Integration between Modeling and Experiments for Crystalline Metals: From Atomistic to Macroscopic Scales IV: Session VI
Sponsored by: TMS Advanced Characterization, Testing, and Simulation Committee, TMS Materials Characterization Committee, TMS: Nanomaterials Committee
Program Organizers: Arul Kumar Mariyappan, Los Alamos National Laboratory; Irene Beyerlein, University of California, Santa Barbara; Levente Balogh, Queen's University; Caizhi Zhou, University of South Carolina; Lei Cao, University of Nevada; Josh Kacher, Georgia Institute of Technology
Wednesday 2:00 PM
October 12, 2022
Room: 401
Location: David L. Lawrence Convention Center
Session Chair: Marc De Graef, Carnegie Mellon University; Andrea Hodge, University of Southern California
2:00 PM Invited
Combinatorial Synthesis and High-throughput Characterization for Alloy Systems: Andrea Hodge1; 1University of Southern California
With the rapid ascend of machine learning as part of materials development, it is important to find synergy between experimental and computational efforts for faster materials discovery. In this talk, an overview and specific methodologies will be discussed using high-throughput experimental techniques ranging from synthesis to mechanical testing. These techniques allow the creation of experimental data sets which can be used to construct materials libraries.In his context, sputtered compositional and microstructural complex metallic alloys will be presented as model systems for high-throughput synthesis and characterization. We will examine the data complexity of going from four to hundreds of compositions in a single sputtering run and how machine learning can be implemented to guide both the synthesis and characterization space.
2:30 PM
Third Generation Thermodynamic Modelling of the Ga-Mn-Ni System: Liangyan Hao1; Wei Xiong1; 1University of Pittsburgh
To evaluate the feasibility of physics-based third generation unary data and thermodynamic models in higher-order systems, Ga-Ni and Mn-Ni systems are successfully optimized using the descriptions for Ga, Ni, and Mn by considering contributions from lattice vibration, electronic excitation and magnetic ordering. The heat capacity of liquid-amorphous phases is modelled by the two-state model. More importantly, the high-temperature heat capacity of solid phases is directly extrapolated from the low-temperature one. To prevent solid phases from restabilizing under very high temperatures, the Equal Entropy Criteria is adopted. The improved magnetic model is adopted to calculate the magnetic properties and magnetic ordering energy. Meanwhile, to avoid nonzero entropy at 0 K, the temperature dependence of the interaction parameters for solution phases and formation energies for stoichiometric compounds is expressed by THETA parameter. The calculated phase diagram and thermochemical properties of Ga-Ni and Mn-Ni systems are in good agreement with experimental results.
2:50 PM
Molecular Dynamics Analysis and Optimization of Ultra High Temperature Ceramic (UHTC) Compositions for Propulsion: Robert Slapikas1; Anindya Ghoshal2; Luis Bravo2; Muthuvel Murugan2; Ryan Mcgowan2; Patrick Albert1; Justin Reiss1; Petr Kolonin3; Susan Sinnott1; Douglas Wolfe1; 1Penn State; 2U.S. Army Research Laboratory; 3Applied Research Laboratory, The Pennsylvania State University
Atomistic simulations were performed to characterize the material properties of UHTC for aero-propulsion usage. The mechanical and thermal properties of transition metal borides ZrB2, HfB2, and UHTC composites containing SiC are determined using mechanical testing and the Green-Kubo formulation. In conjunction with the LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) code, classical molecular dynamics simulations are performed using the bond-order Tersoff interatomic potentials. The mechanical characteristics of the polycrystalline UHTC were measured at strain rates ranging from 1E7 to 1E10 /s, and the results support previously published and completed experimental data. Hardness measurements are made at various loads and grain sizes, indicating the existence of an inverse Hall-Petch relationship. The findings establish a foundation for connecting constituent behavior to composite performance, which is particularly useful when experimental data are unavailable. The various simulations illustrate how the material properties of these UHTC and composites change when the ambient temperature increases.
3:10 PM Invited
Examination of Computed Aluminum Grain Boundary Structures and Interface Energies that Span the 5D Space of Crystallographic Character: Eric Homer1; Gus Hart1; Braxton Owens1; Derek Hensley1; Jay Spendlove1; Lydia Serafin1; 1Brigham Young University
The space of possible grain boundary structures is vast, with 5 macroscopic, crystallographic degrees of freedom that define the character of a grain boundary. While numerous datasets of grain boundaries have examined this space, none has systematically examined the full range of possibilities. We present a computed dataset of more than 5000 unique aluminum grain boundaries in the 5D crystallographic space. Our sampling includes a range of possible microscopic, atomic configurations for each unique 5D crystallographic structure, which we refer to as metastable grain boundary structures. In all, the number of metastable structures associated with all the unique grain boundaries is over 36 million. We will present an overview of the methods used to generate this dataset, an initial examination of the raw data, as well as methods and insights gained in machine learning of grain boundary energy structure-property relationships and relationships between metastable structures for each unique grain boundary.
3:40 PM Break
4:00 PM
Predictive Phase-field Modeling of Nucleation and Growth of β1 Precipitates during Aging of Mg-Nd Alloys: David Montiel1; Stephen DeWitt1; Qianying Shi1; Zhihua Huang1; Katsuyo Thornton1; John Allison1; 1University of Michigan
The spatial distribution and morphology of β1 precipitates are key to determining the precipitation hardening response for a range of Mg-RE alloys. We apply a phase-field model to simulate the nucleation and growth of β1 precipitates in Mg-Nd alloys during aging. The model is implemented in PRISMS-PF, an open-source, massively parallel framework for high-performance phase-field simulations. Our simulation results were validated against experimental measurements at different stages of aging, namely matrix composition data acquired via atom probe tomography (APT), and precipitate sizes and number density, acquired from transmission electron microscopy (TEM) images. We demonstrate that the model is capable of predicting the morphology, orientation and size of precipitates, as well as their number density during nucleation, growth and coarsening. The predictive phase-field model will allow us to optimize aging conditions in order to obtain the desired microstructure and mechanical properties.
4:20 PM
Propagation of Uncertainty in Molecular Dynamic Simulations of Polycrystalline Nickel: Meizhong Lyu1; Anqi Qiu1; Elizabeth Holm1; 1Carnegie Mellon University
Uncertainty quantification (UQ) and uncertainty propagation (UP) have received attention as they relate to the validity and robustness of simulation-based materials research; however, these concepts have not often been applied to microstructural evolution. The sensitivity of the evolution trajectory to the initial conditions is not well-understood even in such familiar processes as polycrystalline grain growth. In an initial Molecular Dynamics (MD) study of four-grain junction decomposition (i.e. the T1 topological transformation), we found that the direction of decomposition depends on the random velocity seed. In this study, we quantify the uncertainty associated with the initial velocity of atoms in MD simulations of grain growth in polycrystalline nickel. We find that UP at the atomic scale can alter the growth or shrinkage trajectory of grains at the microstructural scale; this has significant implications for comparing simulation results with experiments.
4:40 PM Invited
Defects and the Electron Beam Interaction Volume in Electron Back-scattered Diffraction: Marc De Graef1; 1Carnegie Mellon University
Electron back-scattered diffraction (EBSD) has for nearly three decades been used to determine grain orientations ("orientation imaging") as well as the local strain state (using high angular resolution of HR-EBSD). Experimentally, it is well known that a high defect concentration near the surface can substantially deteriorate the pattern quality; in fact, analysis of small lattice orientation changes due to geometrically necessary dislocations (GNDs) provides a means to determine that dislocation density using kernel average misorientation maps and related approaches. The question to be addressed in this contribution is whether or not the actual EBSD patterns themselves contain useful information that can be correlated directly to the local defect density, without the need for averaging procedures over small local neighborhoods. We will describe a modeling approach to take dislocation displacement fields into account in the simulation of EBSD patterns and how pattern sharpness can then be related to defect density.