Session Sheet - Algorithm Development in Materials Science and Engineering: Computational Models and Algorithms in Atomistic Scale

Algorithm Development in Materials Science and Engineering: Computational Models and Algorithms in Atomistic Scale
Sponsored by: TMS Materials Processing and Manufacturing Division, TMS: Computational Materials Science and Engineering Committee, TMS: Integrated Computational Materials Engineering Committee, TMS: Phase Transformations Committee, TMS: Solidification Committee
Program Organizers: Mohsen Asle Zaeem, University of Tennessee, Knoxville; Mikhail Mendelev, NASA ARC; Bryan Wong, University of California, Riverside; Ebrahim Asadi, University of Memphis; Garritt Tucker, Baylor University; Charudatta Phatak, Argonne National Laboratory; Bryce Meredig, Travertine Labs LLC

Tuesday 2:00 PM
March 16, 2021
Room: RM 36
Location: TMS2021 Virtual

Session Chair: Ebrahim Asadi, University of Memphis

2:00 PM
Characterizing Atomistic Geometries and Potential Functions Using Strain Functionals: Edward Kober¹; Colin Adams¹; Jacob Tavenner²; Nithin Mathew¹; ¹Los Alamos National Laboratory; ²Colorado School of Mines
    The analysis of molecular dynamics simulations of the deformation of metals and the validation of the potential functions used in these simulations require a robust set of descriptors that can identify a wide variety of crystal and defect structures. The use of strain tensor functionals for characterizing such arbitrarily ordered atomistic structures is demonstrated here for use in conjunction with machine learning applications. This approach is derived using a Taylor series expansion, ensuring both numerical convergence and direct relationships to physical properties. They are reduced to a minimal non-redundant set, where these naturally describe the deformations in terms of simple concepts: measuring how tetrahedral or cubic a geometry is, how much shear or trigonal deformation is present. The approach has been extended to the analysis of vector (velocities, forces) and tensor (stress, strain) fields as well. The functions can be simply Fourier transformed and thereby also related to diffraction measurements.

2:20 PM
Modeling Static Recrystallization within the SPParKS Kinetic Monte Carlo Framework for Polycrystalline Materials: Austin Gerlt¹; David Newell²; Adam Pilchak²; Eric Payton²; ¹The Ohio State University; ²Air Force Research Lab
    A new extension has been developed for the SPParKS kinetic Monte Carlo grain evolution simulator that allows for the simultaneous modeling of static recrystallization, pinning particles, annealing, and curvature-driven grain boundary minimization. Static recrystallization is modeled via the minimization of stored energy at each voxel site, which can be a proxy for either the local stress state and/or density of dislocations. This model was then calibrated using experimentally collected EBSD data from additively manufactured IN718 specimens annealed between 1 and 8 hours at 1160C. The model has been shown to successfully model the recrystallization rate seen in the additively manufactured IN718 parts when both are fit to the JMAK equation, as well as preserving the grain size and aspect ratio expected at each step of the annealing process.

2:40 PM  Invited
Characterizing the Length Dependence of High-Peierls-Stress Dislocations’ Mobility in BCC Crystals under Deformation at Finite Temperature from the Atomistic to the Mesoscale: Liming Xiong¹; ¹Iowa State University
    It remains a challenge using single-scale approaches to fully address how the collective behavior of atomic-level kinks dictates the mesoscale dislocation dynamics and in turn, the macroscale performance of plastically deformed high-Peierls-barrier materials at finite temperature. Here we present a finite temperature coarse-grained atomistic approach to meet this challenge. Taking bcc iron and tungsten as model materials, we studied the dynamics of screw dislocation with its lengths ranging from 50nm to 5μm. One major finding is: the mobility of a screw dislocation is neither linearly dependent nor independent on its length when the kink nucleation rate is at the same level as the rate of kink annihilation on a very long dislocation line. This result can be used as a key supplement to a variety of higher scale approaches, such as kinetic Monte Carlo, dislocation dynamics, crystal plasticity, and phase field, and lay them on a firm atomistic foundation.

3:10 PM
Dislocation Dipole Study on Material Hardening/Softening: Abu Siddique¹; Tariq Khraishi¹; Hojun Lim²; ¹University of New Mexico; ²Sandia National Laboratories
    Dislocation dynamics simulations often reveal interesting phenomena in regards to material deformation which may not be captured by experiments. In this work, we investigate the effect of dislocation dipoles on plastic material properties under different dipole configurations (i.e. size of the dislocation sources, the distance between active glide planes, and the signs of two dislocations) using a Discrete Dislocation Dynamics code. The simulations show that a dipole is causing hardening effect when the Burgers vectors of the dislocations forming a dipole are of opposite sign and causing a softening effect when they are of the same sign. The distance between the two neighboring dislocations in a dipole was affecting the elastic limit of the material and not the flow stress of the material. Such hardening or flow stress results as in this study can be incorporated in higher scale modeling work.

3:30 PM  Cancelled
Continuum Dislocation Dynamics with Junction Reactions: Computational Modeling and Preliminary Results: Kyle Starkey¹; Anter El-Azab¹; ¹Purdue University
    We aim to understand how dislocations interact in metallic crystals and how these interactions give rise to the complex microstructures and hardening behavior in metals. Continuum dislocation dynamics represent the state-of-the-art modeling approach in this field. Currently, there is a lack of understanding as to how to include dislocation-dislocation interactions into continuum dislocation dynamics models. We present a continuum model for dislocations based on our vector density representation of dislocation lines in the context of dislocation reactions. Dislocations on each slip system are open line segments which are collectively closed on junction nodes such that Frank’s second rule at each node is satisfied. In this model, we represent the dislocation network with bundles of dislocation lines along with point densities to represent junction nodes. We present various graph-theoretic ideas that are naturally incorporated into the model from the addition of a junction point density. Several preliminary simulations will be shown.

3:50 PM
Advancements in Discrete Dislocation Modeling of Slip Transmission through Equilibrium and Non-equilibrium Grain Boundaries: Darshan Bamney¹; Laurent Capolungo²; Douglas Spearot¹; ¹University of Florida; ²Los Alamos National Laboratory
    Recent advancements are presented in the modeling of lattice dislocation transmission through equilibrium and non-equilibrium grain boundaries (GBs) using the discrete dislocation dynamics (DDD) approach. First, the disclination structural unit model (DSUM) is employed to construct disclination dipole-based representations of secondary GB dislocation content in DDD. The equilibrium structures predicted by DSUM are systematically disrupted to simulate non-equilibrium interfaces with extrinsic GB dislocation complexes, mimicking GB damage. Then, a slip transmission algorithm is developed to handle the propagation of dislocations across interfaces. The algorithm uses dislocation propagation criteria based on a combination of geometric parameters and power dissipation and is calibrated using results from atomistic simulations. Additionally, treatment of the evolution of residual dislocations at the interface is included in the framework. Finally, these developments are leveraged to perform a parametric study at the mesoscale to evaluate the influence of the long-range fields generated by non-equilibrium GBs on slip transmission.