Simulation de la formation des microstructures dans les superalliages à base de nickel de type AM1 durant la solidification et les traitements thermiques

Nickel base superalloys are used for the manufacturing of monocrystalline turbine blades, due to their outstanding mechanical properties at high temperature. The mechanical properties strongly depend on the presence of γ’ precipitates in the γ matrix. Improper size or spatial distribution of the precipitates is one of the possible defects that can appear during the processing of turbine blades. In order to anticipate such defects and optimize the processes, numerical simulation can be used to predict the formation and evolution of precipitates as a function of the alloy chemistry and heat treatment conditions. Numerical models can also be used to analyze the influence of microsegregation, which develops during solidification and does not entirely disappear during the solution heat treatment. A modeling approach has been developed to simulate the formation of microstructure in AM1-type nickel based superalloys during the main processing steps. Two distinct numerical models were developed for the simulation of microsegregation, on one hand, and precipitation on the other hand. An existing microsegregation model [1] was modified to take into account important specificities of Ni-base superalloys, such as non-diagonal matrices of diffusion coefficients. The model was applied to several alloys to predict the evolution of the composition profiles, and the formation/dissolution of interdendritic phases during solidification and the solution heat treatment. A new precipitation model was developed in order to predict the formation and evolution of γ' precipitates during heat treatment [2]. The model is based on the description of the precipitate size distribution, which is tracked based on nucleation and growth laws, taking into account coarsening through the Gibbs-Thomson effect. The time evolution of average quantities, such as the average radius, the number density of precipitates and volume fraction is deduced from the particle size distribution. The model is coupled to the phase diagram software Thermo-Calc, which makes it possible to calculate equilibrium concentrations and nucleation driving force in multicomponent industrial alloys. The coupling to Thermo-Calc is also exploited to compute the effective diffusion coefficients from a mobility database. The microsegregation and precipitation models were applied to ternary Ni-Al-Cr alloys and to the industrial CMSX-4 and AM1 superalloys. The results of the simulations were compared with experimental data collected from the literature and atom probe tomography measurements carried out during the project. The simulation results showed to be globally in good agreement with the experiments. The accuracy of the thermodynamic database turned out to be one of the most important factors for quantitative prediction, in particular for precipitation. Chained simulations were performed on Ni-Al-Cr and AM1 alloys in order to assess the sensitivity of precipitation kinetics with respect to microsegregation. The cal