The Predict Of Phase Composition And Pahse Amount And Sublattice Composition In Ni Base Superalloys

The effects of mismatch in atomic size and electronegativity of alloying elements on the elemental concentration of the sublattices of γ' phase in Ni-base superalloys were investigated. The mathematic model of optimizations was established to predict the elemental concentration of sublattices of γ' phase. Ranges of the selected values of concentration variables were determined by use of the iterative and substitutional method. Based on chemical compositions of the γ' phase, the elemental content at the Ni and Al sites in the γ' phase can be calculated with the layered multi-objective optimization algorithm with tolerance. Using the reported data on both the chemical compositions and the lattice parameters of γ' phase, comparisons between the predicted and the reported values of lattice parameters were made through the application of the predicted values of the elemental concentrations of the γ' phase to verify the feasibility and the accuracy of the present method.An optimization method was established to predict phase compositions and amounts in two-phase nickel-base superalloys. Based on the compositions of alloys and their γ' phase, γ phase compositions and γ' phase amounts were calculated with the use of the layered multi-objective optimization algorithmic approach with tolerance. In addition, in order to investigate the relationship between the phase composition and the elemental concentration of the sublattices of the second phase, a optimization method was established to predict the second phase composition and amounts based on the alloy composition and the γ phase composition. The results verified the feasibility and the accuracy of the present method.A neural network method was proposed to predict γ' phase amounts at different temperatures for nickel-base superalloys. Algorithm weight matrixes which accurately express the complex non-linear relationships of γ' phase amounts with temperatures and alloy compositions were obtained based on the back-propagation neural network modeling on the practical values of temperature, alloy composition and γ' phase amount in various superalloys. The feasibility and accuracy of the proposed prediction method were verified through the comparison of the predicted and measured γ' phase amount values.