In the ACODE project, two strategic facets, advanced metallurgy and innovative deposition processes combine in order to improve the performances and lifetimes of nuclear components immersed into extremely aggressive environments. ACODE aims firstly at establishing a lab-scale methodology for selecting and identifying new material compositions that belong to the family of high entropy alloys. It consists of the combinatorial synthesis of thin films allowing for the exploration of a wide range of compositions in a limited number of experiments. The compositions identified in the first step with functional characterizations regarding nuclear applications will be deposited in a second step at a pilot scale on prototype components. In order to evaluate the protection granted by the coatings on prototype components in a more relevant way, they will be tested in more representative conditions. The selected thin films deposition process belongs to the last generation of ionized PVD (Physical Vapor Deposition) technologies. This allows for the growth of coatings with properties tailored to protection in extreme environments (density, homogeneity, interfacial quality…).
PROVIDIA aspires to optimize thin film deposition processes with artificial intelligence through several CEA specific applicative studies that also concern industrial sectors. The optimization task is indeed often hindered by complex and interconnected coating functional properties. Thus, obtaining desired functionalities from a deposition process relies mainly on an empiric development supported by physical and chemical characterizations along with performance evaluation run by experts in a laboratory. We will apply a knowledge-based method coming from two complementary sources in order to optimize these processes by artificial intelligence: a formalization of experts’ knowledge on the one hand and knowledge acquired from data, i.e. machine-learning, on the other hand. An extensive experimental effort should produce enough data to identify and highlight not only links between deposition parameters and the characteristics of the materials coated but also the influence of material characteristics and the performances of coated components.
Depositing a thick coating on a substrate might generate several issues (porosity, micro-cracks, residual stress…) increasing the fragility and the susceptibility of the coated system to degradation phenomena such as corrosion or spalling. When one material exhibits refractory behavior while the other is volatile in the same conditions, it results in the impossibility to perform the coating deposition by the usual routes. CREATE wants to tackle this paradox for materials showing such incompatibility. A stack of intermediate underlayers between the substrate and its coating will be investigated. One relevant deposition process for the growth of 100 µm thick or millimetric layers is laser cladding of a metallic powder projection, which belongs to the family of additive manufacturing. It allows for a fine in situ tuning of composition gradients by gradually varying the balance between different powder materials. The coating selection and conception will be guided by a numerical approach. This concerns metallurgical modeling and comprises mechanical and thermal aspects and the thermodynamics of the materials involved. The thermodynamic systems included range from binary to quaternary.
The primary objective of the POROFAB project is to evaluate the elaboration of parts with complex geometries by additive manufacturing, functionalized afterwards by the gaseous deposition of a thin film. This new production route of innovative components with a high added value is strategic and presents a strong potential for the CEA and external applications (aeronautics, terrestrial and maritime transportation, energy, industrial parts). Several representative model components gathered from the synthesis of several functional industrial specifications will be elaborated by additive manufacturing. They will then be coated by different materials (Al2O3, TiN and Cr-based) using chemical vapor deposition processes (CVD). The numerical modeling of the deposition processes will assist the optimization of the elaboration parameters in addition to classical material characterization in order to ensure a homogeneous coating thickness across the whole component. The functionalized part will finally undergo several tests to evaluate its performances. After demonstrating the viability of this new production route, specific components regarding CEA or external applications will be designed.
The tremendous complexification of specifications requires more and more complex 3D components exhibiting surface properties often different from volume properties. Their multifunctionality can be ensured by a coating grown using chemical vapor deposition (CVD). In this context, the CEA and CIRIMAT wish to sustain their collaboration and develop the knowledge and data acquired over several years on the DLI-MOCVD process (Direct Liquid Injection – MetalOrganic CVD). The main objective of PRECOS is to demonstrate the relevancy of this process at a pilot scale. The first step of PRECOS aims at increasing the reliability of the current pilot deposition chamber, its reproducibility and capacity to treat parts with various shapes and sizes. A numerical model of the coating process will support classical characterization. New chemical systems will be evaluated in a second step to target applicative fields other than the nuclear one. The elaboration of diverse layers of multifunctional materials on components with complex geometries by a simple, clean, economically and industrially efficient and viable DLI-MOCVD process is certainly a challenge. This work has a highly strategic interest for the CEA, at least concerning new energy technologies, current and forthcoming nuclear industries, numerical sciences, technologies for future medicinal applications and also external industrial needs.