The Phase Field Method: Towards a Tool for Materials Design

Central to designing new materials is the ability to link the processing of a material to its microstructure in alloys of commercial importance that typically that contain many elements and phases. For example, it is challenging to predict the -phase morphologies and their size scale in multicomponent commercial Ti alloys, whose microstructures can vary from globular to basket-weave, and thus connect processing to properties. Nearly every structural material is polycrystalline. The grain size clearly affects the mechanical properties of the alloy yet predicting the grain size distribution and morphology during isothermal grain growth, much less as a function of thermal processing, remains a challenge. Many materials are used in their cast form, such as alloys printed using Additive Manufacturing (AM). Thus, the compositional inhomogeneities that result from the solidification process cannot be removed or, as in the AM case, require a costly long-term heat treatment. Understanding the chemical segregation patterns resulting from solidification in multicomponent alloys is thus of great importance. Phase field methods provide a flexible framework that can potentially address all of these critical issues. The method can be used to predict the complex morphologies that result from phase transformations and grain evolution. Moreover, new physics, e.g. stress, can easily be added, and topological transformations, such as precipitate merging and splitting, occur naturally. However, challenges remain that prevent phase field methods to be used routinely in a materials design effort. We illustrate this challenges in many ways. For example, there is a great need for a computationally efficient phase field method that can be used to follow morphological evolution in multicomponent alloys of interest to the Navy during both solidification and -phase precipitation. Phase field models for grain growth that incorporates all five degrees of freedom of the grain boundary energy and mobility remain to be developed. In order to address the computational overhead associated with phase field computations it is also necessary to develop reduced order models that use the results of phase field calculations to yield computationally efficient models that are easily incorporated in a materials design effort. Most importantly, given the flexibility of the phase field models, it will be possible to address problems suggested by researchers at the Naval Research Laboratory and NSWC-Carderock. In order to foster close interactions, the postdoc supported by the program will spend 3 months per year at either location, and 9 months at Northwestern.