The MOSAICS project (Modelling of an austenitic weld inspected by ultrasonic testing) aims at developping modelling codes in order to improve the predication of the ultrasonic propagation in austenitic welds and to help to an inspection diagnosis. The welded components of EDF and DCNS equipments are subject to mechanical and thermal stresses that can generate in-service defects. Therefore Regulatory requirements enforce a volumetric inspection of these components. The ultrasonic (UT) techniques are advocated for their ability to simultaneously detect and size defects. However, for this type of material, the existing industrial processes are neither as effective nor as reliable as is needed. In fact, the multi-pass austenitic welds are characterized by anisotropic and heterogeneous structures which lead to numerous disturbances of the ultrasonic beam. So far, studies have been limited to 2D configurations and to one specific welding process (SMAW). However new industrial applications, with increasing complexity in terms of geometry and material (various welding processes), are emerging . As a consequence, there is a great need for taking into account complex 3D configurations in order to understand the wave/microstructure interaction and then to optimize the UT inspection configurations. . Ultimately, the goal is to provide key elements for decisions relating to the integrity of high-risk structures. The development of several complementary codes (finite elements and semi-analytical models, hybrid model combining the two approaches) will achieve, at all stages of development, a numerical benchmark on perfectly controlled configurations. This task will be supported by all project partners. Moreover, the increasing requirements in terms of inspection performances necessitate an accurate assessment of all influencing parameters in order to reproduce as precisely as possible the propagation phenomena. As the material description is an essential input data, a task of the project will be specifically dedicated to a multiscale welds characterization approach. Various innovative techniques developed by the academic partners of the project will be used in this task. In particular, an ultrasonic measurements device has been designed in order to simultaneously determined the elastic and attenuation coefficients of the wave in an heterogeneous material. Furthermore, an inversion process based on ultrasonic data has been developed to optimize a weld description firstly derived from a phenomenological model. The last task of MOSAICS will be dedicated to the validation of the codes through a direct comparison with experimental results. . Reference data will be acquired using advanced experimental technologies, including phased-array probes,applied on representative mock-ups containing calibrated defects. For all these reasons, MOSAICS is an industrial research project proposed in parts 1 (Modelling and simulation of complex systems) and 5 (Uses and Applications) of the Digital Models Program. The challenges of these developments for the industrial partners concern both the safety, the availability and the lifetime of their facilities. The results will be directly usable in the qualification process of the ultrasonic techniques. The valuation of the project will largely come from the integration of developments in the CIVA software platform and its distribution by the EXTENDE company.

When a road accident happens, the structure of the vehicle is designed to absorb a part of the energy released by the shock, the objective being to protect as much as possible the occupants of the vehicle. The quality of the assembly, and more particularly the spot weldings, is a key stake to ensure the absorption of the energy by deformation of the parts and rupture of fusible points. However, the irregular geometry of the profile and the bad quality of the surface of the spot weldings make particularly difficult the application of inspection methods. Moreover, the automotive industry is subject to a high competitiveness, which in the context of respect of the environment, results in technological breakthroughs as for example the structure lightening of the vehicle and the development of the electric vehicles. These new stakes must be realized by maintaining a high level of safety of the vehicles. ASAP will answer these challenges context by developing a new methodology in non destructive testing allowing an optimized monitoring of the assemblies. The objectives of ASAP will be to study and develop a new phase-array ultrasonic method. One of the innovations of the project will be to implement within this instrumentation dedicated algorithms allowing to adapt in real-time to the complex geometry of the spot welding, to improve the performances of detection, and to have an automatic interpretation of the quality of the spot welding without need for expertise . In addition, a big part of the project will be devoted to the development and the fiabilisation of a macroscopic model of dynamic behavior of the spot weldings. This model coupled with the simulation tools of the non destructive testing method will make possible the validation and the optimization of the methodology for a large range of characteristic parameters of the assemblies and spot weldings. ASAP will gather the industrial RENAULT, two SMEs EXTENDE and M2M, and two research laboratories CEA LIST and LAMIH. The expertise of these partners in complementary fields of research will make possible to lead at the end of the project to the creation of the first apparatus of the market ensuring reliable, fast and simple evaluation of the quality of the spot welding.

The MUSCAD project falls under the SNR orientation “Sensors and Instrumentation” (orientation 15) and the “materials and processes axis” (Axis 3) attached to the challenge “Stimulate Industrial Renewal” (Challenge 3). This is a Collaborative Research Project - Enterprises (ESRP), which brings together public laboratories of INSA Lyon, Aix-Marseille University (AMU) and CEA, an industrial partner, EDF-R&D and the company EXTENDE. Non Destructive Testing (NDT) brings together all the techniques, tests, exams, implemented during manufacture or maintenance, to ensure the integrity of a manufactured product or an industrial plant. These techniques aim at detecting the presence of possible defects estimated critical to the safety and / or proper operation of the inspected component. The NDT methods are part of the industrial tools that help to improve productivity and help to ensure product compliance with quality and safety requirements. NDT techniques based on ultrasonic elastodynamic waves are very widespread in the nuclear industry in particular because they allow control in the component volume. However, complex and heterogeneous structures (coarse grained steels, welding, bimetallic assembling ...) are commonly found in the cooling systems of nuclear reactors pressurized water (PWR) used in all French nuclear power plants. The heterogeneity of these materials degrade performance of controls because of many scattering effects (attenuation, structural noise, dispersion) undergone by the ultrasonic wave. In addition, local variations in their mechanical properties (grain size, crystallographic orientation ...) lead to differences of response according to the inspection position. To assess the reliability of the inspection techniques used on such components, simulation is increasingly used but requires an accurate knowledge of the inspected structure properties. The MUSCAD project aims to develop new complex media characterization tools, in particular those including a heterogeneous and anisotropic microstructure, by solving inverse problem applied on ultrasonic measurements. The objective of this approach is to evaluate the mechanical properties of these materials directly usable as an input of simulation tools in order to improve diagnosis and the controllability of the components. According to used mechanical parameters, the model sensitivity will be quantified on a case of realistic representative study of a weld control. This coupled approach including characterization technique and simulation tools should enable us to make progress on understanding of propagation phenomena and ultrasonic scattering in austenitic welds. We seek to quantify more clearly the influence of the weld structure on the performance of non-destructive ultrasonic testing. The key strategic axes of the project are: - The development of simulation codes dedicated to the modeling of ultrasonic beam propagation in damping and anisotropic materials in order to simulate experimental setup of non-destructive evaluation (NDE) techniques. - The development of ultrasonic characterization tools developed by INSA to more reliable measurements by determining a confidence interval on identified parameters. - The study of the sensitivity of an inspection method to mechanical parameters on a realistic case of ultrasonic NDT. The results of studies conducted in the project will be directly used in the assembly control processes qualification files. The integration of developments in the CIVA software platform opens perennial prospects for new acquisition and inversion strategies . The dissemination of the work will be performed by EXTENDE distributor of the CIVA software.