Research

Overview

The Advanced Computational Materials Engineering Laboratory is focused on the development, implementation, and utilization of computational simulation frameworks to help understand processing-structure-property relationships in polycrystalline metals and metallic alloys. The macroscopic (i.e., component-scale) response of these materials during deformation loading is ultimately influenced by crystal-scale (i.e., microscopic) deformation phenomena, as well as the material's microstructure (i.e., grain geometry and crystallographic orientations). Owing to the difficulty and cost of studying these complex microscale deformation phenomena in polycrystalline materials during in situ deformation loading, computational methods present a powerful avenue to pursue the establishment of processing-structure-property relationships.

Overall, computational methods represent the future of understanding the deformation response of materials, and have become critical in the interrogation of structure-property relationships that provide engineers the ability to better predict macroscopic behavior and the underlying mechanisms that contribute to an engineering component's mechanical behavior. Our aim is to further the state-of-the-art of predictive tools, and promote their adoption in both academic research and engineering design.

Current Projects

Reduced Order Crystal Plasticity Modeling

Crystal plasticity simulations are typically computationally expensive, making reduced-order models attractive for implementation into component-scale finite element frameworks due to their relatively low computational expense. In this study, the relationship between general loading conditions, crystallographic texture, and the resulting macroscopic yield are investigated through the lens of machine learning algorithms. The result is a model describing the macroscopic yield surface of textured polycrystalline samples, distilling the essence of the costly crystal plasticity simulations into efficient property-focused models.

Processing-Structure-Property Relationships of Aluminum Alloys

Various manufacturing processes introduce complex changes to microstructure morphology, in terms of grain size and shape, and the presence of sub-grain features such as precipitates. These features, in turn, affect the macroscopic behavior of the material. This is especially pronounced in some aluminum alloys produced via additive manufacturing techniques. In this study, crystal plasticity modeling is employed to model the influence of grain-scale and sub-grain microstructural features on the macroscopic deformation response, allowing for an understanding of the structure-property relationship.

Dwell-Fatigue Behavior of Titanium Alloys

Titanium alloys utilized in jet aircraft engines are often subjected to dwell fatigue loading under normal operating conditions. This project is focused on understanding the deformation response of materials under such loading. In titanium alloys, this loading is further complicated by relatively complex microstructural configurations, including the presence of microtexture and dual-crystallographic-phase morphologies, which both present simulation challenges. The understanding of this behavior is integral in predicting the lifespan of aerospace components, and may help explain recent high-profile failures in jet engines.

Deformation Twinning in Hexagonal Materials

While many materials deform plastically only via crystallographic slip, some high-strength structural alloys - primarily those with hexagonal crystal structures - exhibit a distinct mode of plastic deformation, deformation twinning. Twinning is distinguished from slip by its relatively rapid, defined shear, and defined crystal reorientation, and presents modeling challenges owing to its discrete (non-diffuse) nature. In this study, high energy X-ray diffraction experiments will be performed to generate an understanding of the local conditions necessary to initiate twinning during deformation loading. A phenomenological model will developed describing the behavior of twins at the grain scale, which will be implemented in a novel finite element framework to be verified and validated against further experimental data.

Modeling Fracture in Polycrystalline Materials

This project is focused on the understanding of fracture behavior (crack initiation and propagation) in polycrystalline materials. The extended finite element method (XFEM) offers a possible route to consider crack behavior at the inter/intra-grain scale.

Resources and Equipment

Software Resources

ACME Lab members develop and employ various software packages. Primarily, we are home to FEPX, a crystal plasticity finite element framework used for simulating the deformation response of crystalline materials, which is capable of modeling large deformations on high-fidelity representations of polycrystals. We additionally develop software to handle other modeling needs - including modeling texture evolution, fracture behavior, and others. Programs developed by lab members may be found on the software page, and are free and open-source.

Computational Equipment

Workstations: ACME Lab is outfitted with multiple high-performance computational workstations for code development, exploratory simulations, and post-processing. These workstations (Hooke, Hill, Schmid, Taylor, Tresca, and vonMises) each contain 48 CPU cores, 128 GB RAM, and a minimum of 3TB of storage.

Cluster: ACME Lab is home to Cauchy, a computational cluster with 16 compute nodes, each containing 64 CPU cores (1024 CPU cores total), 256 GB RAM, and 1TB of storage. Cauchy is also outfitted with a storage node containing 36TB storage.

Other Resources: ACME Lab members are granted unlimited access at The University of Alabama High Performance Computing center, as well as the Alabama Supercomputer Authority. Additionally, we have deployed our software packages at various HPC centers (e.g., DoD), and may use these resources in support of approved projects.

Collaborators

We interact closely with various research groups both nationally and internationally. We are especially interested in working closely with experimentalists (HEDM, SEM/EBSD, etc.) to inform simulations, spur model development, and provide insight into experimental data and trends. Please contact if you are interested in working together.