The crucial part of nuclear waste storage is the construction of sealing structures made of reliable, safe and well–understood materials. We present an extended analysis of long-term multi–sensory monitoring and non–destructive testing (NDT) inspection of two laboratory specimens aiming at potential materials for sealing structures for nuclear waste repositories. Specimens with a volume of 340 litres made from newly developed alkali–activated materials (AAM) and established salt concrete (SC) were analysed using embedded acoustic emission and wireless radio-frequency identification (RFID) sensors, ultrasonic echo imaging, active thermography, and X–ray computed tomography. The monitoring analysis showed lower heat of reaction and 50% less acoustic emission events in AAM compared to SC. However, due to the surface effects of the AAM material, the number of acoustic emission events increased significantly after approximately two months of monitoring. Subsequently performed NDT inspections reliably located embedded sensors and confirmed the absence of major cracks or impurities. The presented laboratory results show the feasibility and potential of comprehensive NDT monitoring and inspection to characterise cementitious and alternative materials as well as the need for multi–parameter long–term monitoring. Thus, our study demonstrates that tailored NDT investigations will help to develop safe sealing structures for nuclear waste repositories.

Corrosion of steel components embedded in concrete is detrimental to the structural integrity and their intrinsic function, such as the leak-tightness of containment liners in nuclear reactor containment buildings. Detection of embedded liner corrosion is a challenging task due to the concrete’s heterogeneity and the large dimensions of containment buildings. This initial study explores early-stage detection using nonlinear wave modulation on corroded liner plates embedded inside small lab-scaled concrete cylinder specimens. The liners were corroded before being embedded in concrete to minimize the risk of concrete cracking, allowing for a focused exploration of early-stage detection capabilities. Impact-based modulation and Luxembourg-Gorky type cross-modulation techniques were employed. Impact-based modulation uses a hammer impact to introduce a strong perturbation on a weak probe signal, analyzed with the Hilbert-Huang transform. Cross-modulation uses a second amplitude-modulated ultrasound to transfer modulation onto the weaker probe, analyzed via spectral analysis. Initial results indicate that impact-based frequency modulation is not highly sensitive to corrosion products alone and may not effectively identify specimens without concrete defects. However, impact-induced amplitude modulation is more responsive to corrosion and cracks in concrete. Cross-modulation shows strong sensitivity to crack-like defects and embedded severe corrosion but may not be optimal for early-stage corrosion detection due to the presence of inherent concrete cracks. Further research is needed to evaluate nonlinear ultrasonic techniques applied to larger concrete specimens and under realistic conditions before making practical recommendations.

Impact Echo (IE) simulations with heterogeneous concrete models have been very rare in the past. However, they could prove to be a valuable tool for enhancing physical understanding and improving data evaluation through artificial intelligence for future IE testing. Such improvements would result in more accurate condition assessments in industrial applications, such as for the inspection of nuclear power plant containment structures, among others. This study presents and discusses parameters influencing the outcome of IE simulations using the elastodynamic finite integration technique (EFIT). These parameters include the effective wave speed in concrete, which depends on the degree of heterogeneity. Additionally, the study explores wave scattering in the Rayleigh limit, identified as a primary noise source during actual IE measurements and discusses approximations for the source function, which ensure stability of the simulations. Lastly, the damping properties of the simulated concrete are discussed, recognizing the challenges in simulating elastic wave damping in low-frequency regimes. Following the discussion of these influencing parameters, the study presents results from 2D and 3D IE simulations of a test specimen, comparing them to real inspection data. Generally, results from 3D simulations exhibit a good agreement with measurement data in various aspects, whereas applications for 2D simulations are more limited yet also valuable.

In recent decades, non-destructive testing (NDT) has become an indispensable component of condition assessment in various industries such as civil infrastructure, automotive, nuclear, and other safety-related sectors. Each NDT method possesses unique strengths and weaknesses, prompting increased interest in combining these methods for comprehensive assessments. However, the full potential of NDT data remains underutilized due to a lack of standardization in estimating NDT reliability. Often, the capabilities of NDT systems for specific inspection tasks are unknown, hindering advancements in NDE 4.0, where knowledge about the reliability of non-destructive tests is crucial. For example, such information is very valuable when labeling data for machine learning purposes to define whether a measurement can be considered sound or defect. This paper underscores the pressing need for international standardization in the reliability assessment of NDT and presents the German approach. This standard outlines procedures for NDT evaluation, including (i) defining the application, (ii) developing and producing test specimens, (iii) statistically evaluating test results (iv) considering human factors, and more. By doing so, the standard ensures the comparability of reliability assessments and enhances their overall acceptance. This study specifically delves into the discussion of reliability evaluation using a model-based approach for ultrasonic inspection of concrete structures, both with and without reinforcement, and under different defect conditions. In this case, honeycombs are chosen as the investigated defect type. Furthermore, an initial case study applies a Probability of Detection (POD) analysis for ultrasonic testing (UT) data obtained from a concrete specimen with artificial defects. This study represents a pioneering effort for UT in the context of concrete structures.