Bulletin of the American Physical Society
APS March Meeting 2019
Volume 64, Number 2
Monday–Friday, March 4–8, 2019; Boston, Massachusetts
Session Y56: Machine Learning Approaches to Understanding Bulk Metallic Glasses and Other Amorphous MaterialsFocus
|
Hide Abstracts |
Sponsoring Units: GSNP GSOFT Chair: Corey O'Hern, Yale Univ Room: BCEC 255 |
Friday, March 8, 2019 11:15AM - 11:27AM |
Y56.00001: Machine Learning Stress Overshoot of Amorphous Solids Shiyun Zhang, Wen Zheng, Ning Xu When undergoing quasistatic shear, slow-quenched glasses exhibit an overshoot in shear stress, which decreases with the increase of quench rate and eventually evolves into a smooth crossover. However, the structure of glasses does not exhibit significant quench rate dependence. It thus remains unclear what determines the emergence of stress overshoot. Here, inspired by image recognition, we propose that stress overshoot of amorphous solids can be predicted from structures of solids using machine learning methods. Our results show that pure geometrical quantities such as local coordination number, bond orientational order and voronoi cell volume, which seem to weakly correlate with soft spots, have displayed very high predictive power of stabilities of amorphous solids, when combined with coarse-grain and machine learning methods. Besides, the stress overshoot in pinned systems suggests that our recently defined order parameter from the normal modes of vibration is a more general structure descriptor to identify stress overshoot. The high accuracy in identifying stress overshoot may imply that machine learning methods successfully capture the spatial correlation of the descriptor in some complicated way. |
Friday, March 8, 2019 11:27AM - 11:39AM |
Y56.00002: Interplay of softness and rearrangements during avalanche propagation Ge Zhang, Sean Ridout, Andrea Liu Disordered solids yield at a common shear strain of about 3%, but the behavior beyond yield is different for different systems and for systems with different histories. Foams can deform indefinitely without fracturing, many systems exhibit crackling noise or avalanche behavior, and still others exhibit shear banding and brittle fracture. Here we study an athermal, jammed packing of Hertzian particles that are sheared quasistatically. We identify the stress drops associated with rearrangements and then use steepest descent dynamics to study the evolution of the avalanches. We find that the avalanches consist of localized events that appear sequentially in well-separated locations of the sample. To understand this behavior, we exploit a machine-learning approach that has been developed to correlate local structure with dynamics in glassy systems. Following earlier work, we define a quantity softness that correlates with dynamical events during the avalanche process. We study the interplay of softness and dynamics: particles with higher softness are more likely to shift, while dynamical events affect local structure and hence softness. This interplay gives insight into the avalanche process. |
Friday, March 8, 2019 11:39AM - 11:51AM |
Y56.00003: A Structural Measure of Effective- (Fictive-) Temperature and its Basis in Statistical Mechanics Michael Falk, Darius D Alix-Williams The concept of fictive-temperature has long been utilized to characterize the processing dependence of glass structure, and has recently been shown to be predictive of metallic glass ductility. Some theories have hypothesized that it is actually a real temperature related to the configurational degrees of freedom of the glass, i.e. an "effective-temperature," notably the shear-transformation-zone (STZ) and soft-glassy-rheology (SGR) theories. Here we derive a thermodynamic integration scheme for calculating effective-temperature based on a 2-temperature hypothesis. To test this scheme we simulate a binary Cu-Zr metallic glass modeled with an EAM potential. Measures of the energy fluctuations associated with both the fast and slow degrees of freedom are measured during the glass quench. The resulting effective-temperature is consistent with estimates of fictive-temperature obtained from simulation in more heuristic ways. The results indicate that effective-temperature can be understood as a purely structural quantity. The method provides a means to measure the effective-temperature in the absence of fluctuations induced by shear and without resorting computationally expensive and impractical methods for explicitly measuring the configurational entropy. |
Friday, March 8, 2019 11:51AM - 12:03PM |
Y56.00004: Interaction potentials for bulk metallic glasses that can generate both brittle and ductile mechanical response Aya Nawano, Jan Schroers, Mark Shattuck, Corey Shane O'Hern Bulk metallic glasses (BMGs) have desirable mechanical properties such as high yield strength and elasticity compared to conventional alloys. However, BMGs are typically brittle, which limits their viability for structural applications. We perform molecular dynamics simulations to understand the ductility of model glass formers that interact via the Lennard-Jones, Stillinger-Weber, and embedded atom method potentials. We prepare binary BMGs over a range of cooling rates and perform athermal quasi-static uniaxial tension tests. We correlate the ductility with the fictive temperature, depth in the potential energy landscape, and measures of local structural order. We show that we can prepare samples that span a wide range of mechanical responses for all of the interaction potentials that we study. We also present a phenomenological spring network model that describes brittle and ductile response in terms of the number of springs that have broken and reformed in response to applied strain. We identify the parameters in the model that control the behavior of the stress versus strain curve, which allows us to achieve quantitative agreement with the results from the simulation of uniaxial tension. |
Friday, March 8, 2019 12:03PM - 12:15PM |
Y56.00005: Correlations in the shear flow of athermal amorphous solids: A principal component analysis Celine Ruscher, Joerg G Rottler Machine learning methods are increasingly being applied to problems in statistical physics, because they unveil aspects that would generally be neglected by traditional approaches. Here we apply principal component analysis, a method frequently used in image processing and unsupervised machine learning, to study the passage from the elastic to plastic flow regime in amorphous materials. Sets of particle displacements are obtained from simulations of a 2D amorphous model system in steady shear flow at different shear rates in the athermal limit. PCA produces a low-dimensional representation of the data, in which the principal directions clearly identify distinct differences between elastic (i.e. reversible) and plastic deformation. When deformation is accumulated over larger strains, shear localizes along bands, and PCA provides a quantitative measure of the increased degree of anisotropy in the flow patterns. We suggest that PCA can be a useful analysis technique that complements a traditional statistical description via correlation functions. |
Friday, March 8, 2019 12:15PM - 12:27PM |
Y56.00006: Analyzing the Internal Structure of Metallic Glasses through X-ray Absorption Fine Structure (XAFS) Spectroscopy Hanyu Zhang, Jennifer Carter, Harold Connamacher This project aims to independently find the structure of metallic glasses through theoretical and experimental measures to see if current theoretical simulations yield the same structures as measured through experimental techniques. This is achieved by analyzing the accuracy of current molecular dynamics (MD) simulations to predict the amorphous structure of metallic glasses as characterized by x-ray absorption fine structure (XAFS) spectroscopy experiments. The influence of composition and processing techniques on the final structure is being explored on metallic glasses made from Ni, Co, Ta, and Nb. This project utilizes Large-scale Atomic/Molecular Massively Parallel Simulator (http://lammps.sandia.gov/), a MD software, to build the structures and Larch (Matthew Newville 2013) for data analysis. Used as a Python package, Larch gives us full control of the usage and implementation of our code. |
Friday, March 8, 2019 12:27PM - 12:39PM |
Y56.00007: Local Yield Stress Analysis in Simulated 3D Glasses Dihui Ruan, Sylvain Patinet, Michael Falk The ‘Local Yield Stress’ (LYS) method was developed to characterize the local structure of a atomistic model of a glass in a way that provides insight into its plastic response. The local yield stress is characterized as the incremental stress needed to induce a structural instability when an outer shell surrounding of a patch of atoms is deformed affinely along various shear orientations, while allowing the structure at the core to rearrange via athermal quasi-static (AQS) simulation. Here we generalize the LYS method to three-dimensional systems. We describe a method to democratically sample local shear orientations and triaxialities. By applying the LYS analysis to as-quenched glasses, we are able to identify the statistics of the local yield stress distribution. In order to assess to predictability and persistence of this measure, a correlation is computed between these local yield stresses and the observed plastic events when the glasses are undergoing shear deformation within in AQS simulations. |
Friday, March 8, 2019 12:39PM - 1:15PM |
Y56.00008: Bulk metallic glass design: What properties determine the glass-forming ability of multi-component alloys? Invited Speaker: Yuan-Chao Hu Bulk metallic glasses (BMGs) possess a number of important properties, such as high strength and thermoplastic formability, which stem from the fact that they are structurally disordered in contrast to crystalline metals. Materials scientists have identified several features that are correlated with the glass-forming ability (GFA) of alloys. For example, good glass-formers are typically multi-component alloys composed of elements with atomic radii that differ by more than 10%. Most BMGs also possess a negative heat of mixing, which disfavors clustering of like atoms and hinders phase separation. However, researchers have not been able to a priori predict a new BMG-forming alloy. In this work, we perform computational studies of binary alloys to understand the relative contributions of geometric frustration and energetic frustration in determining the GFA. From a database of the heats of mixing and cohesive energies of binary atomic systems with atoms A and B, we show that most binary alloys follow a Berthelot combining rule, εAB=(εAAεBB)½ , where εAB, εAA and εBB are the depths of the attractive energy for pair interactions between AB, AA, and BB. We employ this mixing rule in molecular dynamics simulations of binary Lennard-Jones mixtures of atoms with equal sizes, but different cohesive energies. We measure the critical cooling rates of binary systems over the full range of cohesive energies and number fraction fA of A and 1-fA of B atoms. We show that good glass formers satisfy εAA>εAB>εBB when fA<fB. We find that good glass-forming ability is determined by the conditions εBB<<εAA and fA<fB, and not correlated with the magnitude of the heat of mixing. In future studies, we will identify the variables that control the glass-forming ability in ternary alloys with atoms of different sizes and cohesive energies. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700