This presentation will discuss advantages and limitations of various NDT methods, and the technical and non-technical aspects of project approaches, to various civil/structural engineering projects from a consulting/forensic engineering perspective. These themes will be explored in the context of case studies from specialty consulting projects in buildings, bridges, movable structures, and specialty structures from energy and scientific research facilities. The case studies to be examined will include applications of NDT technologies such as acoustic emission and impulse-response, vibration monitoring, and strain and displacement monitoring.

Due to the unique "one-off" nature of these specialty consulting and forensic engineering projects, a substantial portion of the NDT effort was related to development of a rational basis for NDT method selection and application in the absence of an immediately applicable code or standard. Similarly, two-way discussion between the NDT professional and project stakeholders is required to "calibrate" the project to client needs and expectations at key points in the project. Cross-cutting themes and conclusions on the effects of technical and non-technical factors of project success in terms of problem scope, the project owner’s attitudes toward and understanding of NDT, the nature of the NDT consultant’s engagement on the project, and related factors.

As one non-destructive testing method for evaluating stresses acting on concrete, we are advancing research on techniques utilizing ultrasonic methods.
This study demonstrated that the propagation velocity in concrete measured by ultrasonic methods changes elastically in response to increasing stress during monotonic loading, and that the extent of this change is influenced by the concrete's moisture content.
Furthermore, it was demonstrated that the propagation velocity in concrete subjected to sustained loading is affected not only by elastic deformation due to the sustained load but also by creep deformation. The impact of this phenomenon on the acoustic elasticity effect can potentially be quantified using the creep coefficient.

Evaluation of grouting condition of post-tensioned tendons in Prestressed Concrete (PC) structures is an important problem in a field of maintenance of civil engineering infrastructure. Existing methods of non-destructive testing (NDT) used for grouting inspection have some limitations. For example, Impact-Echo method has comparatively low accuracy and X-Ray method is expensive and time- and labor-consuming. At the same time, there are promising alternative methods of NDT, which can be used for grouting inspection. The authors carried out an experiment using several large-scale specimens simulating PC girders. Commercially available Ultrasonic Pulse Echo Tomography device with an array of dry-contact sensors was used. Several cases of grouting condition, PC sheath diameter and geometry and rebar placement were studied. It was confirmed that amplitude of reflected ultrasonic pulse in the case of not grouted PC sheath is larger, as compared with a case of a grouted PC sheath. The authors further developed a more accurate method of evaluating grouting condition, which uses multiple parameters, such as amplitude of ultrasonic pulse reflected from PC sheath, backwall and phase of reflected ultrasonic pulse.

Carbon fiber reinforced plastic (CFRP) composites are increasingly used in many industry fields, including aerospace and automotive applications due to their high stiffness and strength performance. However, the strong anisotropy introduced by random fiber distribution makes accurate characterization and integrity monitoring of the CFRP structure challenging. Specifically, spatial variations in fiber orientation impact mechanical performance. Ultrasonic Lamb waves have been widely studied for structural health monitoring of composite plates, and in this study, the non-destructive evaluation method is proposed to assess fiber bundle distribution in CFRP plate using load-dependent ultrasonic wave propagation. Acoustic emission (AE) sensors are fixed at a specific location on the specimen, while ultrasonic waves are generated at randomly selected positions on the specimen surface. Thin rectangular CFRP plates with spatially varying fiber orientation were prepared. The plate consists of two distinct regions: an upper region with fibers aligned parallel to the tensile loading direction and a lower region with fibers oriented perpendicular to the load. One sample was subjected to tensile loading in the vertical direction, and a reference specimen was tested without an applied load. Ultrasonic Lamb waves are excited and received using surface-mounted piezoelectric transducers, and the group velocity was then evaluated. By applying and releasing mechanical load during measurement, a dataset of ultrasonic wave velocities is obtained under varying load conditions. Experimental observations reveal that ultrasonic waves propagating near the fiber bundle orientation exhibit significant velocity changes with applied load, whereas waves propagating perpendicular to the fiber direction show minimal load sensitivity. By analyzing these directional and load-dependent velocity variations, the spatial distribution of fiber bundles can be evaluated non-destructively without prior knowledge of local fiber orientation by relating lamb wave velocity, fiber orientation, and loads using equations. The proposed approach provides a practical and flexible ultrasonic-based technique for characterizing heterogeneity in short-fiber CFRP. This approach has potential applications in in-situ evaluation of large CFRP structures in future.

This study investigates the applicability of muon tomography to concrete structures with the aim of realizing infrastructure imaging using cosmic-ray muons. Owing to their exceptionally high penetration capability, cosmic-ray muons enable the visualization of deep regions in large-scale structures that are difficult to inspect using conventional non-destructive testing techniques. As “muon transmission tomography,” this capability has been demonstrated in applications such as detecting hidden chambers in pyramids and imaging internal volcanic structures. Furthermore, by exploiting Coulomb multiple scattering, “muon scattering tomography” provides high-resolution imaging and material discrimination, and its use is expanding to various fields, including the inspection of cargo containers and the internal imaging of nuclear reactors. These muon tomography techniques are expected to contribute to improved diagnosis and prediction of deterioration in aging infrastructure by enabling visualization of internal conditions, ultimately supporting more efficient infrastructure maintenance. However, applications to actual infrastructure remain limited, and neither the applicable range nor the measurement methodologies have been systematically established. In this study, concrete structures—representative of major infrastructure—are taken as the target, and key factors relevant to the application of muon tomography, including achievable spatial and contrast resolution and the required measurement time, are investigated primarily through Monte Carlo simulations.

Accurate estimation of fiber bundle breakage (FBB) in CFRP pressure vessels requires a reliable interpretation of acoustic emission (AE) signals. In this study, we propose a methodology to infer the released energy directly from AE waveforms and to estimate FBB events based on the source energy. The approach incorporates a location-based correction, in which the transfer function between the source and each sensor is adaptively adjusted according to the estimated AE source position. This enables a more accurate estimation of the original energy release, which is strongly related to FBB. To validate the transfer functions, impact tests using an air gun were conducted, generating energy release levels comparable to those of actual FBB. The experimentally obtained transfer functions were integrated into the inverse estimation framework. The results demonstrate the feasibility of energy-based fracture detection and highlight the importance of position-dependent transfer function selection for improving estimation accuracy.

Deterioration characteristics of infrastructure exhibit spatio-temporal dynamics that evolve across multiple phases, including construction, early-age behavior, service life, progressive deterioration, and post-event conditions. No single non-destructive testing (NDT) technique can adequately capture the diverse mechanisms governing these phase-dependent phenomena. This keynote introduces the concept of phased monitoring, in which appropriate NDT methods—such as acoustic emission, elastic wave techniques, tomography, fiber-optic sensing, and remote sensing—are selectively and adaptively applied at each stage. By aligning sensing strategies with the dominant physical processes in each phase, phased monitoring enables more reliable damage detection, performance assessment, and informed decision-making, thereby supporting resilient and sustainable infrastructure management.

Surface temperature simulation plays a critical role in concrete infrastructure management, serving both as a tool for structural health monitoring and thermal stress evaluation. The integration of finite element analysis with infrared thermography has proven effective in detecting surface damage in concrete structures. In particular, mass concrete structures such as dams require highly accurate surface temperature simulations for reliable thermal stress assessment. While recent approaches combine passive infrared thermography with numerical analysis to validate thermal images, significant challenges remain in accurately reproducing the temperature field in real-world dam structures. These simulations are based on heat balance equations that incorporate both the thermal properties of concrete and meteorological data. However, practical implementation faces several limitations. Key factors affecting the precision of temperature simulations include meteorological parameters (air temperature, solar radiation, wind speed) and material properties (thermal conductivity, specific heat capacity). Among these, solar radiation presents the greatest difficulty, particularly due to shadow effects caused by surrounding terrain and structures. Most existing studies have focused on cast concrete blocks under controlled conditions, leaving a critical gap in their application to actual dam structures. In such cases, mesh discretization and temporal resolution significantly influence prediction accuracy. The core challenge lies in the accurate measurement of localized environmental conditions, especially solar radiation, which is highly affected by complex and dynamic shadowing patterns that are difficult to observe and quantify.
This study addresses these limitations of conventional surface temperature simulations through two complementary approaches. First, three-dimensional laser scanning is implemented to generate high-resolution point clouds of the dam structure. This data enables precise shadow simulations. The results of shadow analysis are systematically incorporated into solar radiation calculations, improving the accuracy of surface temperature predictions. Second, a time-series deep learning model is developed using annual surface temperature datasets. Its performance is compared with that of traditional physics-based models to assess predictive capability. The study is conducted on a hollow gravity concrete dam, employing a comprehensive measurement protocol that includes meteorological monitoring, dual-mode surface temperature measurements (using both thermocouples and passive infrared thermography), and three-dimensional geometric acquisition through laser scanning. In the heat balance analysis, surface temperature is computed based on thermal equilibrium equations that consider sensible heat, radiation, and ground conduction. For shadow simulation, ray-tracing intersection judgement is conducted using triangular mesh derived from the point cloud data and solar position coordinates. This allows quantification of solar radiation reduction due to shadows. The deep learning model uses all collected meteorological variables including air temperature, humidity, wind speed and solar radiation as input features to train the time-series prediction model.
The results demonstrate that incorporating the shadow effect into heat balance analysis significantly improves simulation accuracy. This enhancement is particularly evident in reproducing temperature drops during shadow-affected periods, which are not well captured using only the original meteorological data. The deep learning framework further improves performance by effectively capturing complex behaviors, such as latent heat effects and shadow influences that are difficult to model using traditional physical approaches. Training the model with three months of autumn data achieves a Mean Absolute Error (MAE) of 1.4°C, outperforming the heat balance analysis, which yields a MAE of 1.9°C. When extended to a one-year dataset, the deep learning model achieves an even lower MAE of 1.1°C, demonstrating its scalability and robustness. Although deep learning models require long-term continuous monitoring data, the integration of shadow simulation, physical modeling, and machine learning offers a promising pathway toward high-precision surface temperature simulation. Such integrated systems are crucial for structural health monitoring and thermal stress management in concrete dam structures.

Nondestructive testing (NDT) can be formulated mathematically as an inverse problem, in which the spatial distribution and severity of internal damage are inferred from measurement data. In practical inspection settings, however, constraints such as limited sensor coverage, measurement cost, and noise often prevent the acquisition of sufficient data to uniquely determine the unknown parameters of interest. Under such conditions, Bayesian inference provides a powerful framework for probabilistic estimation. Bayesian inference—often referred to as Bayesian sensing when applied directly to measurement data—models the measurement process as a forward model to define the likelihood function, while prior knowledge is mathematically expressed through appropriate prior distributions. By combining these components using Bayes’ theorem, the posterior distribution of damage features, including their magnitude and spatial patterns, can be objectively and quantitatively characterized based on the available data. In this talk, I will introduce the fundamental concepts and mathematical formulation of Bayesian sensing in an accessible manner, and present representative applications. Examples include tomographic imaging in Earth sciences and structural assessment in civil engineering, demonstrating the versatility and practical utility of Bayesian approaches in nondestructive inspection.

In this study, we focus on the attenuation characteristics of elastic waves and develop an elastic-wave attenuation tomography method that provides high sensitivity, linear formulation, and improved computational efficiency. A sparse modeling approach is incorporated to enhance reconstruction accuracy, and simulation studies show that the proposed method clearly outperforms conventional L2-regularized inversion. To improve practical feasibility in field applications, optimal sensor configurations were investigated to reduce measurement effort while maintaining imaging quality. Moreover, since sensor contact pressure strongly affects the measured attenuation in practical settings, a correction method to compensate for this influence is introduced. These combined techniques constitute a high-performance attenuation tomography approach for concrete structures, achieving higher imaging accuracy and lower computational cost compared with traditional velocity-based tomography.

To elucidate the fracture behavior of concrete members, acoustic emission (AE) testing is commonly employed in conjunction with mechanical loading tests. In AE analysis, parameters such as average frequency, peak frequency, and centroid frequency are compared at each loading step using their histogram distributions. However, these parameters represent single-valued quantities and therefore cannot fully capture the distributional characteristics of the power spectrum. As a result, they may be insufficient for describing the underlying fracture behavior.
To address this limitation, Shannon Wavelet Entropy (SWE), which quantitatively characterizes the complexity or irregularity of AE signals, was applied to AE data obtained from a reinforced concrete (RC) column subjected to the quasi-static cyclic loading test. Because SWE reflects the degree of dispersion in the frequency bandwidth of each AE waveform, it enables a quantitative distinction between complex fracture processes in the early stages and more dominant, localized fracture mechanisms in the later stages.
Consequently, SWE allowed for statistical evaluation of the fracture behavior at each drift ratio δ. Furthermore, by expressing SWE as a scalar damage variable, the degree of stiffness degradation corresponding to each δ could be effectively demonstrated.

This study proposes a damage-tolerant design concept for composite high-pressure hydrogen vessels, assuming immediate detection of fiber failure—the critical damage mode—by an acoustic-emission–based structural health monitoring system. Residual-strength requirements for composite high-pressure hydrogen vessels were defined by adapting the No-growth approach and damage categorization methods used in composite aircraft structures. This presentation demonstrates that AE-based SHM enables the practical application of a damage-tolerant design concept to composite high-pressure hydrogen vessels.

In Japan, a significant proportion of transport infrastructure is over 50 years old, and the challenge of delivering comprehensive maintenance is further compounded by workforce shortages. Therefore, maintenance of infrastructure based on preventive maintenance is required. In preventive maintenance, it is necessary to detect damage as early as possible. To address this issue, we have developed a technology called "Bridge Deck Internal Soundness Mapping" to detect signs of deterioration before damage becomes apparent, targeting bridge decks, which are important components of bridges. This technology utilizes active elastic waves generated by moving vehicles on the road surface, which propagate through the interior of concrete bridge decks. Multiple AE (Acoustic Emission) sensors installed on the underside of the deck detect these elastic waves and their sources are located. When internal damage exists, the located source positions tend to shift or fail to be determined, making it possible to estimate the internal condition of the deck based on the distribution of these sources. To date, we have conducted demonstration tests on actual highway bridges. In this paper, aiming for practical implementation, we experimentally investigated the influence of vehicle conditions—such as driving position, weight, and speed—by installing concrete deck specimens on a test bridge. The results indicate that by appropriately correcting measurement and evaluation data obtained from real bridges, this technology can be widely and effectively applied for practical use under various bridge conditions.
Meanwhile, applications using multiple AE sensors have faced challenges due to the need for long wiring to each sensor, resulting in increased installation time and cost. To address this, we developed a wireless AE sensor system equipped with novel time estimation technology. This system employs compact wireless sensor units powered by dry batteries, contributing to improved work efficiency at bridge sites. We also report on the application of this system to field measurements on actual bridges.

The application of 3D concrete printing (3DCP) technology in the construction sector is accelerating worldwide. In Japan, its use in permanent formworks for civil engineering structures has been increasing. 3DCP has the potential to provide the solutions to social challenges such as labor shortages on construction sites, shorter construction periods, and reduced environmental impact. In 2025, the Japan Society of Civil Engineers (JSCE) published the “Technical recommendation for Concrete Structures using 3D-Printed permanent Formwork,” which summarize the fundamental principles and points of consideration in planning, design, manufacturing, construction, and maintenance of structures incorporating permanent formworks manufactured by construction 3D printers using cementitious materials. This article provides an overview of these guidelines.

The use of structures fabricated by 3D concrete printing (3DCP) has been rapidly expanding in recent years, as the technology enables formwork-free construction, geometric freedom, reduced labor requirements, and faster on-site production. Among the various 3DCP techniques, extrusion-based systems are the most widely applied due to their simplicity and compatibility with cementitious materials. However, during the layer-by-layer deposition process, the extruded filaments may not fully bond or may partially separate, leading to the potential formation of internal voids along filament interfaces. Such voids can adversely affect the mechanical performance and long-term durability of printed components. Therefore, although material design, print paths, and printing parameters should be optimized to prevent void formation, it remains crucial to establish reliable methods for quantitatively detecting the location and size of voids when their presence is suspected after printing.

In this study, wall-type structures fabricated by extrusion-based 3DCP were investigated. Artificial voids were embedded within the printed specimens, and the feasibility of detecting these voids using ultrasonic pulse velocity testing and infrared thermography was evaluated.

Although there are limitations regarding the smallest detectable void dimensions, the results demonstrate that voids above a certain size can be successfully identified using these non-destructive testing methods.

To realize the practical application of cement-based 3D printed concrete (3DPC), which is expected to reduce labor and eliminate formwork as the additive manufacturing process in construction and building, it is necessary to address deformation of the extruded material from nozzle of printer caused by self-weight and layer-by-layer deposition in fabricated layered structures. Predicting such deformation and designing the object surface geometry accordingly will be essential. In this study, estimation model for the buildable layered structure, which depends on viscoelastic material property of the extruded material, is investigated based on the deformation due to the loading given in printing of each layer. In addition, cement-based materials exhibit changes in their material properties over time. Therefore, these deformations evolve as progresses of layering steps. It is unavoidable to ensure the geometrical information in three-dimensions of the layered structure by taking into account the deformations caused by self-weight and the load from subsequent layers over time. Thus, 3DPC printable requirement for changing mix proportion and printing condition is evaluated using non-destructive and micro-destructive testing of extruded material at various mortar hardening processes.

The fundamental challenge of additive manufacturing techniques for cementitious materials, known as 3D concrete printing (3DCP), is the anisotropic mechanical behaviour caused by the layer-by-layer printing and deposition process. This process leads to potentially weak interfaces between successive filaments. Considering that these interfaces are distributed internally, applying innovative non-destructive testing (NDT) techniques is essential to characterise the complex fracture behaviour. The present study applies an integrated analysis utilising Acoustic Emission (AE) and Digital Image Correlation (DIC) to investigate the anisotropic crack kinematics of 3D printed concrete. We examined specimens printed with different printing paths, resulting in the fundamental fracturing mode, namely trans-layer fracture and inter-layer fracture. During three-point bending tests on notched beams, AE captured microcracking and DIC tracked surface strain distribution. The experimental work revealed that despite the good printing quality, it acts as a localised weak zone. Specifically, the integrated analysis demonstrated the difference in the crack initiation resistance between trans-layer fracture and inter-layer fracture. These results provide fundamental insights into the mechanical anisotropy, highlighting the efficiency of NDT approaches for developing reliable design concepts for future infrastructure.

3D Concrete Printing (3DCP) has progressed rapidly over the past decade, evolving from an experimental fabrication technique into a construction technology that is being increasingly implemented worldwide. Alongside ongoing advances in material rheology and process control, growing attention has been directed toward structural performance, constructability, and engineering reliability under practical conditions.

This keynote lecture reviews recent developments in 3DCP from a structural engineering perspective, with particular emphasis on civil engineering applications. In many recent projects, 3D printing has been adopted not to directly fabricate fully load-bearing members, but to produce cementitious components that function as permanent formwork integrated into reinforced concrete structures. This strategy enables improved productivity while preserving consistency with established structural design concepts and construction practices.

From an engineering viewpoint, 3D-printed permanent formwork introduces challenges that differ fundamentally from those of conventional reinforced concrete. Layered fabrication inherently leads to anisotropic material properties, while the completed structure behaves as a composite system consisting of printed elements, cast-in-place concrete, and multiple interfaces. Furthermore, manufacturing and construction processes have a direct influence on structural performance, requiring these processes to be explicitly considered within structural planning and design.

To systematically address these issues, the Japan Society of Civil Engineers (JSCE) published technical recommendations in August 2025 covering structural planning, design, manufacturing, construction, quality control, inspection, and maintenance of concrete structures using 3D-printed permanent formwork. This presentation introduces the key concepts underlying these recommendations and discusses how accumulated practical experience has been translated into an engineering framework.

The lecture concludes by outlining future directions for 3DCP as a structural system, emphasizing the need for integrated consideration of design, construction, and performance to enable safe, reliable, and scalable applications in structural and infrastructure engineering.

A distinct approach to 3D printing construction has been introduced in Thailand through the development of hollow-section printed walls combined with internal truss members. This strategy reduces structural weight, minimizes material consumption, and lowers CO₂ emissions while maintaining structural efficiency. To support the safe design of such systems, extensive investigations have been conducted to clarify the behavior of printed structures, with particular emphasis on the interfaces between layers that govern the anisotropic response of 3D-printed elements. Two series of tests were carried out to determine the interface normal and shear strengths of printed mortar under varying printing interval times, initial setting times, mortar strengths, ages, and temperatures. Predictive equations derived from these results provide improved accuracy for finite element analysis of 3D-printed structures.
Beyond material-scale characterization, structural performance has been examined through testing of 3D-printed wall components. The influence of wall geometry on axial behavior has been demonstrated, with different printing patterns producing distinct stress-transfer paths that affect cracking, load-carrying capacity, and ductility. Among the tested patterns, the diamond wall exhibited the most favorable performance, combining higher first-crack load, improved ductility, uniform stress distribution, shorter printing time, and reduced mortar use. Full-scale tests on walls with large openings further highlighted the role of reinforcement strategies in stiffness, cracking behavior, and failure modes. Additional studies on walls with internal trusses and grouted cores under eccentric compression revealed complex strain behavior governed by local topology variations and core stiffness.
A numerical modeling approach has been developed in which isotropic mortar properties are used while printed layers are explicitly modeled with tiebreak contact interfaces to simulate separation and sliding. Recent numerical studies on seismic performance also indicate promising behavior for single-V internal bracing patterns. In summary, these findings support the advancement of reliable, efficient, and sustainable 3D-printed construction in Thailand.