under various groups of parameters, R² is greater than 0.97, and the bolt preload measurement error range for flange bolts is between –5.4% and 5.6%. These findings suggest that the method presented in this paper has a promising future.

Carbon fiber reinforced polymer (CFRP) has been extensively utilized in the aerospace industry due to their light weight and high strength, however, they are susceptible to defects such as delamination and debonding during service. To enhance material safety, reliability and defect detection efficiency in infrared non-destructive testing (NDT), this study treats each pixel in the thermal image of the specimen surface as an independent entity. Temporal thermal wave signal features are extracted, and after non-dimensional processing, the features are mapped back to each pixel to reconstruct the characteristic distribution on the specimen surface, leading to the proposal of the dynamic thermal regression (DTR) algorithm. The DTR technology, along with the dynamic thermal tomography (DTT) and thermal signal reconstruction (TSR) techniques, were applied to the original infrared image sequences. The results demonstrate that applying these image processing techniques significantly enhances defect detection in CFRP. Furthermore, the DTR technique effectively reduces the acquisition time for infrared NDT image sequences, shortens the sequence length, and thereby improves the efficiency of infrared NDT.

High-pressure hydrogen valves are subjected to the instantaneous impact of hydrogen flow and repeated start-stop action during service, and there is a potential risk of leakage. This paper investigates monitoring and identification of hydrogen valves leakage to ensure their operational reliability. Firstly, an acoustic signal monitoring system was built based on a high-pressure hydrogen gas-tightness test platform, and the time-domain feature of valves under different leakage conditions was analyzed. Secondly, the frequency-domain feature is extracted using a combination of  variational modal decomposition and wavelet packet decomposition. Ultimately, the backward propagation network (BP) and convolutional neural network (CNN) are used to recognize patterns of acoustic signals, with the time-domain and frequency-domain parameters as feature inputs independently. The results show that the accuracy of BP and CNN networks based on frequency domain features has significantly improved, 93.33 and 91.67%, respectively. This paper obtained the feature extraction and pattern recognition method for hydrogen valves, which provides a reference for accurate and efficient recognition of the leakage condition of high-pressure hydrogen valves in the service process.

We consider the 2D magnetostatics inverse problem for a uniformly magnetized body and reduce it to a nonlinear 1D integrodifferential equation determining the body (cavity) shape based on the measured strength of the external magnetic field. We design a numerical algorithm for solution of this equation based on minimization of a function of several variables and develop a FORTRAN program implementing this algorithm. To test and illustrate our approach we find a solution for the cross section of a homogeneous infinite cylinder in a nonmagnetic and opaque medium based on the known strength of the external field.

This paper proposes a roll defect recognition method based on C-GAN and CNN-Attention, addressing the challenges of limited data and low recognition accuracy in ultrasonic defect detection for rolls. Initially, an ultrasonic testing experimental system is employed to inspect artificially prepared roll defect samples, leading to the collection of actual defect data. Subsequently, a C-GAN data augmentation model is developed to learn the distribution patterns of various defects, generating high-quality new samples that align with the distribution of each defect type, thereby expanding the training dataset. Utilizing this augmented data, a convolutional neural network defect classification method that incorporates an attention mechanism is designed to further enhance prediction accuracy. By integrating an attention module to assign weights to each feature channel, improved feature representations are achieved, optimizing the learning mechanism of the CNN. The model attains a recognition accuracy of  95.83%, demonstrating the effectiveness of this method in roll defect recognition applications.