The microstructures of materials dictate their macroscopic mechanical behavior (i.e., plasticity and fracture).  Our group studies the links between microstructure and mechanical behavior through novel experiments and computational modeling to develop predictive and extensible mechanical models.  We perform mechanical tests over a range of length scales to probe sub-grain level to macroscale deformation.  We compare the experimental results with microstructural finite element representative volume element models and crystal plasticity models developed in our group.  The hybrid experimental-computational approach allows for the identification of deformation and failure mechanisms of different metals and alloys, which enables the development of physically based, predictive plasticity and fracture models.

The ductile fracture behavior of metals depends on the stress state under which they are loaded.  Thus, to efficiently use these materials, predictive models that describe the failure of these materials under realistic complex loading states are needed.  Our lab uses a combined experimental-computational approach to investigate and model large deformation and fracture behavior.  Experimentally, we use a custom-built dual-actuator mechanical test frame along with specially designed test specimens to measure the multiaxial deformation behavior of a range of metal alloys, from advanced high strength steels to additively manufactured alloys.  We couple these experiments with computational simulations to develop plasticity and fracture models for these materials.

In additive manufacturing (AM) of metals, 3D components are built in a layer-by-layer fashion using different processes including powder-based laser powder bed fusion (L-PBF), laser-based directed energy deposition (DED), and binder jetting.  AM technology allows for the fabrication of design-driven components rather than process-dependent geometries. However, an understanding of the mechanics of materials made by AM, based on their unique processing conditions and thermal histories, is required for the adoption of AM in load-bearing applications.  Our group works to understand the multiaxial mechanical behavior of materials made by AM through combined experimental and computational methods.

Our group focuses on using machine learning to interpret the process-structure-property relationship for the PBF-LB AM materials from two prospectives:
• Statistical analysis and machine learning: identification of the complex process-structure-property relationship, with a combination of experimental investigation, including processing design, microstructure characterization, and mechanical testing, and data-driven models and feature important analysis.
• Computer vision and deep learning: in situ processing signals that are indictive of various processing conditions, are used to predict the sample mechanical properties, with the computer vision and deep learning techniques, toward application in real-time process diagnosis and control system.

Functionally graded materials (FGMs) have the potential to expand the design space within additive manufacturing (AM) for spatially tailored properties within single components. In FGMs, material properties vary spatially due to intentional changes in chemistry or microstructure within a single component.  Our group studies metallic FGMs fabricated using directed energy deposition (DED) additive manufacturing.  With DED AM, the ratios of powder metal feedstock being fed into the melt pool can be varied, allowing for targeted changes in chemistry, and therefore, phases and properties, as a function of location.  We use a combination of experimental characterization and computational simulations to analyze the microstructure, chemistry, phase composition, and mechanical properties of FGM systems, and use the resulting information to inform the design of future gradient pathways. 

Additive manufacturing provides the ability to fabricate parts with geometries that are challenging or impossible to produce with traditional subtractive methods, including thin lattice structures and other topologically optimized geometries. These structures offer high elastic modulus, specific strength, and energy absorption capacity; thus these structures provide routes for lightweighting of components. However, in parts with small feature sizes, fewer grains are present across each feature and microstructure is more heavily affected by local changes in processing variations and thermal history than bulk parts. Our group investigates how part geometry impacts these process-structure-property relationships in additively manufactured small-scale geometries.

Alloys currently used in additive manufacturing are primarily limited to those designed for welding and casting, thus, they are not optimized for the complex thermal histories seen in fusion-based AM of rapid solidification followed by repeated thermal cycling with the addition of layers. Understanding and designing advanced alloys for AM requires knowledge of non-equilibrium microstructure as a function of alloying elements and thermal history. We use computational methods informed and validated by experimental methods to develop alloys specifically tailored for AM.