Parametrically Homogenized Crystal Plasticity Model for Nickel-based Superalloys: Intragranular Microstructures To Polycrystalline Aggregates
Johns Hopkins University
The deformation mechanics of nickel-based superalloys at every length scale are influenced by a unique array of dislocation mechanisms evolving at the nanoscale. The origin of the lower scale phenomena within a single crystal is primarily due to the existence of an ordered intermetallic γ′ precipitate phase embedded within a γ solid solution matrix phase. The chemical composition and the γ−γ′ matrix-precipitate microstructure of Ni-based superalloys are highly engineered to optimize for creep, fatigue, thermal, and corrosion-resistant properties. The precise morphology and spatial configuration of the γ′ precipitates control and impede the flow of dislocations and, therefore, govern the overall viscoplastic response of the material. This thesis develops a hierarchical multiscale crystal plasticity framework for single crystal Ni-based superalloys. An emphasis is placed on the creation of an image-based model that captures the complex γ−γ′ configuration of experimental three-dimensional microstructures. In generating such a model, important considerations emerge regarding microstructure generation, representative volume element analysis, boundary condition selection, and homogenization methodology. Each of these issues is addressed with a combination of computational tools from statistics, machine learning, optimization, and continuum mechanics. A multiscale modeling pipeline is established in the development of a parametrically-homogenized crystal plasticity model (PHCPM). The PHCPM explicitly incorporates morphological statistics of the γ−γ′ intragranular microstructure in the crystal plasticity constitutive coefficients. This enables highly efficient and accurate image-based polycrystalline microstructural simulations. The single crystal PHCPM development process involves: (i) construction of statistically-equivalent representative volume elements, (ii) image-based modeling with dislocation-density crystal plasticity model, (iii) identification of representative aggregated microstructural parameters, (iv) selection of PHCPM framework and (v) self-consistent homogenization. Finally, polycrystalline studies are performed throughout to demonstrate the universality of SERVE techniques for other classes of microstructures, the role of cube slip on texture at high temperatures, and the effectiveness of the PHCPM framework for higher scale simulations.
superalloy, multiscale modeling, phcm, serve, rve, crystal plasticity, homogenization, precipitate, computational model, polycrystal