arXiv subjects

Venkatasubramanian Viswanathan

Publications and source records attributed to Venkatasubramanian Viswanathan.

At least 37 records · Page 2Linked to original sources

arXiv2023

Accurate Surface and Finite Temperature Bulk Properties of Lithium Metal at Large Scales using Machine Learning Interaction Potentials

The properties of lithium metal are key parameters in the design of lithium ion and lithium metal batteries. They are difficult to probe experimentally due to the high reactivity and low melting point of lithium as well as the microscopic scales at which lithium exists in batteries where it is found to have enhanced strength, with implications for dendrite suppression strategies. Computationally, there is a lack of empirical potentials that are consistently quantitatively accurate across all properties and ab-initio calculations are too costly. In this work, we train Machine Learning Interaction Potentials (MLIPs) on Density Functional Theory (DFT) data to state-of-the-art accuracy in reproducing experimental and ab-initio results across a wide range of simulations at large length and time scales. We accurately predict thermodynamic properties, phonon spectra, temperature dependence of elastic constants and various surface properties inaccessible using DFT. We establish that there exists a Bell-Evans-Polanyi relation correlating the self-adsorption energy and the minimum surface diffusion barrier for high Miller index facets.

cond-mat.mtrl-sci↗

arXiv2023

Chemellia: An Ecosystem for Atomistic Scientific Machine Learning

Chemellia is an open-source framework for atomistic machine learning in the Julia programming language. The framework takes advantage of Julia's high speed as well as the ability to share and reuse code and interfaces through the paradigm of multiple dispatch. Chemellia is designed to make use of existing interfaces and avoid ``reinventing the wheel'' wherever possible. A key aspect of the Chemellia ecosystem is the ChemistryFeaturization interface for defining and encoding features -- it is designed to maximize interoperability between featurization schemes and elements thereof, to maintain provenance of encoded features, and to ensure easy decodability and reconfigurability to enable feature engineering experiments. This embodies the overall design principles of the Chemellia ecosystem: separation of concerns, interoperability, and transparency. We illustrate these principles by discussing the implementation of crystal graph convolutional neural networks for material property prediction.

cs.CE↗

arXiv2023

Importance of equivariant and invariant symmetries for fluid flow modeling

Graph neural networks (GNNs) have shown promise in learning unstructured mesh-based simulations of physical systems, including fluid dynamics. In tandem, geometric deep learning principles have informed the development of equivariant architectures respecting underlying physical symmetries. However, the effect of rotational equivariance in modeling fluids remains unclear. We build a multi-scale equivariant GNN to forecast fluid flow and study the effect of modeling invariant and non-invariant representations of the flow state. We evaluate the model performance of several equivariant and non-equivariant architectures on predicting the evolution of two fluid flows, flow around a cylinder and buoyancy-driven shear flow, to understand the effect of equivariance and invariance on data-driven modeling approaches. Our results show that modeling invariant quantities produces more accurate long-term predictions and that these invariant quantities may be learned from the velocity field using a data-driven encoder.

physics.flu-dyn↗

arXiv2023

Anomalous interfacial electron transfer kinetics in twisted trilayer graphene caused by layer-specific localization

Interfacial electron-transfer (ET) reactions underpin the interconversion of electrical and chemical energy. Pioneering experiments showed that the ET rate depends on the Fermi Dirac distribution of the electronic density of states (DOS) of the electrode, formalized in the Marcus Hush Chidsey (MHC) model. Here, by controlling interlayer twists in well-defined trilayergraphene moires, we show that ET rates are strikingly dependent on electronic localization in each atomic layer, and not the overall DOS. The large degree of tunability inherent to moire electrodes leads to local ET kinetics that range over three orders of magnitude across different constructions of only three atomic layers, even exceeding rates at bulk metals. Our results demonstrate that beyond the ensemble DOS, electronic localization is critical in facilitating interfacial ET, with implications for understanding the origin of high interfacial reactivity typically exhibited by defects at electrode electrolyte interfaces.

cond-mat.mes-hall↗

arXiv2023

Multiscale Graph Neural Network Autoencoders for Interpretable Scientific Machine Learning

The goal of this work is to address two limitations in autoencoder-based models: latent space interpretability and compatibility with unstructured meshes. This is accomplished here with the development of a novel graph neural network (GNN) autoencoding architecture with demonstrations on complex fluid flow applications. To address the first goal of interpretability, the GNN autoencoder achieves reduction in the number nodes in the encoding stage through an adaptive graph reduction procedure. This reduction procedure essentially amounts to flowfield-conditioned node sampling and sensor identification, and produces interpretable latent graph representations tailored to the flowfield reconstruction task in the form of so-called masked fields. These masked fields allow the user to (a) visualize where in physical space a given latent graph is active, and (b) interpret the time-evolution of the latent graph connectivity in accordance with the time-evolution of unsteady flow features (e.g. recirculation zones, shear layers) in the domain. To address the goal of unstructured mesh compatibility, the autoencoding architecture utilizes a series of multi-scale message passing (MMP) layers, each of which models information exchange among node neighborhoods at various lengthscales. The MMP layer, which augments standard single-scale message passing with learnable coarsening operations, allows the decoder to more efficiently reconstruct the flowfield from the identified regions in the masked fields. Analysis of latent graphs produced by the autoencoder for various model settings are conducted using using unstructured snapshot data sourced from large-eddy simulations in a backward-facing step (BFS) flow configuration with an OpenFOAM-based flow solver at high Reynolds numbers.

cs.LG↗

arXiv2022

Nonequilibrium Electrochemical Phase Maps: Beyond Butler-Volmer Kinetics

Electrochemical kinetics at electrode-electrolyte interfaces are crucial to understand high-rate behavior of energy storage devices. Phase transformation of electrodes is typically treated under equilibrium thermodynamic conditions, while realistic operation is at finite rates. Analyzing phase transformations under nonequilibrium conditions requires integrating nonlinear electrochemical kinetic models with thermodynamic models. This had only previously been demonstrated for Butler-Volmer kinetics, where it can be done analytically. In this work, we develop a kinetic modeling package in the Julia language capable of efficient numerical inversion of rate relationships for general kinetic models using automatic differentiation. We demonstrate building nonequilibrium phase maps, including for models such as Marcus-Hush-Chidsey that require computation of an integral, and also discuss the impact of a variety of assumptions and model parameters (such as temperature, reorganization energy, activity, and ideal solution interaction energy), particularly on high-rate phase behavior. Even for a fixed set of parameters, the magnitude of the critical current can vary by in excess of a factor of two amongst kinetic models.

cond-mat.mtrl-sci↗

arXiv2022

By how much can closed-loop frameworks accelerate computational materials discovery?

The implementation of automation and machine learning surrogatization within closed-loop computational workflows is an increasingly popular approach to accelerate materials discovery. However, the scale of the speedup associated with this paradigm shift from traditional manual approaches remains an open question. In this work, we rigorously quantify the acceleration from each of the components within a closed-loop framework for material hypothesis evaluation by identifying four distinct sources of speedup: (1) task automation, (2) calculation runtime improvements, (3) sequential learning-driven design space search, and (4) surrogatization of expensive simulations with machine learning models. This is done using a time-keeping ledger to record runs of automated software and corresponding manual computational experiments within the context of electrocatalysis. From a combination of the first three sources of acceleration, we estimate that overall hypothesis evaluation time can be reduced by over 90%, i.e., achieving a speedup of $\sim$$10\times$. Further, by introducing surrogatization into the loop, we estimate that the design time can be reduced by over 95%, i.e., achieving a speedup of $\sim$$15$-$20\times$. Our findings present a clear value proposition for utilizing closed-loop approaches for accelerating materials discovery.

cond-mat.mtrl-sci↗

arXiv2022

Differentiable physics-enabled closure modeling for Burgers' turbulence

Data-driven turbulence modeling is experiencing a surge in interest following algorithmic and hardware developments in the data sciences. We discuss an approach using the differentiable physics paradigm that combines known physics with machine learning to develop closure models for Burgers' turbulence. We consider the 1D Burgers system as a prototypical test problem for modeling the unresolved terms in advection-dominated turbulence problems. We train a series of models that incorporate varying degrees of physical assumptions on an a posteriori loss function to test the efficacy of models across a range of system parameters, including viscosity, time, and grid resolution. We find that constraining models with inductive biases in the form of partial differential equations that contain known physics or existing closure approaches produces highly data-efficient, accurate, and generalizable models, outperforming state-of-the-art baselines. Addition of structure in the form of physics information also brings a level of interpretability to the models, potentially offering a stepping stone to the future of closure modeling.

physics.flu-dyn↗

arXiv2022

AutoMat: Accelerated Computational Electrochemical systems Discovery

Large-scale electrification is vital to addressing the climate crisis, but several scientific and technological challenges remain to fully electrify both the chemical industry and transportation. In both of these areas, new electrochemical materials will be critical, but their development currently relies heavily on human-time-intensive experimental trial and error and computationally expensive first-principles, meso-scale and continuum simulations. We present an automated workflow, AutoMat, that accelerates these computational steps by introducing both automated input generation and management of simulations across scales from first principles to continuum device modeling. Furthermore, we show how to seamlessly integrate multi-fidelity predictions such as machine learning surrogates or automated robotic experiments "in-the-loop". The automated framework is implemented with design space search techniques to dramatically accelerate the overall materials discovery pipeline by implicitly learning design features that optimize device performance across several metrics. We discuss the benefits of AutoMat using examples in electrocatalysis and energy storage and highlight lessons learned.

cond-mat.mtrl-sci↗

arXiv2022

Effect of disorder and doping on electronic structure and diffusion properties of Li$_{3}$V$_{2}$O$_{5}$

V$_{2}$O$_{5}$ in its $ω$ phase (Li$_{3}$V$_{2}$O$_{5}$) with excess lithium is a potential alternative to the graphite anode for lithium-ion batteries at low temperature and fast charging conditions due to its safer voltage (0.6 V vs Li$^{+}$/Li(s)) and high lithium transport rate. In-operando cationic disorder, as observed in most ordered materials, can produce significant changes in charge compensation mechanisms, anionic activity, lithium diffusion and operational voltages. In this work, we report the variation in structural distortion, electronic structure and migration barrier accompanied by disorder using first-principles calculations. Due to segregation of lithium atoms in the disordered state, we observe greater distortion, emergence of metallic behaviour and potential anionic activity from non-bonding oxygen states near the Fermi level. Redox capacity can be tuned by doping with 3d metals which can adjust the participating cationic states, and by fluorine substitution which can stabilize or suppress anionic states. Moreover, suppression of anionic activity is found to decrease structural distortion, crucial for mitigating voltage fade and hysteresis. Diffusion barrier calculations in the presence of disorder indicate the activation of the remaining 3D-paths for lithium hopping which are unavailable in the ordered configuration, explaining its fast-charging ability observed in experiments.

cond-mat.mtrl-sci↗

arXiv2022

Super-linear Scaling Behavior for Electric Vehicle Chargers and Road Map to Addressing the Infrastructure Gap

Enabling widespread electric vehicle (EV) adoption requires substantial build-out of charging infrastructure in the coming decade. We formulate the charging infrastructure needs as a scaling analysis problem and use it to estimate the EV infrastructure needs of the US at a county-level resolution. Surprisingly, we find that the current EV infrastructure deployment scales super-linearly with population, deviating from the sub-linear scaling of gasoline stations and other infrastructure. We discuss how this demonstrates the infancy of EV station abundance compared to other mature transportation infrastructures. By considering the power delivery of existing gasoline stations, and appropriate EV efficiencies, we estimate the EV infrastructure gap at the county level, providing a road map for future EV infrastructure expansion. Our reliance on scaling analysis allows us to make a unique forecast in this domain.

econ.EM↗

arXiv2022

Chemomechanics: friend or foe of the "AND problem" of solid-state batteries?

Solid electrolytes are widely considered as the enabler of lithium metal anodes for safe, durable, and high energy density rechargeable lithium-ion batteries. Despite the promise, failure mechanisms associated with solid-state batteries are not well-established, largely due to limited understanding of the chemomechanical factors governing them. We focus on the recent developments in understanding solid-state aspects including the effects of mechanical stresses, constitutive relations, fracture, and void formation, and outline the gaps in the literature. We also provide an overview of the manufacturing and processing of solid-state batteries in relation to chemomechanics. The gaps identified provide concrete directions towards the rational design and development of failure-resistant solid-state batteries.

physics.app-ph↗

arXiv2022

Score-Based Generative Models for Molecule Generation

Recent advances in generative models have made exploring design spaces easier for de novo molecule generation. However, popular generative models like GANs and normalizing flows face challenges such as training instabilities due to adversarial training and architectural constraints, respectively. Score-based generative models sidestep these challenges by modelling the gradient of the log probability density using a score function approximation, as opposed to modelling the density function directly, and sampling from it using annealed Langevin Dynamics. We believe that score-based generative models could open up new opportunities in molecule generation due to their architectural flexibility, such as replacing the score function with an SE(3) equivariant model. In this work, we lay the foundations by testing the efficacy of score-based models for molecule generation. We train a Transformer-based score function on Self-Referencing Embedded Strings (SELFIES) representations of 1.5 million samples from the ZINC dataset and use the Moses benchmarking framework to evaluate the generated samples on a suite of metrics.

cs.LG↗

arXiv2021

Principles of the Battery Data Genome

Electrochemical energy storage is central to modern society -- from consumer electronics to electrified transportation and the power grid. It is no longer just a convenience but a critical enabler of the transition to a resilient, low-carbon economy. The large pluralistic battery research and development community serving these needs has evolved into diverse specialties spanning materials discovery, battery chemistry, design innovation, scale-up, manufacturing and deployment. Despite the maturity and the impact of battery science and technology, the data and software practices among these disparate groups are far behind the state-of-the-art in other fields (e.g. drug discovery), which have enjoyed significant increases in the rate of innovation. Incremental performance gains and lost research productivity, which are the consequences, retard innovation and societal progress. Examples span every field of battery research , from the slow and iterative nature of materials discovery, to the repeated and time-consuming performance testing of cells and the mitigation of degradation and failures. The fundamental issue is that modern data science methods require large amounts of data and the battery community lacks the requisite scalable, standardized data hubs required for immediate use of these approaches. Lack of uniform data practices is a central barrier to the scale problem. In this perspective we identify the data- and software-sharing gaps and propose the unifying principles and tools needed to build a robust community of data hubs, which provide flexible sharing formats to address diverse needs. The Battery Data Genome is offered as a data-centric initiative that will enable the transformative acceleration of battery science and technology, and will ultimately serve as a catalyst to revolutionize our approach to innovation.

physics.soc-ph↗

arXiv2021

Autonomous optimization of nonaqueous battery electrolytes via robotic experimentation and machine learning

In this work, we introduce a novel workflow that couples robotics to machine-learning for efficient optimization of a non-aqueous battery electrolyte. A custom-built automated experiment named "Clio" is coupled to Dragonfly - a Bayesian optimization-based experiment planner. Clio autonomously optimizes electrolyte conductivity over a single-salt, ternary solvent design space. Using this workflow, we identify 6 fast-charging electrolytes in 2 work-days and 42 experiments (compared with 60 days using exhaustive search of the 1000 possible candidates, or 6 days assuming only 10% of candidates are evaluated). Our method finds the highest reported conductivity electrolyte in a design space heavily explored by previous literature, converging on a high-conductivity mixture that demonstrates subtle electrolyte chemical physics.

cs.LG↗

arXiv2021

Validation and parameterization of a novel physics-constrained neural dynamics model applied to turbulent fluid flow

In fluid physics, data-driven models to enhance or accelerate solution methods are becoming increasingly popular for many application domains, such as alternatives to turbulence closures, system surrogates, or for new physics discovery. In the context of reduced order models of high-dimensional time-dependent fluid systems, machine learning methods grant the benefit of automated learning from data, but the burden of a model lies on its reduced-order representation of both the fluid state and physical dynamics. In this work, we build a physics-constrained, data-driven reduced order model for the Navier-Stokes equations to approximate spatio-temporal turbulent fluid dynamics. The model design choices mimic numerical and physical constraints by, for example, implicitly enforcing the incompressibility constraint and utilizing continuous Neural Ordinary Differential Equations for tracking the evolution of the differential equation. We demonstrate this technique on three-dimensional, moderate Reynolds number turbulent fluid flow. In assessing the statistical quality and characteristics of the machine-learned model through rigorous diagnostic tests, we find that our model is capable of reconstructing the dynamics of the flow over large integral timescales, favoring accuracy at the larger length scales. More significantly, comprehensive diagnostics suggest that physically-interpretable model parameters, corresponding to the representations of the fluid state and dynamics, have attributable and quantifiable impact on the quality of the model predictions and computational complexity.

physics.flu-dyn↗

arXiv2021

Cell-Level State of Charge Estimation for Battery Packs Under Minimal Sensing

This manuscript presents an algorithm for individual Lithium-ion (Li-ion) battery cell state of charge (SOC) estimation in a large-scale battery pack under minimal sensing, where only pack-level voltage and current are measured. For battery packs consisting of up to thousands of cells in electric vehicle or stationary energy storage applications, it is desirable to estimate individual cell SOCs without cell local measurements in order to reduce sensing costs. Mathematically, pure series connected cells yield dynamics given by ordinary differential equations under classical full voltage sensing. In contrast, parallel--series connected battery packs are evidently more challenging because the dynamics are governed by a nonlinear differential--algebraic equations (DAE) system. The majority of the conventional studies on SOC estimation for battery packs benefit from idealizing the pack as a lumped single cell which ultimately lose track of cell-level conditions and are blind to potential risks of cell-level over-charge and over-discharge. This work explicitly models a battery pack with high fidelity cell-by-cell resolution based on the interconnection of single cell models, and examines the observability of cell-level state with only pack-level measurements. A DAE-based state observer with linear output error injection is formulated, where the individual cell SOC and current can be reconstructed from minimal number of pack sensing. The mathematically guaranteed asymptotic convergence of differential and algebraic state estimates is established by considering local Lipschitz continuity property of system nonlinearities. Simulation results for Graphite/NMC cells illustrate convergence for cell SOCs, currents, and voltages.

eess.SY↗

arXiv2021

Differentiable Physics: A Position Piece

Differentiable physics provides a new approach for modeling and understanding the physical systems by pairing the new technology of differentiable programming with classical numerical methods for physical simulation. We survey the rapidly growing literature of differentiable physics techniques and highlight methods for parameter estimation, learning representations, solving differential equations, and developing what we call scientific foundation models using data and inductive priors. We argue that differentiable physics offers a new paradigm for modeling physical phenomena by combining classical analytic solutions with numerical methodology using the bridge of differentiable programming.

cs.LG↗