The recent development of experimental techniques that rapidly reconstruct the three-dimensional microstructures of solids has given rise to new possibilities for developing a deeper understanding of the evolution of microstructures and the effects of microstructures on materials properties. Combined with three-dimensional (3D) simulations and analyses that are capable of handling the complexity of these microstructures, 3D reconstruction, or tomography, has become a powerful tool that provides clear insights into materials processing and properties. This introductory article provides an overview of this emerging field of materials science, as well as brief descriptions of selected methods and their applicability.

This article first provides a brief review of the status of the subfield of three-dimensional (3D) materials analyses that combine serial sectioning, electron backscatter diffraction (EBSD), and finite element modeling (FEM) of materials microstructures, with emphasis on initial investigations and how they led to the current state of this research area. The discussions focus on studies of the mechanical properties of polycrystalline materials where 3D reconstructions of the microstructure—including crystallographic orientation information—are used as input into image-based 3D FEM simulations. The authors' recent work on a β-stabilized Ti alloy is utilized for specific examples to illustrate the capabilities of these experimental and modeling techniques, the challenges and the solutions associated with these methods, and the types of results and analyses that can be obtained by the close integration of experiments and simulations.

With the emergence and evolution of serial sectioning techniques that allow for three-dimensional data collection and the continuing increase in computational power, it is now possible to analyze and compute the evolution of three-dimensional nano- and microstructures. Structures can be accurately characterized, and it is possible to correlate processing paths with materials properties with great precision. Examples of the analysis and computations of the evolution of three-dimensional microstructures are discussed. The focus is on experiments that use serial sectioning methods to determine three-dimensional structure and on phase-field simulations of microstructural evolution that employ experimental three-dimensional data as initial conditions.

Synchrotron x-ray microtomography is a characterization technique increasingly used to obtain 3D images of the interior of optically opaque materials with a spatial resolution in the micrometer range. As a nondestructive technique, it enables the monitoring of microstructural evolution during in situ experiments. In this article, examples from three different fields of metals research illustrate the contribution of x-ray tomography data to modeling: deformation of cellular materials, metal solidification, and fatigue crack growth in Al alloys. Conventionally, tomography probes the 3D distribution of the x-ray attenuation coefficient within a sample. However, this technique is also being extended to determine the local crystallographic orientation in the bulk of materials (diffraction contrast tomography), a key issue for the modeling of microstructure in metals.

Three-dimensional x-ray diffraction (3DXRD) allows nondestructive characterization of grains, orientations, and stresses in bulk microstructures and, therefore, enables in situ studies of the structural dynamics during processing. The method is described briefly, and its potential for providing new data valuable for validation of various models of microstructural evolution is discussed. Examples of 3DXRD measurements related to recrystallization and to solid-state phase transformations in metals are described. 3DXRD measurements have led to new modeling activity predicting the evolution of metallic microstructures with much more detail than hitherto possible. Among these modeling activities are three-dimensional (3D) geometric modeling, 3D molecular dynamics modeling, 3D phase-field modeling, two-dimensional (2D) cellular automata, and 2D Monte Carlo simulations.