Traitement thermique d'un métal

Le traitement thermique d'une pièce de métal consiste à lui faire subir des transformations de structure grâce à des cycles prédéterminés de chauffage et de refroidissement afin d'en améliorer les caractéristiques mécaniques : dureté, ductilité, limite d'élasticité Ce procédé est souvent couplé avec l'emploi d'une atmosphère contrôlée lors de la mise en température de la pièce, soit pour éviter son oxydation, soit pour effectuer un apport ou changement moléculaire de surface (traitement de surface). Généralités vignette|Solution solide de substitution (haut) et d'insertion (bas) dans un cristal. Dans les métaux, les atomes sont organisés sous la forme de cristaux : ils forment une structure ordonnée. Des atomes étrangers — impuretés, éléments d'alliage — peuvent s'introduire dans ce réseau, soit en substitution des atomes « de base », soit en insertion, c'est la notion de solution solide. Par ailleurs, il peut y avoir des cristaux de plusieurs types, comme des inclusion

Development of High-Strength Aluminum Alloys for Additive Manufacturing

Marvin Lennard Schuster

Metal additive manufacturing (AM) offers the possibility to rapidly produce complex geometries that are not achievable with conventional manufacturing methods. The two most common technologies, Laser Powder Bed Fusion (LPBF) and Direct Metal Deposition (DMD), are characterized by high process temperatures, fast heating, and cooling rates, and an associated far-from-equilibrium solidification. This is problematic when processing materials that were developed and optimized many years ago for conventional manufacturing processes. Common difficulties are the occurrence of gas pores in the component, which are caused by the evaporation of volatile alloying elements during the process, the formation of hot cracks due to large solidification intervals, and the formation of metastable phases. For the processing of high-strength, precipitation-hardening Al-Cu aluminum alloys, this prevents the use of non-modified conventional alloys. However, the development of novel alloys tailored to these processes not only allows these difficulties to be overcome but also opens up the possibility of improving material properties through the targeted exploitation of process-inherent properties. This includes the processing of oxide particle-reinforced alloys, which can not be produced by conventional casting processes. The standard heat treatment usually performed for precipitation hardening, which has also been optimized for conventional materials, must be adapted accordingly to achieve optimum material properties. Within the scope of this work, the microstructure, precipitation, and defect formation after LPBF as well as the heat treatment behavior are thoroughly investigated based on an Al-Cu-Mg-Zr alloy for the first time. The obtained findings are applied to design a novel 2618 Al-Cu alloy for LPBF and DMD, with suitable heat treatment being developed for LPBF. The alloy is characterized with respect to microstructure, precipitation formation, and mechanical properties before, during, and after heat treatment. Furthermore, by modifying elemental Al-Zr powder blends with nm-sized Al2O3, an oxide dispersion strengthened alloy is produced by LPBF and investigated in detail.

EPFL2022

Microstructure of single crystal Ni-based superalloys during blade manufacturing process and creep deformation

Stéphane Pierret

The aim of the thesis is to obtain more understanding about the influence of plastic deformation on the microstructural changes observed in single crystal (SX) Ni-based superalloys using diffraction techniques. This work was organised around two main problematics: (1) rafting in turbine blades after the manufacturing process and (2) rafting and mosaicity evolution in two SX alloys with different composition during creep deformation. SX Ni-based superalloys are first choice materials for turbine blade application in land-based gas turbines and aero-engines. Superior mechanical properties of these materials at high temperatures are achieved due to the optimisation of the alloying composition and the microstructure. SX Ni-based superalloy turbine blades solidify by dendritic solidification with dendrites aligned along the directions of lowest Young modulus (interdendritic spacing around 300μm). At the end of the manufacturing process, the microstructure consists of high volume fraction of cubic γ’ precipitates (around 500nm) coherently embedded in the γ matrix. However specific parts of engine-ready turbine blades exhibit directionally coarsened γ’ precipitates in a preferred direction (rafting) without the influence of an applied load. It is of prime importance to understand the origin of rafting prior to the introduction of the blades in the engine since rafting can alter the mechanical properties of the alloy. Chemical segregations and the introduction of plastic deformation prior to thermal treatments were identified as potential driving force for rafting during load-free high temperature annealing. Most of the SX Ni-based superalloys used in the current blade production in Alstom contain significant amount of refractory elements which are expensive and increase the density of the alloy. To answer the demand of cheaper and lighter components, Alstom launched a development program to design novel SX superalloys. One of these projects led to the production of a Re-free superalloy, so-called MD2. Extensive mechanical investigations on MD2 are on-going and compared to the behaviour of the MK4HC alloy (3wt% Re) currently used in turbine blade production. One of the major differences in terms of mechanical properties between Re-containing and Re-free SX alloys is the faster kinetics of coarsening and rafting during creep deformation at high temperature in the latter alloys. It prevented so far their application in the hottest areas in gas turbines. It is therefore of utmost importance to acquire more knowledge about the mechanisms associated with the microstructure evolution during creep of the new MD2 SX alloy in comparison with the MK4HC alloy. The evolution of the lattice parameter misfit in different directions with respect to the applied load during creep deformation was associated with the evolution of the microstructure from cubic to elongated γ’ precipitates It could be shown that rafting occurs after the cooling stage following the solution heat treatment (SHT) during the manufacturing of the blade with investigations of the microstructure. The influence of chemical segregations on rafting was excluded with the investigation of the blade alloy composition in different parts of the blade cooled from the SHT and diffraction measurements in stress-relieved combs prepared from the same blade. The formation of a rafted microstructure during the manufacturing process was associated with the formation of plastic strains during cooling after the SHT. This was confirmed after gathering results from investigations of the microstructure, simulation of the cooling stage and neutron diffraction measurements on the blade. The investigations of the microstructure performed on the engine-ready blade revealed an inversion of the γ’ raft orientation below (parallel to [001]) and above the platform of the blade (perpendicular to [001]). This was attributed to an inversion of the sign of residual strain between the previously mentioned positions as revealed by neutron diffraction measurements performed along the airfoil of the blade cooled after the SHT. Microscopy also showed that the rafted microstructure was more pronounced below the platform (long γ’ rafts) than above the platform (shorter γ’ rafts with retained cubic precipitates). The concept of the threshold plastic strain necessary to induce rafting during annealing subsequent to plastic deformation was used to describe the difference in rafted microstructure. It is suggested that plastic strains largely exceed the threshold below the platform leading to fully formed rafted microstructure. On the other side, plastic strains of the order to the threshold strain are suggested above the platform in order to explain the partially formed rafted microstructure Neutron diffraction measurements revealed that the lattice parameter misfit was larger in the MK4HC than in the MD2 alloy at 20°C in the unloaded state. The larger misfit in the MK4HC alloy was attributed to the larger concentration of elements dissolving in the γ phase and to the presence of Re. During creep deformation, coherency stresses are relieved at the γ/γ’ interface in the γ channels perpendicular to the applied load by dislocations. On the other hand, dislocations increase the internal stress at the γ/γ’ interface in the γ channels parallel to the applied load. This leads to the rafting of the γ’ precipitates perpendicular to the applied load. This was observed in the MK4HC. The absence of clear misfit evolution in the MD2 alloy was attributed to the anomalous microstructure at the beginning of the creep deformation. The mosaicity of the MK4HC and MD2 alloys revealed crystallographic misorientation between neighbouring dendrites up to 0.5°. The misorientation was associated with the presence of dislocations in the regions between the dendrites formed during the dendritic solidification of the alloys from existing work. It was shown that the composition of the SX alloys and short periods at high temperature in the unloaded state had only little influence on the misorientation between neighbouring dendrites. Upon creep loading to 180MPa, the mosaicity markedly changed in the two SX alloys independent of the creep temperature. In particular, the number of single misoriented domains measured by X-ray diffraction decreased. This was attributed to the fast propagation of the dislocations present in the interdendritic structure inducing a rotation of the individual dendrites. During creep deformation, the mosaicity increased as visible from the broadening of the maximum spread in misorientation. The formation of inhomogeneous plastic strain fields in the γ/γ’ microstructure is believed to create new mosaic blocks in the dendritic structure of the alloys leading to the observed broadening of the mosaicity. The influence of plastic deformation on the evolution of the mosaicity was demonstrated. The mosaicity significantly changed during creep of the MD2 at 1050°C/180MPa and 1000°C/180MPa whereas no significant evolution could be measured during creep at 950°C/180MPa. This was attributed to a temperature dependence of the yield strength of the MD2 alloy. It therefore induces difference in dislocation propagation at 950°C compared to 1050°C and 1000°C which retards the broadening of the mosaicity during creep at 950°C. Furthermore, the mosaicity formed during creep deformation was retained upon unloading and cooling of the SX samples to room temperature.

EPFL2012

Investigation of physiological roles of AMPK-activated protein kinase y3

Philipp Jurijvic Rhein

Clinical trials have shown that direct activators of an evolutionary-conserved metabolic sensor, AMP-activated protein kinase (AMPK), are beneficial in preventing/treating a range of metabolic disorders, including type 2 diabetes. These activators, including 991/MK-8722, bind at regulatory site, termed the allosteric drug and metabolite (ADaM-site) at the interface between the catalytic alpha subunit and regulatory beta subunit. In addition, another approach for direct activation of AMPK is through mimicking its natural ligands, AMP/ADP, by binding to the nucleotide site in the y3 subunit, e.g. 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR). Interestingly, their ability to influence glucose homeostasis is reported to be through activation of AMPK complexes found within the skeletal muscle. The AMPKy3 is unique as an AMPK subunit isoform since it is selectively expressed in glycolytic skeletal muscle fibres. To gain more insight into the y3 isoform-specific regulation of glucose metabolism, we used a y3 knock-out (y3-/-) mouse model to investigate the effects of the AICAR and 991/MK-8722. We confirm that y3 isoform expression and activity is high glycolytic muscles, including extensor digitorum longus (EDL), whereas there is very little detected in oxidative soleus (SOL) muscle. We show that AICAR and 991/MK-8722 stimulated glucose uptake in both EDL and SOL muscles. Interestingly, only the ability of AICAR to induce glucose uptake was blunted in EDL muscle taken from y3-/- mice. Consistent with this, whole-body glucose clearance was also impaired in y3-/- mice in response to an AICAR, whilst MK-8722 lowered blood glucose similarly between WT and y3-/- mice. Despite the presence of y1-containing complexes, which are the most predominant complexes in both EDL and SOL muscles, this suggests that different AMPKy isoforms play different roles in glucose homeostasis in response to AICAR. We further looked at de novo glycogen synthesis and glucose utilisation in skeletal muscle. Despite similar basal glucose uptake between muscles from WT and y3-/-, y3-deficiency results in lower basal glycogen content only in glycolytic muscles. We show that this is not due to an inability to synthesise glycogen de novo, since MK-8722 was able to increase glycogen levels in EDL. Instead, we show that the decrease in basal glycogen in y3-/- EDL, may be linked lower expression of UDP-glucose pyrophosphorylase 2 (UGP2). This suggests that AMPKy3 may play a role in steering the molecular fate of glucose inside the cell. Taken together, whilst we show that AMPKy3 is dispensable for ex vivo glucose transport and in vivo glucose homeostasis in response to ADaM-site AMPK activators, nucleotide-mediated activation of y3 by AICAR is impaired. Further research is required to understand the nucleotide regulation of different y-containing AMPK complexes. This differential activation mechanism could be exploited to specifically target AMPKy3 complexes in the muscle, and potentially avoid deleterious effects of activation of AMPK complexes in other tissues.

EPFL2021