Open Access

Material defects in railway wheelset axles are exceedingly rare, yet non-destructive testing (NDT) personnel are expected to detect them reliably. Earlier work on human-related risks has shown that this low frequency of true indications limits inspectors’ exposure to meaningful signals and may contribute to performance challenges such as fatigue, reduced alertness, and skill degradation. Building on these findings, two explorative qualitative studies were conducted—one on manual and the other on automation-assisted ultrasonic testing (UT)—to examine whether the rare occurrence of signals may create conditions for cognitive underload. Both studies investigate whether, and in which parts of the workflow, cognitive underload may occur and how it manifests. Hierarchical task analyses were conducted based on literature, training materials, interviews, and field observations. Building on this, qualitative interviews were developed and analyzed to determine whether indications of cognitive underload can be identified. The findings provide first insights into underload-related risks in both manual and automation-assisted UT and suggest measures to support personnel engagement and prevent potential performance shortfalls. This work contributes to a broader research effort aimed at understanding human factors in NDT and strengthening the reliability of inspection systems in increasingly automated environments.

This article presents non-destructive testing data from six prestressed reinforced concrete sleepers of type B70. The investigation methods employed were visual inspection and impact-echo. Impact Echo testing was performed with three different actuator types, including air-coupled impact-echo actuation. For data acquisition, the same accelerometer was used for all actuators. Two of the six sleepers were factory-new. The other four were in service for several years and show serious deterioration and crack patterns. The crack types and corresponding damage classes are documented in the crack catalogues. This is the first time air-coupled impact-echo data have been published. The data consist of a crack catalogue based on visual inspection, time series of the accelerometer mounted on the top surface, and high-resolution photographs of the sleepers. This dataset is unique as it enables a direct comparison of different types of actuation on slender geometries made of concrete. Air-coupled and manual impact-hammer actuation was performed on all sleepers. Actuation using a commercial device with an electromechanically mounted impact-hammer was applied only to one intact and one damaged sleeper, as the resulting spectrograms are very similar to those obtained using manual impact-hammer actuation. The dataset enables the community to verify and reevaluate the data. This should give confidence in the new approach and motivate others to develop advanced data processing to increase prediction accuracy. Furthermore, the actuation needs just pressurised air, delivered through common air gun nozzles. Thus, other research groups and inspection companies might adopt this method and use our data or processing approach for fast and efficient comparison. This article accompanies the research article “Air-coupled impact-echo actuation applied on concrete sleepers for crack detection” published in the Journal of Nondestructive Evaluation (Springer Nature). There, the data are analysed in detail and the application of air-coupled impact-echo for the condition assessment of concrete sleepers is evaluated.

We report on using embedded fiber optic sensors to detect wire breaks and monitor traffic loads in solid bridges. These sensors provide real-time data on loads, structural behavior, and potential damage. We also develop methods to analyze large amounts of sensor data to support bridge monitoring and maintenance.

For high-strength metallic materials such as Ti-6Al-4V, failure mechanisms in the very-high-cycle fatigue regime (VHCF) between 10 8 and 10 10 load cycles differ fundamentally from those in high-cycle fatigue (HCF). Cracks usually initiate inside the component, originating from local defects or material inclusions. Recent research has demonstrated that introducing a cavity into fatigue specimens through additive manufacturing can be used to predetermine crack initiation sites in VHCF samples, thereby enabling systematic investigation of internal crack-formation processes. In this work, the theory of critical distances (TCD) is applied to numerically identify the crack initiation site on the surface of an artificially placed defect inside a VHCF sample. The computations are based on the actual geometry of the defect which is obtained through X-ray computed tomography (XCT). The predicted crack initiation site was subsequently used to define an initial crack for three-dimensional propagation simulations using FRANC3D. Since microstructural heterogeneity and crystallographic effects are not considered, the framework is not intended as a physics-based VHCF life-prediction model, but rather as an engineering approach for identifying likely crack initiation regions and predicting subsequent crack propagation from realistic internal defects. The presented explicit crack-growth simulations investigate the influence of mesh density, crack-growth increment, initial crack orientation, and kink-angle criterion on the predicted crack path. As they do not incorporate cycle-dependent crack growth, the results are intended to characterize the geometric and numerical factors governing crack propagation and to provide the ground work for future three-dimensional simulations with resolved internal defects and experimentally calibrated VHCF-specific crack-growth laws.

Polymeric components in hydrogen infrastructure face a significant challenge: the risk of damage from rapid gas decompression (RGD), which can ultimately compromise operational safety. This study investigates the impact of high-pressure hydrogen exposure on various rubber formulations: carbon black (CB) and silica (Sil) filled acrylonitrile butadiene rubber (NBR), as well as CB-filled ethylene-propylene-diene monomer rubber (EPDM). The materials were subjected to both static and dynamic high-pressure hydrogen environments. Their physical, mechanical, and chemical properties immediately after decompression and 48 h later were analysed. Our findings reveal a trend of the material to undergo tensile stiffening up to 24% after cyclic hydrogen exposure. The mechanism behind the increased stiffness is associated with a combination of microstructural mechanisms, polymer-filler interfacial rearrangement, and possible additive migration, depending on the grade. Silica-filled NBR demonstrates a faster volume recovery among all grades, but lower mechanical stability compared to its CB-filled counterparts, while the presence of plasticisers appears to accelerate crack formation under both static and dynamic loading conditions. The non-plasticised CB-filled NBR grade showed the best overall balance amongst the tested materials, particularly in tensile property retention, structural integrity, and resistance to crack formation under the applied exposure conditions. These findings contribute to the development of safer and more resilient materials for operational performance in hydrogen applications.