Comparison of Grain-Growth Mean-Field Models Regarding Predicted Grain Size Distributions.

Mean-field models have the ability to predict the evolution of grain size distribution that occurs through thermomechanical solicitations. This article focuses on a comparison of mean-field models under grain-growth conditions. Different microstructure representations are considered and discussed, especially regarding the consideration of topology in the neighborhood construction. Experimental data obtained with a heat treatment campaign on 316L austenitic stainless steel are used for the identification of material parameters and as a reference for model comparisons. Mean-field models are also applied to both mono- and bimodal initial grain size distributions to investigate the potential benefits of introducing neighborhood topology in microstructure prediction models. This article demonstrates that improvements in the predictions can be obtained in monomodal cases for topological models. In the bimodal test, no comparison with experimental data was performed as no data were available. But relative comparisons between models indicated few differences in the predictions. Although of interest, the consideration of neighborhood topology in grain-growth mean-field models generally results in only small improvements compared to classical mean-field models, especially in terms of implementation complexity.

Simulation of Continuous Dynamic Recrystallization Using a Level-Set Method.

Dynamic recrystallization is one of the main phenomena responsible for microstructure evolution during hot forming. Consequently, obtaining a better understanding of dynamic recrystallization mechanisms and being able to predict them is crucial. This paper proposes a full-field numerical framework to predict the evolution of subgrain structures upon grain growth, continuous dynamic recrystallization, and post-dynamic recrystallization. To be able to consider a subgrain structure, two strategies are proposed. One relies on a two-step tessellation algorithm to generate a fully substructured microstructure. The second strategy enables for the simulation of the formation of new subgrains during hot deformation. Using these tools, the grain growth of a fully substructured microstructure is modeled. The influence of microstructure topology, subgrain parameters, and some remaining stored energy due to plastic deformation is discussed. The results highlight that the selective growth of a limited number of subgrains is observed only when mobility is a sigmoidal function of disorientation. The recrystallization kinetics predicted with different criteria for discrimination of recrystallized grains are quantitatively compared. Finally, the ability of the framework to model continuous dynamic and post-dynamic recrystallization is assessed upon a case study representative of the hot extrusion of a zircaloy-4 billet (T=650 °C;ε˙=1.0s-1;εf=1.35). The influence of grain boundary properties and nucleation rules are quantified to evaluate the model sensitivity and suitability. Application of these numerical tools to other thermomechanical conditions and microstructures will be presented in an upcoming article.

Level-Set Modeling of Grain Growth in 316L Stainless Steel under Different Assumptions Regarding Grain Boundary Properties.

Two finite element level-set (FE-LS) formulations are compared for the modeling of grain growth of 316L stainless steel in terms of grain size, mean values, and histograms. Two kinds of microstructures are considered: some are generated statistically from EBSD maps, and the others are generated by the immersion of EBSD data in the FE formulation. Grain boundary (GB) mobility is heterogeneously defined as a function of the GB disorientation. On the other hand, GB energy is considered as heterogeneous or anisotropic, which are, respectively, defined as a function of the disorientation and both the GB misorientation and the GB inclination. In terms of mean grain size value and grain size distribution (GSD), both formulations provide similar responses. However, the anisotropic formulation better respects the experimental disorientation distribution function (DDF) and predicts more realistic grain morphologies. It was also found that the heterogeneous GB mobility described with a sigmoidal function only affects the DDF and the morphology of grains. Thus, a slower evolution of twin boundaries (TBs) is perceived.

Comparative Study and Limits of Different Level-Set Formulations for the Modeling of Anisotropic Grain Growth.

In this study, four different finite element level-set (FE-LS) formulations are compared for the modeling of grain growth in the context of polycrystalline structures and, moreover, two of them are presented for the first time using anisotropic grain boundary (GB) energy and mobility. Mean values and distributions are compared using the four formulations. First, we present the strong and weak formulations for the different models and the crystallographic parameters used at the mesoscopic scale. Second, some Grim Reaper analytical cases are presented and compared with the simulation results, and the evolutions of individual multiple junctions are followed. Additionally, large-scale simulations are presented. Anisotropic GB energy and mobility are respectively defined as functions of the mis-orientation/inclination and disorientation. The evolution of the disorientation distribution function (DDF) is computed, and its evolution is in accordance with prior works. We found that the formulation called "Anisotropic" is the more physical one, but it could be replaced at the mesoscopic scale by an isotropic formulation for simple microstructures presenting an initial Mackenzie-type DDF.

A 2D Front-Tracking Lagrangian Model for the Modeling of Anisotropic Grain Growth.

Grain growth is a well-known and complex phenomenon occurring during annealing of all polycrystalline materials. Its numerical modeling is a complex task when anisotropy sources such as grain orientation and grain boundary inclination have to be taken into account. This article presents the application of a front-tracking methodology to the context of anisotropic grain boundary motion at the mesoscopic scale. The new formulation of boundary migration can take into account any source of anisotropy both at grain boundaries as well as at multiple junctions (MJs) (intersection point of three or more grain boundaries). Special attention is given to the decomposition of high-order MJs for which an algorithm is proposed based on local grain boundary energy minimisation. Numerical tests are provided using highly heterogeneous configurations, and comparisons with a recently developed Finite-Element Level-Set (FE-LS) approach are given. Finally, the computational performance of the model will be studied comparing the CPU-times obtained with the same model but in an isotropic context.