Nondestructive Inspection Techniques Research Articles

The use of nonlinear guided-wave features for detection and assessment of intermetallic compounds within dissimilar joints – an experimental investigation

Abstract Welding dissimilar materials is widely employed in industrial construction and manufacturing to enhance cost-effectiveness and performance, often utilizing non-fusion methods like solid-state and high-energy beam welding. However, a significant challenge is the formation of intermetallic compounds (IMCs) at the joint interface, which can weaken the bond and increase brittleness, leading to hidden internal cracks. Nonlinear ultrasound detection methods are employed as advanced, non-destructive testing techniques for early damage inspection in various materials. This research investigates the assessment of the thickness of the intermetallic layer within dissimilar joints using non-linear ultrasound-wave features. Numerical and experimental investigations were performed using four friction stir welding (FSW) lap joints, between AA5052-H32 aluminum and ASTM 516-70 steel, with various intermetallic thicknesses. The methodology involved examining the generation of second-order harmonic frequency by exciting Lamb waves (LW) at specific frequencies. To determine the necessary LWs' excitation frequency, synchronism, and non-zero power flux conditions were employed. The collected signals were measured and analyzed in the time and frequency domains to understand the behavior of the non-linear parameter β′ with the thickness of the intermetallic layer. The results show that β′ changes in a linear manner with the thickness of the intermetallic compound layer (several micrometers in thickness). This provides strong evidence that nonlinear LW features are sensitive to microstructural variations in the FSW joints, which enables them to effectively evaluate their strength.

Enhancing Inductive IR Thermography by Using FFT-Equalization, Motion Tracking Detection and VDSR Super-resolution Processing

Among non-destructive inspection techniques, infrared thermography stands out as a promising technology that enables real-time inspection of large areas without the need for physical contact. In this study, we employed the dynamic induction thermography method, which is one of the active infrared thermography techniques, to detect defects on the back side of the S275 material specimen. This technique involves creating relative movement between the IR camera and the specimen. We acquired sequence images at different moving speeds using the induction thermography technique, and then used the FFT with Gaussian filtering to solve for non-uniform heat sources. To further enhance the resolution, we applied the VDSR technique, which is based on deep neural networks. The effectiveness of this approach was validated both qualitatively and quantitatively. Finally, we utilized the MOT algorithm to automatically detect defects in the image with the highest thermal contrast, which was captured at a speed of 15 mm/s. In this study, we demonstrate the effectiveness of thermal equalization using the GF-based FFT algorithm, as well as the superresolution conversion achieved through the VDSR-based deep neural network. Additionally, we present a mechanism for automated slit detection using the MOT algorithm.

EVALUATING MODEL ROBUSTNESS FOR DEFECT IDENTIFICATION AND CLASSIFICATION IN COMPOSITE AEROSTRUCTURE MATERIAL

Abstract Aircraft structures are required to have a high level of quality to satisfy their need for light weight, efficient flight, and withstanding high loads over their lifespan. These aerostructures are typically made from composite material due to their good tensile strength and resistance to compression. To ensure their structural integrity, the composite material requires inspection for common flaws such as porosity, delaminations, voids, foreign object debris, and other defects. Ultrasonic testing (UT) is a popular non-destructive inspection (NDI) technique used for effectively evaluating composite material. Current inspection methods rely heavily on human experience and are extremely time consuming. Therefore, there is a need for the development of techniques to reduce the manual inspection time. This work compares the performance of different deep learning-based methods in the identification and classification of defects. Deep learning has shown great promise in numerous fields, and we show its effectiveness in the evaluation of composite aerostructure material. The methods developed here are both highly reliable with a top recall value of 98.64% as well as extremely efficient requiring an average of 4 seconds during the inferencing stage to evaluate new composites. Finally, we investigate model robustness to concept drift by measuring its performance over time.

The Health Monitoring of Bonded Composite Joints Using Both Thickness and Radial Impedance Resonance Responses.

With the advent of bonded composites in today's aircraft, there is a need to verify the structural integrity of the bonded joints that comprise their structure. To produce adequate joint integrity, strict process control is required during bonding operations. The latest non-destructive joint inspection techniques cannot quantify the strength of the bond and only indicate the presence of disbonds or delaminations. Expensive and timely proof-load testing of the joints is required to demonstrate structural performance. This work focuses on experimentally evaluating joint-health monitoring with piezoelectric sensors exposed to repeated loadings until failure. Single-lap-shear composite joints are structurally tested while using sensor electromechanical impedance response as a health indicator. Based on these experiments, validation of this novel method is achieved through detailed evaluation of the sensor impedance response characteristics during loading, which enable initial and prognostic joint health assessments. The experimental results indicate that the embedded piezoelectric sensors are able to measure the sensor impedance radial and thickness resonance response changes prior to joint failure, without sacrificing the joints' structural performance.

Comparison of Different NDT Techniques for Evaluation of the Quality of PCBs Produced Using Traditional vs. Additive Manufacturing Technologies.

Multilayer printed circuit boards (PCBs) can be produced not only in the traditional way but also additively. Both traditional and additive manufacturing can lead to invisible defects in the internal structure of the electronic component, eventually leading to the spontaneous failure of the device. No matter what kind of technology is used for the production of PCBs, when they are used in important structures, quality control is important to ensure the reliability of the component. The nondestructive testing (NDT) of the structure of manufactured electronic components can help ensure the quality of devices. Investigations of possible changes in the structure of the product can help identify the causes of defects. Different types of manufacturing technologies can lead to diverse types of possible defects. Therefore, employing several nondestructive inspection techniques could be preferable for the inspection of electronic components. In this article, we present a comparison of various NDT techniques for the evaluation of the quality of PCBs produced using traditional and additive manufacturing technologies. The methodology for investigating the internal structure of PCBs is based on several of the most reliable and widely used technologies, namely, acoustic microscopy, active thermography, and radiography. All of the technologies investigated have their advantages and disadvantages, so if high-reliability products are to be produced, it would be advantageous to carry out tests using multiple technologies in order to detect the various types of defects and determine their parameters.

Eddy currents operation parameters optimization

With the development of non-destructive inspection techniques, more and more challenging situations have arisen and the correct choice of operating parameters can be decisive for a good detection sensitivity. Based on that, an algorithm was developed to obtain the best combination of an eddy current sensor operation parameters. The results obtained demonstrate the improvement in the cracks detection in welded parts after optimization.

In Situ Measurement of the Machining State in Small-Diameter Drilling by Acoustic Emission Sensing

In drilling small holes with diameters of 1 mm or less, minute clogging and twining of chips or the adhesion of the workpiece material can become factors in causing breakages of the drill bit; moreover, it can be difficult to identify the machining state. Acoustic emission (AE) sensing is a nondestructive inspection technique that measures the elastic-stress waves that are generated when a material is deformed and fractured. AE sensing permits highly sensitive measurements to be made without changing the rigidity of the experimental system, unlike force sensing of cutting resistance, etc. In the present study, attempts were made to identify the machining state and tool wear, and to predict abnormalities in small-diameter drilling by using the change in the frequency of AE signal waveforms arising from deformation and fracture. It was shown that it is possible to predict the breakage of the drill bit by detecting high-frequency AE signals at about 1 MHz, caused by the adhesion of the workpiece material. In addition, a correlation map of the AE frequency spectrum for identifying the machining state in a drilling operation is suggested.

Coherent Off-Axis Terahertz Tomography with a Multi-Channel Array and f-theta Optics.

Terahertz tomography is a promising method among non-destructive inspection techniques to detect faults and defects in dielectric samples. Recently, image quality was improved significantly through the incorporation of a priori information and off-axis data. However, this improvement has come at the cost of increased measurement time. To aim toward industrial applications, it is therefore necessary to speed up the measurement by parallelizing the data acquisition employing multi-channel setups. In this work, we present two tomographic frequency-modulated continuous wave (FMCW) systems working at a bandwidth of 230-320 GHz, equipped with an eight-channel detector array, and we compare their imaging results with those of a single-pixel setup. While in the first system the additional channels are used exclusively to detect radiation refracted by the sample, the second system features an f-θ lens, focusing the beam at different positions on its flat focal plane, and thus utilizing the whole detector array directly. The usage of the f-θ lens in combination with a scanning mirror eliminates the necessity of the formerly used slow translation of a single-pixel transmitter. This opens up the potential for a significant increase in acquisition speed, in our case by a factor of four to five, respectively.

Signal Process of Ultrasonic Guided Wave for Damage Detection of Localized Defects in Plates: From Shallow Learning to Deep Learning

Ultrasonic-guided waves are one of the non-destructive inspection techniques used in structural health monitoring (SHM) for localized damage detection. Propagation, scattering, and dispersion of the Lamb waves Even though the great progress in the monitoring technology of the Lamb wave, the detection, identification, and quantification of structure damage still face challenges because of the complexity of the process in the propagation, scattering and dispersion of the Lamb wave. Machine learning methods, including shallow learning (support vector machine (SVM)) and deep learning methods (convolutional neural networks (CNN) and long short-term memory (LSTM)), in recent years have brought revolutionary opportunities for decoding the information of the Lamb wave. Therefore, the methodology structure was proposed from dataset collection, data preprocessing (including feature extraction, feature combination, and feature transformation), data training, and classification. Two different cases of damage types and damage sizes were designed in a COMSOL environment. The shallow learning method of the SVM model and the deep learning method of the CNN-LSTM model were compared with the defined time-series features and transformed images. The results showed that both shallow learning methods and deep learning methods can be used in the application of signal classification, while the deep learning method of CNN-LSTM exhibited higher accuracy in image classification, as compared to the SVM. The robustness of the proposed models has also been verified under noise interference. The results demonstrated that the deep learning architecture of CNN-LSTM has the potential to attain greater precision because of better feature extraction and processing ability than the shallow learning model of SVM. In addition, the performance of signal classification and image classification of SVM and CNN-LSTM models dramatically decreased as the noise levels increased.

Fast and accurate prediction of cure quality and mechanical performance in fiber‐reinforced polymer composite using dielectric variables and machine learning

AbstractFiber‐reinforced polymer (FRP) composites are widely used in the aerospace and automotive industries due to their high strength‐to‐weight ratio. However, the manufacturing process of FRP composites is intricate, requiring precise control over multiple parameters, which can be challenging. To ensure top‐notch product quality and bolster confidence in the durability of composite materials, it requires a uniform curing evaluation technique and a predictive strength model. This study presents a rapid nondestructive quality inspection technique for composites, involving three distinct studies that employ machine learning techniques in combination with the frequency‐dependent dielectric response of materials. The first study introduces a nondestructive method for swiftly inspecting the cure state (degree of cure) of composite samples. This technique combines broadband dielectric spectroscopy with supervised machine learning algorithms, particularly support vector machines. The second study employs advanced artificial neural networks like multi‐layer perceptron to predict the tensile strength, a measure of mechanical performance, of composite materials. The results demonstrate an accurate classification of the curing state with 96.7% accuracy and a prediction of tensile strength with 87.5% accuracy. The final study explores the application of machine learning in quality monitoring of prepreg (raw materials for FRP) aging at room temperature.Highlights Integration of machine learning with frequency‐dependent dielectric measurement offers efficient method for inspecting and monitoring the curing state and mechanical performance of composite materials. Machine learning and frequency‐dependent dielectric response used to accurately assess the curing state of fiber‐reinforced polymer composites with 96.7% accuracy, which saves the time and effort and also reduces the need for destructive testing. Advanced artificial neural networks, specifically multi‐layer perceptron, employed to predict mechanical performance of composite materials with 87.5% accuracy, enhancing confidence in product reliability. Novel dielectric measurement technique integrated with machine learning algorithms enables prepreg age monitoring at room temperature.

CT-Based Dimensional Metrology for Quality Assessment of the Internal Structure of Additive Manufactured Aluminum Parts

Two common types of internal defects of additively manufactured (AM) samples are lack-of-fusion and gas-entrapped porosities. These internal defects can have different physical origins and particular local characteristics (e.g., different shape, size). Thus, the use of reliable non-destructive inspection techniques is essential for the accurate assessment of integrity, allowing the applicable AM processing parameters correction. To overcome this challenge, this work aims to evaluate the accuracy of volumetric characteristics measured by computed tomography for porosity evaluations in AM samples, including assessment of measurement uncertainty. The effect of different cumulative thickness on the evaluated measurements accuracy is also assessed. The results show that deviations of defect size measurements can be below 2% if the proposed procedure is followed. In addition, the expanded uncertainty can be up to 10% of the measured magnitude when the cumulative thickness is increased to 70 mm. The physical relationships obtained between the cumulative thickness and the individual measurements are also presented.

Defect detection and imaging in elastic materials with complex geometries

Additive manufacturing has made it possible to easily create parts with complex shapes, even out of metal materials. However, in order to guarantee the quality of such parts, there is a need to establish a nondestructive inspection technique that is feasible for such complex structures. In this presentation, I will introduce a technology that can visualize internal damage even in complex objects by utilizing the ultrasonic field generated and diffused inside the object with non-contact manner. Ultrasonic waves generated by laser irradiation resonate with the natural frequency of defects such as delamination near the surface of an object. By scanning the laser beam for generating ultrasonic waves and detecting vibrations at a fixed position on the surface of the object, an image of defects in the vicinity of the surface can be obtained as a distribution of vibration energy. The distribution of vibration energy does not depend on reflected or scattered waves from the object's wall, but only on information from the point where the laser is irradiated. This technique is expected to be especially useful for in-situ inspections of the additive manufacturing objects.

Scalable in situ non-destructive evaluation of additively manufactured components using process monitoring, sensor fusion, and machine learning

Laser Powder Bed Fusion (L-PBF) Additive Manufacturing (AM) is among the metal 3D printing technologies most broadly adopted by the manufacturing industry. However, the current industry qualification paradigm for critical-application L-PBF parts relies heavily on expensive non-destructive inspection techniques, which significantly limits the use-cases of L-PBF. In situ monitoring of the process promises a less expensive alternative to ex situ testing, but existing sensor technologies and data analysis techniques struggle to detect sub-surface flaws (e.g., porosity and cracking) on production-scale L-PBF printers. In this work, an in situ NDE (INDE) system was engineered to detect subsurface flaws detected in X-Ray Computed Tomography (XCT) directly from process monitoring data. A multilayer, multimodal data input allowed the INDE system to detect numerous subsurface flaws in the size range of 200–1000 µm using a novel human-in-the-loop annotation procedure. Furthermore, a framework was established for generating probability-of-detection (POD) and probability-of-false-alarm (PFA) curves compliant with NDE standards by systematically comparing instances of detected subsurface flaws to post-build XCT data. We also introduce for the first time in the AM in situ sensing literature the a90/95 – the flaw size corresponding to a 90% detection rate on the lower 95% confidence interval of the POD curve. The INDE system successfully demonstrated POD capabilities commensurate with traditional NDE methods. Traditional ML performance metrics were also shown to be inadequate for assessing the ability of the INDE system’s flaw detection performance. It is the belief of the authors that future studies should adopt the POD and PFA approach outlined here to provide better insight into the utility of process monitoring for AM.

Corrosion under insulation: a rewiew

Corrosion under insulation (CUI) occurs from electrochemical reactions on the surface of thermally insulated industrial equipment and has water as an electrolyte. CUI generates many costs for companies due to maintenance and can also cause problems to human and environmental integrity. In this work, vast bibliographic research was carried out on corrosion under insulation, in articles published between 1998 and 2022, with the objective of evaluating the CUI in Industry, factors that influence it, corrosion mechanism, non-destructive inspection techniques and techniques that can work as a mitigation for this problem. As a result, it was confirmed that corrosion under insulation is a serious problem for the industry, especially the petrochemical one, as its detection is not so simple and is expensive when done visually. To solve this difficulty in identification and detection, non-destructive tests are used, through various technologies, it is possible to prevent the insulation from being damaged and, in some cases, the inspection can be carried out without stopping the process. In addition, serious situations can be generated due to the non-monitoring of the CUI, such as: Economic, human and environmental integrity impacts.

Non-Contact MEMS-Sensor Array Inspection of Composites and Metallic Parts Using Lamb Waves

Abstract Composite materials are becoming more popular in the aerospace industry, because of their physical properties. In quality assurance and in-service inspection, there is a need for fast, non-contact, high-quality, non-destructive inspection techniques. The most common approach is to perform the inspection using water-coupled high-frequency transducers. Full wavefield techniques are promising to replace the conventional inspection approach. However, these are currently performed by a laser vibrometer setup, which has drawbacks. As an alternative, a low-cost micro-electro-mechanical system (MEMS) sensor array and dedicated processing scheme are presented enabling fast inspection of large samples. This inspection approach uses a piezoelectric actuator to excite the composite or metallic part with Lamb waves. An array of MEMS sensors records the energy that radiates into the surrounding air. A dedicated processing scheme will translate the measured wavefield into a thickness map of the inspected part. For composite parts, the material’s anisotropy needs to be taken into account for accurate thickness mapping. In principle, all relevant defects show up as local thickness reductions. The results in this paper are obtained with a MEMS-sensor array of 128 elements capable of detecting ultrasound up to 250 kHz at a typical stand-off distance of 100 mm. Defects up to 6 mm in diameter could be detected in thick panels, and defects as small as 2.5 mm could be detected in thin panels. A full-size fuselage experiment shows that the method is also suited for fast inspection of large inspection areas.

Damage progression monitoring using self-sensing capability and acoustic emission on glass fiber / epoxy composites and damage classification through principal component analysis

In this study, the synergistic combination of acoustic emission (AE) and self-sensing capability provided by the integration of carbon nanotube (CNT) networks was used to a better damage progression monitoring of glass fiber epoxy composites under flexural loading. The specimens were prepared with different stacking sequences ([06], [04], [0/90]S, and [90/0]S) using the vacuum assisted resin infusion procedure. The acoustic signals recorded during the tests were analyzed by a classification methodology which consists of the k-means method and principal component analysis (PCA), this analysis allowed the identification of the various damage mechanisms such as matrix cracking, fiber/matrix debonding, delamination and fiber breakage. The electromechanical response, through the change in electrical resistance signal, and some AE parameters such as the cumulative energy and/or strength cumulative were able to capture the occurrence of specific failure events during the composites damage evolution. Damage evaluation of failed specimens by means of scanning electron microscopy was performed, and it exhibited a less damage severity on specimens with CNT due to the reinforcement effect. This work demonstrates the complementarity of both non-destructive inspection techniques for monitoring damage progression in composite structures.

Investigation of neutron imaging applications using fine-grained nuclear emulsion

Neutron imaging is a nondestructive inspection technique that has a wide range of applications. One of the important aspects of neutron imaging is achieving a micrometer-scale spatial resolution. The development of a high-resolution neutron detector is a challenging task. As one potential solution to this task, we investigate whether neutron detectors based on fine-grained nuclear emulsions are suitable for high-resolution neutron imaging applications. High track density is necessary to improve the quality of neutron imaging. However, the available track analysis methods are difficult to apply under high track density conditions. Simulated images are used to determine the required track density for neutron imaging. A track density of the order of 104 tracks per 100×100μm2 is sufficient to utilize neutron detectors for imaging applications. Contrast resolution was also investigated for image datasets with various track densities and neutron transmission rates. Moreover, experiments were performed for the neutron imaging of gadolinium-based gratings with known geometries. The grating structures were successfully resolved. The calculated 1σ 10%–90% edge response using the grayscale optical images of the grating slit with a periodic structure of 9 μm was 0.945±0.004 μm.

Nonlinear ultrasonic characterization of joint structures using backscatter guided waves

Abstract Joint structures made from plates or shells are frequently used in large constructions, such as oil storage tanks in petrochemical industries. The corrosion-induced metal loss often appears on the back side of the bottom plate supported by ground or sand. The welding residual stress and weight of shell walls accelerate the corrosion. Determining the severity of metal loss under the joint is challenging. This work uses the mode pair technique for nondestructive inspection of inaccessible regions in joint structures. A finite element analysis model simulates the transient responses of the incident and mode-converted fundamental and second harmonic plate waves. The second harmonic s0 plate wave satisfying the phase velocity matching condition accompanies the transmitted fundamental SH0 plate wave. A directional filter identifies every mode-converted plate wave by comparing the frequency spectra with the dispersion curves. Linear ultrasonic guided waves can detect inaccessible regions using mode-converted reflections. The symmetry of the linear mode-converted fundamental plate waves depends on the defect position and geometry, not the size. The mode-converted second harmonic guided wave can achieve better spatial resolution for defect sizing. The nonlinear parameter shows a proportional trend with an increasing thickness reduction. Experimental and numerical evidence reveals nonlinear ultrasonic guided wave technique has the potential for defect detection in joint structures.

A novel method for defects marking and classifying in MFL inspection of pipeline

Magnetic Flux Leakage (MFL) testing is the most widely used non-destructive technique for the inner inspection of oil and gas pipelines. Accurate quantification of defects is a long-standing difficulty in the field of pipeline leak detection. Scientific marking and classification of defects is an important prerequisite for accurate quantification. A novel method of marking and classifying defects with MFL signals is proposed, which is aiming at the problem that it is difficult to mark and classify defects accurately due to the tremendous amount of oil and gas pipeline leakage magnetic detection data. An improved CLIQUE algorithm is used to mark the defect areas of segmented pipelines to estimate the number and location of defects. Then the 3D MFL characteristic signals of the marked areas are studied and extracted. The SSA_BP neural network is trained to classify the defects. The effectiveness and accuracy of the proposed method are tested and verified by using finite element simulation defects and actual defects. The results show that the method is more efficient in marking defects and more detailed in marking areas.

Assessment of 3D micro-computed tomography for PV interconnection technology development

X-ray (micro-)computed tomography is a non-destructive characterization method using photons to obtain multiple radiographic projections, which generates a 3D volume rendering of a scanned object after reconstruction. It is recently used in the research and development of photovoltaic modules to investigate smaller volumes in a PV module, for example, detection of cavities in solder joints and visualization of module volumes up to connection ribbon sizes. This work shows that current advanced devices are able to scan larger volumes, using a single-cell 20 × 20 × 0.5 cm 3 module with back-contact multi-ribbon interconnection technology as an example. The generated 3D volume renderings show the interconnection structures and give information on ribbon-metallization interconnections. Next to this, also the solder reflow volume is visible, indicating that the technology has great potential to be used in further interconnection technology development. Finally, a destructive cross section analysis using scanning electron microscopy is performed; it has a higher resolution and can differentiate elements with similar photon absorption coefficients, thus enabling microstructure and intermetallic compound analysis. The findings show the potential for micro-computed tomography as a non-destructive and relatively fast visualization method and how it can give complementary information to other characterization techniques for photovoltaic interconnection research. • X-ray computed tomography is a non-destructive visual inspection technique. • Scanning electron microscopy and x-ray tomography are complementary. • X-ray tomography can aid PV interconnection technology developments.