Bulletin of the American Physical Society
2021 Annual Meeting of the APS Four Corners Section
Volume 66, Number 11
Friday–Saturday, October 8–9, 2021; Virtual; Mountain Daylight Time
Session J06: Modeling and Computational Approaches in Material Science |
Hide Abstracts |
Chair: Ian Leahy, University of Colorado Boulder |
Saturday, October 9, 2021 10:45AM - 11:09AM |
J06.00001: Neural Networks and Machine Learning in Condensed Matter Physics Research: Two Examples Invited Speaker: John Colton Neural networks are a powerful modern tool whereby a computer learns to make predictions based on a set of training data. This talk will discuss some of the fundamental aspects of neural networks in the context of two recent examples from my research. The first application is predicting the temperature of quantum dots based on measurements of the quantum dot photoluminescence, in order to potentially create small localized temperature sensors. The second application is predicting the properties (frequency, mode pattern) of a specific resonant mode of a cylindrical metal shell (a "cavity") filled with bits of high dielectric material, in order to design better microwave resonators for e.g. use in spin resonance experiments. [Preview Abstract] |
Saturday, October 9, 2021 11:09AM - 11:21AM |
J06.00002: Improved Uncertainty Quantification of Interatomic Models using Sloppy Model Analysis Dylan Bailey, Mark Transtrum, Yonatan Kurniawan, Cody Petrie Interatomic models (IMs) are useful in predicting material properties of interest. However, The development of a single IM can take months to years and relies on expert intuition, and is then normally only valid for a singular application. Extending existing IMs to new applications is an active area of research. Uncertainty quantification (UQ) can help to inform us how well an IM predicts in a new regime to which it was not trained. The predictions of many IMs are insensitive to large, coordinated changes in many of their parameters, a phenomenon known as sloppiness. Our previous work has shown that sloppiness poses challenges both for the implementation and interpretation of traditional UQ analysis. To address these issues we use the Manifold Boundary Approximation Method to systematically remove sloppy parameters and perform UQ on the reduced model. I report on our progress on a model of MoS$_2$ monolayer using the Stillinger-Weber potential. We find that in comparison to the original the confidence region of the reduced model becomes less sensitive to the choice of confidence level. [Preview Abstract] |
Saturday, October 9, 2021 11:21AM - 11:33AM |
J06.00003: Intuitive and Dynamic Software for muSR Analysis Kevin Petersen Muon spin spectroscopy is a method of investigating the magnetic properties of various types of condensed matter by implanting a beam of spin polarized muons in a small sample of the material and analyzing the precession or relaxation of the muon spin in the local magnetic field. The data we get from these experiments and the programs currently used to visualize or analyze the date can be cumbersome to download and work with. Our project was to create a smaller program that was easier to install but still provided the necessary features and a more intuitive interface with which to analyze the data. [Preview Abstract] |
Saturday, October 9, 2021 11:33AM - 11:45AM |
J06.00004: Non-Linear Optimization for Enhanced Parameter Retrieval in Magnetic Resonance Fingerprinting John Lundstrom, Megan E. Poorman, Andrew Dienstfrey, Kathryn E. Keenan Quantitative Magnetic Resonance Imaging (qMRI) is emerging as a critical tool in medical diagnostics. Methods for qMRI demand long scan times to provide high-quality quantitative maps. Magnetic Resonance Fingerprinting (MRF) is a novel method for simultaneous multi-parametric qMRI, which is five times faster and feasibly more accurate than traditional qMRI methods. The current MRF analysis pipeline uses dictionary matching to infer MR parameters. The algorithm compares under-sampled MRI data to a model-based dictionary of unique MR parameter combinations; quantitative MRF maps are determined by the closest match. A discrete table of parameter values constrains dictionary matching, whereas in reality, these values are continuous. To improve conventional dictionary based MRF, we add non-linear optimization (NLO) with the goal of obtaining more precise and accurate parameter results. We determine the robustness of NLO with respect to initial condition and noise using a Monte Carlo simulation and obtain novel results, which quantify the error associated with NLO. Estimating uncertainty will improve understanding of MRF function and its utility in a clinical setting. [Preview Abstract] |
Saturday, October 9, 2021 11:45AM - 11:57AM |
J06.00005: Turbulent Hydrodynamic Flow of a Dirac Fluid in a Two Dimensional Solid Mark Watson In the present numerical study we explore the possibility of a turbulent flow in the electric transport of a two dimensional solid, with particular focus on graphene. We use a relativistic hydrodynamic simulation to analyze the flow of the massless charge carriers in a solid with impurities. We find evidence of the possibility of a chaotic and perhaps pre-turbulent flow. Experimental consequences are discussed. [Preview Abstract] |
Saturday, October 9, 2021 11:57AM - 12:09PM |
J06.00006: Challenges for prior selection in sloppy, multiparameter models Yonatan Kurniawan, Cody Petrie, Mark Transtrum, Kinamo Williams Uncertainty quantification is an important tool for assessing the credibility of a model. However, multi-parameter models from many fields are often insensitive to large, coordinated changes in many of their parameter combinations, a phenomenon known as sloppiness. Often, confidence regions in a model's parameter space do not close and the range of physically allowed parameter values is effectively infinite. In Markov Chain Monte Carlo (MCMC) sampling of a Bayesian posterior, sloppiness leads to a phenomenon known as parameter evaporation, in which the samples prefer the sub-optimal region at some sampling temperature. I demonstrate this phenomenon on several illustrative examples and discuss how the choice of prior in the Bayesian posterior can bias sampling results in subtle, unexpected ways. [Preview Abstract] |
Saturday, October 9, 2021 12:09PM - 12:21PM |
J06.00007: Developing an Ab Initio-Kinetic Model for the Prediction of Passivation Behavior Rachel Gorelik, Peter Crozier, Arunima Singh With the US economy incurring {\$}200 billion annual losses due to corrosion, there is still a significant need for effective a priori models which enable the prediction of materials' corrosion resistance, particularly within the field of materials discovery. While most current predictive models use fully empirical parameters, there is a value in the development of kinetic models which utilize solely quantum-mechanics-derived inputs. As a first step in this direction, we therefore develop an ab initio-kinetic model by applying the Pilling-Bedworth Rule (PBR), a metric commonly used to predict the passivation protectiveness of a given material based on mechanical driving forces. With a previously developed ab initio methodology for determining the passivation products of any material, we automate a methodology utilizing the PBR for all materials currently in the Materials Project database, and furthermore we extend it to include materials which are predicted to form more than one passivation product on their surfaces. Upon development, this model can serve as a preliminary, low-cost screening step which can be applied to any material for predicting its electrochemical stability. [Preview Abstract] |
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