Wire Arc Additive Manufacturing (WAAM) is a metal Additive Manufacturing (AM) technique that can produce fully dense metallic structures with virtually no porosity and at high productivity, compared to other currently available AM techniques such as Laser Powder Bed Fusion (L-PBF). As development of the technique is still ongoing, monitoring or post-fabrication inspection methods are under active investigation. In this work, we apply Resonant Ultrasound Spectroscopy (RUS) to samples fabricated from two different wires (construction steel and austenitic stainless steel) and quantitatively characterize isotropic and anisotropic elastic behaviour of the obtained dense parts. We find that an isotropic elastic model fits the construction steel samples well. For the 316 L polycrystal however, the isotropic elastic model is unsatisfactory, and an effective orthotropic elastic model is found to fit the resonance data. EBSD and XRD measurements are used to confirm and explain this difference in elastic behaviour between steel grades by the presence of a strong texture in the 316 L samples. Additionally, the texture data measured by EBSD are used to infer single crystal constants from the polycrystal resonance data using the Hill averaging scheme for one of the 316 L samples. We end by discussing the differences between the two elastic models used in the study (orthotropic and texture based) as well as the link between the measured resonances and microstructural descriptions of the samples.

In recent years, Resonant Ultrasound Spectroscopy (RUS) has been extensively applied to objects produced by additive manufacturing to characterize elastic material properties, detect defects or geometrical deviations. In this talk, we analyze samples that were produced using the wire-arc additive manufacturing (WAAM) process using different grades of steel wires. Resonance spectra were obtained and allowed to classify samples as either elastically isotropic or anisotropic. Detailed investigation on anisotropic samples (produced with 316L wire) under an orthotropy hypothesis showed that the samples were markedly softer along the layer deposition direction. Subsequent investigation using EBSD confirmed the results obtained with RUS. They also allowed to quantitatively model the elastic constants using the Voigt-Reuss-Hill averaging theory, which were in good agreement with the ones obtained using the RUS inverse problem.

The mechanical behaviour of 316L stainless steel single crystal is characterised at room temperature and 300°C. Elasticity moduli at room temperature are obtained with resonant ultrasound spectroscopy. Their dependence on temperature is calibrated with molecular dynamics simulations. The plastic behaviour is characterised by tensile tests on millimetre-sized single crystal specimens and compression tests on micrometre-sized single crystal specimens. A constitutive model of crystal plasticity based on dislocation density hardening at finite strains is developed and implemented in an open-source material subroutine compatible with several finite element (FE) and fast Fourier transform (FFT) solvers. Tensile curves at room temperature and 300°C are used to calibrate the interaction coefficients for self and coplanar dislocation interactions. The dislocation mean free path for obstacle dislocations and the annihilation distance are also calibrated. The calibrated model predicts tensile curves in excellent agreement with experimental data. In addition, the predicted plastic strain fields are in good agreement with the experimental fields obtained by digital image correlation. Semi-quantitative agreement between simulation and experimental data is obtained for micro-compression tests without further calibration of the model. Finally, an extension to polycrystals with grain size effects is finally proposed. The predicted strain hardening behaviour is compared with experimental data on stainless steel polycrystals.

Grade2XL is a European project funded by the H2020 program that gathers 21 academic and industrial partners. Started in March 2020 for 4 years, the principal objective is to print extra-large structures using the WAAM (Wire Arc Additive Manufacturing) method with a complete control of the fabrication process. Since WAAM is a relatively new technology for printing large specimens, it is important to develop quality assurance methods that can be used during the fabrication and once the structure has been built, which is the focus of CEA List in the project. Regarding methods applied built structures, CEA List has investigated Resonant Ultrasound Spectroscopy (RUS) for the material characterization and more specifically the evaluation of the elastic constants or more conventional ones for the inspection with phased-array ultrasonic techniques (PAUT) or eddy current (EC) for the detection of indications are implemented and optimised them for the inspection of the final structures. Concerning the development of in-line monitoring methods, CEA List has been working on laser Doppler vibrometry (LDV) and X-ray fluorescence (XRF) spectrometry, in order to get structural information during the printing but also detect the appearance of abnormal events and correlate them to the appearance of defects generated during the WAAM process. This proceeding presents the advantages of those methods for the inspection of extra-large structures made by WAAM and discusses the first results obtained in the framework of the project. Keywords: Additive Manufacturing, NDT, Imaging techniques, Eddy current, In-line monitoring, laser vibrometry, X-ray fluorescence spectrometry, Resonant Ultrasound Spectroscopy.

Dans la perspective d’améliorer les procédés d’imagerie ultrasonore sur des assemblages soudés, les caractéristiques mécaniques effectives du matériau et leur variabilité au sein de la soudure sont une donnée nécessaire. De nombreuses techniques reposant sur l’utilisation d’ondes élastiques ultrasonores ont déjà été mises en oeuvre afin d’obtenir avec précision les constantes élastiques et ainsi d’évaluer l’anisotropie apparente de ces milieux complexes. Dans le cadre de cette étude, la Spectroscopie de Résonance Ultrasonore (RUS) a été mise en oeuvre pour caractériser des échantillons prélevés dans un « mur » fabriqué par soudage dans l’hypothèse d’un comportement élastique orthotrope, homogène à l’échelle de l’échantillon. Les fréquences de résonance mesurées ont permis de résoudre le problème inverse à l’aide de méthodes d’optimisation utilisant le gradient et d’obtenir les constantes élastiques pour chacun des échantillons. Les résultats sont comparés à ceux préalablement obtenus par l’utilisation d’une technique de mesure de vitesse de phase en ondes ultrasonores pulsées dans le cadre du projet ANR MUSCAD.

Les procédés de fabrication additive par fusion laser sur lit de poudre (FLLP) permettent aujourd’hui de fabriquer des composants métalliques à géométrie complexe, aux propriétés performantes. Parmi les différents paramètres impliqués dans le procédé, la température joue un rôle fondamental, car elle contrôle la fusion de la poudre, la solidification et la formation de la microstructure à partir du bain liquide. Différentes techniques existent pour mesurer la température à la surface d’une pièce, comme des mesures par pyromètres ou caméra infra-rouge. Cependant, le champ de température interne reste difficile à estimer par ces méthodes conventionnelles. L’objectif de nos travaux est d’étudier l’évolution du champ de température d’un objet en cours de construction, en proposant une technique de suivi in situ, basée sur la sensibilité des ondes élastiques à la température du milieu de propagation. Un dispositif expérimental a été développé afin de mesurer simultanément des temps de vol en impulsion-écho pendant l’élaboration d’une pièce cylindrique ainsi que températures à l’aide de thermocouples. Un modèle thermique par éléments finis a été développé afin de corréler les variations de temps de vol et de température observées au cours de la fabrication. Dans ce papier, la technique ultrasonore proposée ainsi que les mesures expérimentales réalisées sont exposées. Le modèle et la confrontation de ses résultats aux données expérimentales sont présentés.

The objective of the present work is to investigate the potential of Resonant Ultrasound Spectroscopy (RUS) as an innovative technique for the nondestructive analysis of Laser Powder Bed Fusion (LPBF) AlSi7Mg0.6 parts. The acoustic resonance measurements are tested against an experimental database that covers a broad range of process parameters and large variations of the standard volumetric energy density. Two other nondestructive techniques are used to assess the potential of the RUS measurements of additively manufactured samples: the easy-to-use Archimedes density measurement and the cost-intensive computerized X-ray tomography. Our results show that RUS provides both qualitative and quantitative insights that allow the detection of the amount of lack of fusion porosities and the estimation of the elastic properties of the fabricated samples. Quantitative correlations between the three nondestructive testing methods are obtained, hinting at how RUS could be used effectively for systematic post-production testing of LPBF samples.

Accurate defect characterization is desirable in the ultrasonic nondestructive evaluation as it can provide quantitative information about the defect type and geometry. For defect characterization using ultrasonic arrays, high-resolution images can provide the size and type information if a defect is relatively large. However, the performance of image-based characterization becomes poor for small defects that are comparable to the wavelength. An alternative approach is to extract the far-field scattering coefficient matrix from the array data and use it for characterization. Defect characterization can be performed based on a scattering matrix database that consists of the scattering matrices of idealized defects with varying parameters. In this article, the problem of characterizing small surface-breaking notches is studied using two different approaches. The first approach is based on the introduction of a general coherent noise model, and it performs characterization within the Bayesian framework. The second approach relies on a supervised machine learning (ML) schema based on a scattering matrix database, which is used as the training set to fit the ML model exploited for the characterization task. It is shown that convolutional neural networks (CNNs) can achieve the best characterization accuracy among the considered ML approaches, and they give similar characterization uncertainty to that of the Bayesian approach if a notch is favorably oriented. The performance of both approaches varied for unfavorably oriented notches, and the ML approach tends to give results with higher variance and lower biases.

This paper presents a collaborative work that started out under the framework of the Generation IV International Forum regarding ultrasound telemetry in sodium-cooled fast reactors. The liquid sodium environment imposes harsh constraints on immersed transducers for telemetry. An alternative approach consists in locating a transducer outside of the reactor core and exciting guided waves that are conducted by a waveguide inside the liquid sodium from where they radiate inside the reactor and can be used for telemetry and associated inspections. This work presents a numerical method coupling the propagation of elastic guided waves in long waveguides with a liquid environment and applies it to conduct a parametric study on several waveguide designs. The simulated results are compared to data from real waveguides obtained in a water tank, carried out with 9 sensors, in order to cover a large design space and ensure the validity of the modeling approach. Good agreement is observed with respect to the main measured quantities. The limits of the developed model are discussed with the help of these results. In particular, variations linked to untracked experimental parameters are shown to have an impact on results.

In Guided Wave (GW) Structural Health Monitoring (SHM), a baseline, i.e. a set of measurements taken on the inspected structure in a pristine state, is often required to separate the contributions of the defect(s) from the other propagating wave packets. Due to the sensitivity of GWs to Environmental and Operational Conditions (OECs), most GW-SHM techniques are limited in terms of range of applicability [1]. More specifically, the inspection of the unknown state must be conducted under the same EOCs as the ones of the pristine state. Besides measuring baselines on the structure under all the EOCs of interest, which is prohibitively expensive, a potential solution is to compensate the EOC effects on the measured signals. Several solutions in the literature, such as Baseline Signal Stretch [2] and Dynamic Time Warping [3], have been proposed to solve this problem, but are somewhat limited in terms of amplitude of compensation or range of application. In this paper, a model-based machine learning procedure to compensate the measured signals in an unknown state and known EOCs is presented. The compensation model is trained on experimental data acquired at various temperatures on a structure representative of the one of interest. In other words, an experiment under various EOCs must be conducted on a simplified version of the structure with at least two transducers. Material and transducers of the simplified experiment must be identical to the ones of the real structure, but the actual geometry might differ. The compensated signals are then used to conduct guided wave imaging, allowing immediate defect detection and localization. Results are shown for both aluminum and composite panels.

In this work, we assess the inversion performance in terms of crack characterization and localization based on synthetic signals associated to ultrasonic and eddy current physics. More precisely, two different standard iterative inversion algorithms are used to minimize the discrepancy between measurements (i.e., the tested data) and simulations. Furthermore, in order to speed up the computational time and get rid of the computational burden often associated to iterative inversion algorithms, we replace the standard forward solver by a suitable metamodel fit on a database built offline. In a second step, we assess the inversion performance by adding uncertainties on a subset of the database parameters and then, through the metamodel, we propagate these uncertainties within the inversion procedure. The fast propagation of uncertainties enables efficiently evaluating the impact due to the lack of knowledge on some parameters employed to describe the inspection scenarios, which is a situation commonly encountered in the industrial NDE context.

This paper presents a global strategy aiming at solving efficiently inverse problems classically faced in the community of non-destructive testing, like flaw characterization and sensors optimization, for instance. The approach is based on intensive use of simulation tools that are obtained from complex physical models developed at CEA LIST and dedicated to ultrasonic and electromagnetic testing applications. Results presented in different physical contexts show not only its efficiency to solve parametric estimation problems, but also its capability to account for variability due to uncertainty in some model parameters and to evaluate the so-called well-posedness of inverse problems considered.

Models for the simulation of ultrasonic inspections of flat and curved plate-like composite structures, as well as stiffeners, are available in the CIVA-COMPOSITE module released in 2016. A first modelling approach using a ray based model is able to predict the ultrasonic propagation in an anisotropic effective medium obtained after having homogenized the composite laminate. Fast 3D computations can be performed on configurations featuring delaminations, flat bottom holes or inclusions for example. In addition, computations on ply waviness using this model will be available in CIVA 2017. Another approach is proposed in the CIVA-COMPOSITE module. It is based on the coupling of CIVA ray-based model and a finite difference scheme in time domain (FDTD) developed by AIRBUS. The ray model handles the ultrasonic propagation between the transducer and the EDIT) computation zone that surrounds the composite part. In this way, the computational efficiency is preserved and the ultrasound scattering by the composite structure cm be predicted. Alternatively, a high order finite element approach is currently developed at CEA but not yet integrated in CIVA. The advantages of this approach will be discussed and first simulation results on Carbon Fiber Reinforced Polymers (CFRP) will be shown. Finally, the application of these modelling tools to the construction of metamodels is discussed.

The in-service inspection of the sodium-cooled fa st reactor prototype ASTRID requires substantial R&D efforts for selecting, dev eloping and qualifying ultrasonic techniques and tools. Several ultrasonic transducer s have been developed and tested: the TUSHT model by the CEA, the TUCSS model by AREVA NP /INTERCONTROLE (both using piezoelectric materials), and electromagnetic acoustic transducers (EMAT) by the CEA. Each type of transducer has several advantages and disa vantages regarding the ultrasonic applications considered for ASTRID: transducer loca ti n, ultrasonic techniques (“contact” or “immersion”, bulk or guide waves, etc.), temperatur e, and the type of measurement (telemetry/vision or defect detection). This articl e describes the development and qualification programmes that are currently underway. This progra mme aims at ultimately selecting the most appropriate transducer for each ASTRID ultraso nic application. This involves tests on both simple and elaborate targets (with defects to be detected or specific shapes to be measured/seen), together with tests in water at roo m temperature and in sodium at 200°C. Simulation tools have also been improved by develop ing the CIVA software platform.

The present work describes a non-destructive testing method aimed at securing high manufacturing quality of the innovative compact heat exchanger developed under the framework of the CEA R&D program dedicated to the Advanced Sodium Technological Reactor for Industrial Demonstration (ASTRID). The heat exchanger assembly procedure currently proposed involves high temperature and high pressure diffusion welding of stainless steel plates. The aim of the non-destructive method presented herein is to characterize the quality of the welds obtained through this assembly process. Based on a low -frequency model developed by Baik and Thompson [1], pulse-echo normal incidence measurements are calibrated according to a specific procedure and allow the determination of the welding interface stiffness using a nonlinear fitting procedure in the frequency domain. Performing the characterization of plates after diffusion welding using this method allows a useful assessment of the material state as a function of the diffusion bonding process.

This paper describes the development of an in-house phased array EMAT transducer for longitudinal wave inspection in liquid sodium. The work presented herein is part of an undergoing project aimed at improving in-service inspection techniques for the ASTRID reactor project. The design process of the phased array EMAT probe is briefly explained and followed by a review of experimental test results. We first present test results obtained in the laboratory while the last part of the paper describes the liquid sodium testing and the produced ultrasound images.

Sodium cooled Fast Reactors (SFR) use liquid sodium as a coolant. Liquid sodium being opaque, optical techniques cannot be applied to reactor vessel inspection. This makes it necessary to develop alternative ways of assessing the state of the structures immersed in the medium. Ultrasonic pressure waves are well suited for inspection tasks in this environment, especially using pulsed electromagnetic acoustic transducers (EMAT) that generate the ultrasound directly in the liquid sodium. The work carried out at CEA LIST is aimed at developing phased array EMAT probes conditioned for reactor use. The present work focuses on the experimental validation of a newly manufactured 8 element probe which was designed for beam forming imaging in a liquid sodium environment. A parametric study is carried out to determine the optimal setup of the magnetic assembly used in this probe. First laboratory tests on an aluminium block show that the probe has the required beam steering capabilities.

Ultrasonic guided wave techniques are seen as suitable candidates for the inspection of welded structures within sodium cooled fast reactors (SFR), as the long range propagation of guided waves without amplitude attenuation can overcome the accessibility problem due to the liquid sodium. In the context of the development of the Advanced Sodium Test Reactor for Industrial Demonstration (ASTRID), the French Atomic Commission (CEA) investigates non-destructive testing techniques based on guided wave propagation. In this work, guided wave NDT methods are applied to control the integrity of welds located in a junction-type structure welded to the main vessel. The method presented in this paper is based on the analysis of scattering matrices peculiar to each expected defect, and takes advantage of the multi-modal and dispersive characteristics of guided wave generation. In a simulation study, an algorithm developed using the CIVA software is presented. It permits selecting appropriate incident modes to optimize detection and identification of expected flawed configurations. In the second part of this paper, experimental results corresponding to a first validation step of the simulation results are presented. The goal of the experiments is to estimate the effectiveness of the incident mode selection in plates. The results show good agreement between experience and simulation.