A reference-based inversion framework is presented for measuring the absolute acoustic nonlinearity parameter, β, of solids using nonlinear immersion ultrasonics. The proposed approach eliminates the need for direct transducer calibration by normalizing the measured second harmonic response of an unknown specimen with respect to a reference solid of known nonlinearity. A physics-based theoretical model describing finite-amplitude wave propagation in multilayer media is combined with through-transmission immersion measurements to isolate the contribution of the solid layer from surrounding fluid layers. The inversion is performed by matching the experimentally measured ratio of second harmonic amplitudes between the test and reference specimens with the corresponding ratio predicted by the model, which accounts for nonlinear accumulation, diffraction, and attenuation across all layers. The framework is validated experimentally using immersion measurements on cast aluminum, Ti-64, and SS 420. The inverted β values fall within the range typically reported for metallic materials, demonstrating the robustness of the proposed reference-based methodology and its suitability for repeatable, calibrationindependent nonlinear material characterization.
This paper analyses the usability of amplitude, damping, energy and frequency of sonic signals in standing trees for defect detection while using sonic tomography as non-destructive testing method in arboriculture. Data of sound (healthy) and defect trees was collected with a new sonic tomograph, the ”impactor”. The analysis was done using ”Extreme Gradient Boosting” as classification algorithm. The results show that it is possible to detect defects using the mentioned signal features. The algorithm was able to classify almost 80 % of the data correctly, where amplitude contains the most relevant information.
This study presents an enhanced Total Focusing Method (TFM) for ultrasonic phased array imaging, where geometric time-of-flight (ToF) calculations are replaced by wave-based ToF values obtained from Elastodynamic Finite Integration Technique (EFIT) simulations. Unlike conventional TFM, which assumes straight-line propagation, this approach accounts for refraction, scattering, and mode conversion. Experimental validation on aluminum specimens with side-drilled holes (SDHs) of varying depths and sizes demonstrates superior clarity, sharper reflector definition, and reduced artifacts. Both experimental and simulation-based comparisons between conventional and EFIT-based TFM are presented, with quantitative evaluation based on signal-to-noise ratio (SNR) and SDH size accuracy.
The emerging challenges brought about by sustainability requirements are leading to new applications of renewable resources. Unlike most conventional materials, renewable raw materials lower their carbon footprint by storing CO2 while they grow, making them highly attractive for industrial use. New composite concepts and joints incorporating these materials are being analysed with a view to develop non-destructive testing (NDT) solutions for quality control for subsequent production processes. NDT methods with potential to control specifically determined inspection tasks, such as computed tomography, ultrasonic, microwave or terahertz testing and active thermography, are being evaluated in terms of their respective strengths and limitations. Assessments of testability are discussed on the basis of both application-related examples and experimental investigation. The focus of this work is adapting active thermography to the considered renewable raw materials (timber and bamboo), their novel joints (resistance butt welds), and sustainable composites (COMBOO material). New approaches for thermal excitation provide initial information on detectable material and structural inhomogeneities. Analyses of the resistance butt welds with computer tomography reveal details of the inner structure and possible defects in the adhesion zone, making important information for further NDT development available.
The adoption of automated visual inspection systems is growing across various industries, such as manufacturing and energy, and is expected to expand significantly into other sectors, including aerospace. However, these systems often encounter challenges when visually inspecting highly reflective metallic surfaces, as varying light conditions can obscure critical surface details, thus risking errors in defect detection. This paper addresses these challenges by developing methods to detect specular light reflections in order to automatically assess the quality of inspection images. This enables automated systems to take inspection images from different angles, to avoid undesired reflections. We show that U-Net based architectures trained on a novel dataset of inspection images and reflection masks lead to good detection under challenging conditions. The results demonstrate that Convolutional Neural Network (CNN)-based models, particularly U-Net++ with a ResNet-50 encoder, outperform Transformer-based approaches, achieving the highest accuracy in identifying reflective areas. While the proposed UNETR-Attention Fusion (UNETR-AF) model shows promise for smaller reflections, it struggles with larger ones. This research offers a practical solution for industries aiming to improve visual inspection reliability, particularly for safety-critical applications. By enabling automated systems to handle reflective surfaces effectively, it addresses a significant gap in current inspection technologies.
Projection welding is a highly efficient welding process that can be applied with a high degree of automation. It is used in various industries for the production of automotive bodies, rail vehicles, kitchen appliances, and in the electronics industry. These industries face several challenges, including increasing safety requirements, growing material diversity, and increasingly regulated resource efficiency in the context of the circular economy. For these reasons, reliable quality assurance of projection welds is becoming increasingly important. Visual inspection is usually not applicable for quality assurance, as the welds are not visible from the outside. Monitoring of process parameters is used alongside regular destructive testing to assess joint quality. However, the latter contradicts good resource efficiency. No industrial standard has yet been established in the field of non-destructive testing (NDT). Obvious testing methods from resistance spot welding, such as manual ultrasonic testing, have not yet become established. The reasons lie in the variety of possible projection welds and the resulting high geometric variability of the components to be tested. This paper presents investigation results obtained from laboratory and practical ultrasonic testing of projection welded nuts. Test results obtained using coded excitation scanning acoustic microscopy (CESAM) are compared and discussed with those from manual ultrasonic testing. Possible development steps are derived by comparing these with the corresponding results from destructive testing of the welds.
Lightweight constructions offer the potential to save weight and thus climate-damaging CO₂ through less accelerated mass. Complex structures and hybrid material combinations are therefore being used in more and more applications. A key aspect of these structures are the joining processes used to connect different components made of various materials. Adhesive bonds offer great advantages here as a material-locking connection compared to ‘classic’ joining methods, as they can be used to join almost any combination of materials. With the adhesives currently available, high-performance materials can be joined together with high strength without additional components such as screws or bolts. In order to ensure safe operation, quality assurance is absolutely essential after production and during operation of such joints. Non-destructive testing methods, such as non-contact air-coupled ultrasonic testing, offer great potential here. In order to investigate the potential of these methods for defect detection in more detail, different types of defects were introduced on demonstrator components. These are tested using both air-coupled ultrasound and X-ray radiographic testing. It has been shown that pores and insufficiently cured adhesive can be visualized using airborne ultrasound.
Magnetic Barkhausen Noise measurements are advantageous for the fatigue characterisation of ferromagnetic materials since they can provide information regarding microstructure, residual stresses, hardness and the presence of microstructural defects. The fatigue characterisation of a quenched and tempered SAE 4140 steel was realised using a servo-hydraulic fatigue testing rig instrumented with an infrared camera and a Magnetic Barkhausen Noise sensor for capturing the material response. Magnetic Barkhausen measurements require correlation against a standardised measurement method such as X-ray diffraction technique. Therefore, micromagnetic measurements were carried out ex-situ at different fatigue stages and then correlated with X-ray diffraction data as well as thermographic data. Since the measured parameters correspond to the induced partial damage, they can be used in lifetime prediction methods and enable an accelerated estimation of the lifetime of materials, thus proving to be effective in monitoring the structural integrity of the material.
The publication presents quality control in the reconstruction of large components using additive processes instead of new production by casting. The investigations on an additively manufactured demonstrator component with various inhomogeneities are described. The material properties were determined and compared with the properties of an adequate cast component. Non-destructive testing methods were used to detect and evaluate irregularities. The results were verified using reference methods. The focus was on the ultrasonic phased array method. The concept and test configuration were developed through prior sound field simulations. The test results show the possibilities and limitations of the method and provide approaches for further improving the use of the UT-PA method for evaluating component quality.
This study evaluates the effectiveness of thermographic signal reconstruction using polynomial approximations for preprocessing thermographic images obtained through Pulsed Thermography in the segmentation of defects in CFRP laminates. Neural networks were trained with both processed and unprocessed images and evaluated using IoU and F1-Score metrics. The preprocessing significantly increased defects' signal-to-noise ratio and detection accuracy, achieving an IoU of 92% and an F1-Score of 98%. The technique proved effective in enhancing defect segmentation and definition, particularly in scenarios with low signal-to-noise ratio.
Reliable inspection of micro-scale components is essential in fields such as electronics and medical device manufacturing. To address this, a method combining digital speckle holography with microscopy has been developed. The resulting “Microferoscope” enables precise detection of structural defects in areas of just a few square millimetres. The method identifies minute deformations induced by controlled external stimuli, such as thermal or mechanical loading. Characteristic deformation patterns reveal structural anomalies. A dedicated module with orthogonal illumination allows for high-resolution measurement of deformations perpendicular to the component surface. Additionally, a three-dimensional module using laser diodes of different wavelengths enables the detection of both in-plane and out-of-plane deformations. Although high sensitivity in out-of-plane detection enhances defect recognition, it also increases susceptibility to environmental disturbances. This study compares both modules on micro-components to evaluate their sensitivity, robustness, and practical applicability.
This paper presents the design, modeling, and experimental validation of a non-destructive testing (NDT) system based on Eddy Current Testing (ECT) for the inspection of various conductive structures, including hollow tubes, flat plates, and complex geometries. The system integrates a motor-driven mechanical platform, a customfabricated eddy current probe, and real-time impedance measurement using an LCR meter. Its modular architecture allows the adjustment of the probe and motion configuration depending on the specimen under test. Finite Element Method (FEM) simulations were conducted using FEMM to model electromagnetic behavior and evaluate defect detectability.
Manufacturing aerospace engine components requires extensive development and rigorous testing, often involving numerous experiments and destructive analyses that lead to high costs and significant material waste. Surface Integrity (SI) in low-pressure aero-engine regions is a critical quality factor, predominantly influenced by thermo-mechanical loads during manufacturing and the prior processing history of semi-finished products. Traditionally, SI assessment relies on destructive post-production testing, limiting efficiency and resource utilization. This paper demonstrates that high-frequency eddy current technology can non-destructively quantify changes in residual stresses and phase contributions in nickel-based superalloys by calibrating electrical conductivity. Controlled variations in SI characteristics were induced using a laser heat treatment process. Calibration of high-frequency eddy current signals against these variations revealed a strong correlation between electrical conductivity and the microstructural changes associated with residual stress and phase alterations. The findings indicate that this non-destructive evaluation method is promising for real-time, in-process monitoring of SI, potentially reducing reliance on costly and wasteful destructive testing in aerospace component manufacturing.
Manganese steel turnout frogs are widely utilized in the Austrian railway system due to their excellent mechanical properties. However, their coarse grain structure presents challenges for traditional non-destructive testing (NDT) methods, such as eddy current and ultrasonic testing. Active inductive thermography offers an effective solution for inspecting such austenitic materials. This study showcases the results of a first in-field test using a mobile demonstrator designed for detecting and characterizing surface defects. Defect detection is achieved through a scanning approach, while subsequent static measurements enable crack characterization, including estimations of crack depth and penetration angle.
The increasing integration of automation and artificial intelligence (AI) in non-destructive testing (NDT) is not only changing the inspection processes themselves, but also the way decisions are made. While technical systems can reduce error-proneness and support data processing, the ultimate responsibility remains with the human. This paper examines the role of intuition in decision-making and analyses typical errors of judgment using prospect theory and insights from cognitive psychology. It also shows how well-informed decisions can be supported in AI-supported NDT processes - through training, explainable systems, user-centred design, suitable metrics, and a targeted distribution of tasks between people and technology. Rather than replacing human intuition, AI systems should be designed to complement it. To engage effectively with such systems, inspectors require not only technical expertise, but also competencies in risk assessment, probabilistic reasoning, and critical reflection on both their own judgments and the outputs provided by AI.
This paper introduces a novel robotic endoscope equipped with an optical microphone for non-destructive testing (NDT) of structural components using ultrasound guided waves. Conventional NDT often requires disassembling components, a labor-intensive and time-consuming process, especially for defects in carbon fiber reinforced polymers that are not visible externally. To address these challenges, a concept from minimally invasive medicine is transferred and a robotic prototype is developed that integrates endoscopic mobility, robotic precision and ultrasonic inspection capabilities. A test stand demonstrates the robot’s ability to navigate confined areas and perform internal ultrasonic measurements. By scanning the surface of components, the system generates full wavefield images and post-processing reveals alterations in the structure, offering insights into structural damage that traditional visual inspections may not detect. This technology is particularly promising for applications like the inspection of hydrogen storage tanks, where conventional methods are limited.
Data annotation is a mandatory step that enables machine learning methods based on supervised learning. This work introduces a pipeline that aims to reduce the annotation effort by applying data fusion of multiple data modalities e.g. machine data, process acoustics and inspection results related to robotic friction stir welding (FSW). This data is prepared for training machine learning models for assessing the weld quality based on the process acoustics acquired by acoustic emission sensors online in real-time. Non-Destructive Testing (NDT) based on light microscopy and computed tomography (CT) gather weld quality information which are used to annotate the structure-borne sound signals. This work demonstrates how the involved data modalities are put into a shared context by temporal synchronization and definition of common coordinate bases to enable transferring annotations across domains, effectively reducing the amount of annotation effort by a human inspector.
Infrared thermography is a widely recognized non-destructive testing (NDT) method used in material research and defect detection across various industrial applications. Moreover, thermography plays a crucial role in preserving cultural heritage, including historical paintings and buildings. This study focuses on the application of thermography in inspecting the historic Bücker Bü 181 aircraft, which was used in Germany during World War II. Over time, the original appearance of aircraft has often been altered as part of preservation efforts, either before or during their time in museums, leading to deviations from their historically original state. Additionally, the operational history of such objects is frequently undocumented or entirely lost, making it difficult to understand the presence of artifacts and historically significant data. These factors present major challenges in cultural heritage preservation, and destructive methods cannot be used to investigate such invaluable objects. Therefore, thermography is implemented as a non-destructive and contactless examination method. Active flash thermography combined with phase analysis is a powerful tool for evaluating multilayer systems. In this study, multiple layers of old paint on the object posed a challenge in assessing defect conditions and retrieving other critical information beneath the surface coatings. Nevertheless, pulse thermography not only demonstrated its capability to identify defects and markings in multilayered coatings but also provided insights into the internal structure and subsections of the investigated aircraft.
A reliable design of dynamically loaded components requires a thorough understanding of the fatigue properties of metallic materials. This article focuses on extending conventional methods to include non-destructive testing (NDT)-based methods, which enable a process-oriented evaluation of fatigue behaviour instead of a lifetime driven approach. The use of NDT-generated data to describe cyclic deformation curves not only enables a significant increase in information, but also a generation of virtual S-N curves with a significantly reduced number of specimens. Therefore, synergies between the two disciplines of NDT-based and conventional destructive materials testing are used to enable a better understanding of ongoing processes. The main focus lies on the development of thermographic evaluation approaches, whereby further methods such as resistance measurements are used additionally. The test materials are two unalloyed steels of grade SAE 1020 and SAE 1045, as well as the low-alloyed 20MnMoNi5-5 steel.
While the use of contrast agents in in-vitro radiology is a typical procedure in medical diagnostics, their use in industrial testing is practically non-existent. In addition to the challenging implementation and high costs, the use of contrast media in NDT, in contrary to medical applications, is also limited by a variety of alternative methods for solving the inspection task. One application in which contrast agents in the form of tracer fibers could be useful in the future are short and long fiber-reinforced plastics. Conventionally used reinforcing fibers have a very small diameter (7 to 20 µm) and a low contrast to the plastic matrix, which is why high magnifications are necessary in computed tomography (CT) imaging to resolve individual fibers. On the other hand, the fiber morphology (fiber orientation, fiber length and fiber content) has a very high influence on the component strength and stiffness. Therefore, knowledge of fiber morphology is essential for the design of fiber-reinforced plastics. Microcomputed tomography can achieve the required high resolutions for structures smaller than 10 µm, but the analyzed component sample must only be a few millimetres in size to ensure sufficiently high spatial and contrast resolution. In this work the use of tracer fibers is intended to overcome this limit in component volume. For a practical application strongly attenuating AR glass fibers are used for the investigations, which were homogeneously distributed in the plastic in addition to the reinforcing fibers during the manufacturing process. The investigations show that at a concentration of 0.1 vol-% of tracer fibers can map the fiber orientation very well and even provide sufficient information about the fiber length distribution. The article highlights the possibilities and limitations of this approach and shows how the use of these tracer fibers can lead to an increase in the field of view by the factor of 5 and also leads to a significant reduction in computing time.
