Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Fractional operators”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 541 records · Page 30

Mobile Sorption-based Thermal Battery for Harvesting Low-Temperature Geothermal Energy

Around 20% of the total primary energy in the United States is consumed for thermal demands of buildings such as space cooling, dehumidification, and space heating (EIA 2018). Low-temperature geothermal energy is abundant and can effectively satisfy buildings’ thermal demands. However, low-temperature geothermal energy is underutilized because the energy density of geothermal fluid is too low to justify the costs associated with transporting it between existing geothermal resources and buildings. The mobile sorption-based thermal battery (MSTB) system has been developed using three-phase (i.e., vapor–liquid, solution–solid, crystal) sorption technology to harvest low-temperature heat and store it with a much higher energy density than the geothermal fluid. The energy density of salt crystals is over six times higher than geothermal fluid, which makes long-distance transportation of salt crystals economically feasible. Salt crystals can be used to dehumidify air or provide space cooling in buildings, which alleviates peak demand on the electricity grid by offsetting electricity use for these end uses. This helps improve the grid’s stability and resilience. High-energy storage density, fast crystallization, and dissolution of salt crystals are all critical to the viability and performance of the MSTB system. Therefore, the design and operation of MSTB systems need to ensure effective generation and dissolution of salt crystals inside the MSTB. To achieve this target, this seedling project developed an experimental apparatus for characterizing the crystallization and dissolution processes. The energy density and potential latent cooling capacity of the MSTB are also evaluated based on lab test results. The crystallization results showed that the generated lithium chloride hydrate crystals are fluffy, the crystallization process lasts about 50 min, and the maximum crystal fraction (i.e., the ratio of crystal mass to the mass in the MSTB) can be up to 51.1% of the total mass in the MSTB at a solution flow rate of 1.58 g/s. The dissolution results show that the salt crystals in the MSTB can be fully dissolved within 15–28 min, based on different test conditions. Reducing solution flow rate and cooling water temperature can achieve increased energy storage density and crystal fraction. While the increase in the discharge rate (i.e., latent cooling capacity for dehumidifying air) is achieved by increasing flow rate and temperature of inlet diluted solution, as well as by using a pump for internal solution circulation, the discharge rate increases by 38%, from 0.95 kW to 1.31 kW. Compared with increasing the inlet solution flow rate, power consumption of salt solution transportation can be reduced by using a pump for internal solution circulation. The crystallization test results also showed that the maximum energy storage density is 981.8 kJ/kg, and the maximum discharge rate of the dissolution tests is ≤1.79 kW. Both are above the target values of 900 kJ/kg and 1.75 kW) for this project. The work reported here proves the feasibility and advancement of the MSTB system, which is helpful to the further study and improvement of the MSTB system.

15 GEOTHERMAL ENERGY↗

Artificial Photosynthesis with Next Generation Molecular Catalysts

The objective of this project was to develop a new class of molecular photocatalysts for CO 2 reduction that will enable breakthrough advances in artificial photosynthesis. A major limitation of existing photocatalytic assemblies is that only a small fraction of absorbed photons are converted into fuel due to inefficiencies in electron transport between the different system components. In the case of molecular catalysts for reduction of CO 2 , highly reducing electron equivalents are required for catalysis. As a result, the electron flux between the photosensitizer and CO 2 reduction catalyst is heavily biased towards the photosensitizer. One method to direct the electron flux towards productive photocatalysis is to design molecular catalysts for CO 2 reduction that operate at more favorable potentials than existing catalysts. The conventional approach is to tune the potential of the catalyst through selection of the supporting ligands, however, this only leads to incremental improvements due to the existence of an unfavorable scaling relationship between the rate and potential of catalysis. Our hypothesis is that the catalytic rate and potential can be simultaneously enhanced using bimetallic catalysts in which one metal improves the potential of catalysis and the second metal increases the rate of catalysis. During the course of this project, a few studies reported some success in improving photocatalytic CO 2 reduction with heterobimetallic complexes, demonstrating the potential of this approach to catalyst design.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Knock-limited combustion of ethanol-, isobutanol-, and 2-methyl-3-buten-2-ol-gasoline blends in a direct-injected spark-ignition engine

Knock-limited combustion of alcohol-gasoline fuel blends was studied in a direct-injection spark-ignition engine. Ethanol, isobutanol (2-methyl-1-propanol), and methylbutenol (2-methyl-3-buten-2-ol) were splash-blended with a blendstock for oxygenate blending (BOB) gasoline. Ethanol was blended in fractions of 10%, 20%, and 30% by volume (E10, E20, and E30). Isobutanol and methylbutenol were blended to match the oxygen weight percentage of the ethanol blends, resulting in blends of 16%, 32%, and 49% (I16, I32, and I49) isobutanol by volume and 18%, 37%, and 56% (M18, M37, and M56) methylbutenol by volume. Neat BOB and gasoline primary reference fuels (PRF) with octane numbers (ON) of 87 (PRF87) and 100 (PRF100) were included for reference. The engine was operated at fixed speed, equivalence ratio, and injection timing while knock-limited spark advance was located across a range of intake pressures. Low-level alcohol blends E10 and M18 had similar knocking behavior and appeared to be slightly more knock-resistant than I16. For mid-level blends, E20 showed further improvement over I32 while M37 showed noticeably better knock resistance, able to match the knock-resistance of PRF100 for loads above 1000 kPa gross indicated mean effective pressure (GIMEP). For high-level blends, the difference between E30 and I49 was similar to their mid-level counterparts. However, M56 showed significant improvement over E30 and I49 throughout the entire range of loads tested. M56 exceeded the knock-resistance of PRF100 for loads above 700 kPa GIMEP. The improved knock-resistance combined with the increased volumetric energy density of methylbutenol resulted in improved fuel consumption and indicated efficiency. The overall results of this study indicate that the order of knock-resistance of these three alcohols, for direct-injection operation with injection during the intake stroke, is isobutanol < ethanol < methylbutenol when compared on an equal oxygen weight basis.

33 ADVANCED PROPULSION SYSTEMS↗

Post-Deployment Characterization of Glass Fiber-Reinforced Thermoset and Thermoplastic Composite Tidal Turbine Blades: Preprint

In 2021, the National Renewable Energy Laboratory (NREL) supported Verdant Power with the most successful tidal energy deployment in U.S. history. Three of their tidal turbines were deployed as part of the Roosevelt Island Tidal Energy project. Initially, the three rotors were manufactured from glass fiber-reinforced epoxy composites. Midway through the deployment, one rotor was replaced with one manufactured at NREL. Instead, it was infused with a novel, infusible thermoplastic resin system. Since the deployment, one epoxy rotor and one thermoplastic rotor were returned to NREL for continued materials and manufacturing research. The two rotors underwent full-scale structural testing before being sectioned and cut into specimens for a variety of manufacturing quality tests, thermos-mechanical characterization, and evaluation of material performance in marine environments, to understand the key differences between the fiberglass reinforced epoxy and Elium composites used for the respective rotors. Matrix burnoff tests showed that the Elium blades had a considerably higher fiber volume fraction compared to the epoxy blades (61% vs. 49%). Environmental aging of the specimens showed that the epoxy laminates absorbed more water over the conditioning period, however, it was determined that the Elium laminates had higher diffusion coefficients, so initially absorbed water faster. Finally, one full epoxy blade and one full Elium blade were conditioned at ambient temperatures for up to 11 months, while taking periodic mass measurements. The datasets were extrapolated out to assume a full 20 year operational life span and it was determined that the blades would not reach full saturation during that time span.

composite manufacturing↗

A ruthenium oxide thermometer for dilution refrigerators operating down to 5 mK

At the lowest temperatures achieved in dilution refrigerators, ruthenium oxide resistance thermometers often saturate and therefore lose their sensitivity. In an effort to extend the range of such temperature sensors, we built a thermometer which maintains sensitivity to 5 mK. A key component of this thermometer is an in situ radio frequency filter which is based on a modern rf absorption material. Here, we show that the use of such a filter is only effective when it is encased in the same rf-tight enclosure as the ruthenium oxide sensor. Our design delivers an attenuation level that is necessary to mitigate the effects of parasitic heating of a fraction of pW present in our circuit. Furthermore, we show that the likely origin of this parasitic heating is the black body radiation present within the experimental space of the refrigerator. We found that the equilibration time of the thermometer increases rapidly as the temperature is lowered; below 5 mK this thermometer becomes impractical because of the prohibitively long equilibration times.

47 OTHER INSTRUMENTATION↗

High-order matrix-free incompressible flow solvers with GPU acceleration and low-order refined preconditioners

In this work, we present a matrix-free flow solver for high-order finite element discretizations of the incompressible Navier-Stokes and Stokes equations with GPU acceleration. For high polynomial degrees, assembling the matrix for the linear systems resulting from the finite element discretization can be prohibitively expensive, both in terms of computational complexity and memory. For this reason, it is necessary to develop matrix-free operators and preconditioners, which can be used to efficiently solve these linear systems without access to the matrix entries themselves. The matrix-free operator evaluations utilize GPU-accelerated sum-factorization techniques to minimize memory movement and maximize throughput. The preconditioners developed in this work are based on a low-order refined methodology with parallel subspace corrections, as described for diffusion problems in [1]. The saddle-point Stokes system is solved using block-preconditioning techniques, which are robust in mesh size, polynomial degree, time step, and viscosity. For the incompressible Navier-Stokes equations, we make use of projection (fractional step) methods, which require Helmholtz and Poisson solves at each time step. The performance of our flow solvers is assessed on several benchmark problems in two and three spatial dimensions.

97 MATHEMATICS AND COMPUTING↗

Spatial quantification of dynamic inter and intra particle crystallographic heterogeneities within lithium ion electrodes

The performance of lithium ion electrodes is hindered by unfavorable chemical heterogeneities that pre-exist or develop during operation. Time-resolved spatial descriptions are needed to understand the link between such heterogeneities and a cell's performance. Here, operando high-resolution X-ray diffraction-computed tomography is used to spatially and temporally quantify crystallographic heterogeneities within and between particles throughout both fresh and degraded Li x Mn 2 O 4 electrodes. This imaging technique facilitates identification of stoichiometric differences between particles and stoichiometric gradients and phase heterogeneities within particles. Through radial quantification of phase fractions, the response of distinct particles to lithiation is found to vary; most particles contain localized regions that transition to rock salt LiMnO 2 within the first cycle. Other particles contain monoclinic Li 2 MnO 3 near the surface and almost pure spinel LixMn2O4 near the core. Following 150 cycles, concentrations of LiMnO 2 and Li 2 MnO 3 significantly increase and widely vary between particles.

25 ENERGY STORAGE↗

Design and optimization of a high-energy optical parametric amplifier for broadband, spectrally incoherent pulses

Spectrally incoherent laser pulses with sufficiently large fractional bandwidth are in demand for the mitigation of laser–plasma instabilities occurring in high-energy laser–target interactions. Here, we modeled, implemented, and optimized a dual-stage high-energy optical parametric amplifier for broadband, spectrally incoherent pulses in the near-infrared. The amplifier delivers close to 400 mJ of signal energy through noncollinear parametric interaction of 100-nJ-scale broadband, spectrally incoherent seed pulses near 1053 nm with a narrowband high-energy pump operating at 526.5 nm. Mitigation strategies for high-frequency spatial modulations in the amplified signal caused by index inhomogeneities in the Nd:YLF rods of the pump laser are explored and discussed in detail.

47 OTHER INSTRUMENTATION↗

Exploring the Impact of N Solubility and Trace Elements on the Creep Properties of P92 Steel

Trace elements can have major consequences on the properties of creep resistant martensitic steels. To better understand the changes in microstructure and mechanical performance associated with the variation of trace elements and enable design of steels operating at more demanding conditions, we formulated three versions of P92 within the specified allowable N, B, C and Si ranges. Different service conditions were explored, and >80% decrease in creep life was observed at 625°C 155MPa for the highest B and N containing alloy. Multiscale characterization (OM, SEM, TEM) revealed key changes due to the trace element variation. The high trace content alloy formed deleterious BN precipitates with morphology that would promote crack nucleation, but also formed higher fractions of beneficial MX precipitates. The alloy with the lowest trace content showed the best creep performance – a result of the refined precipitate populations and the absence of large-scale inclusions and BN precipitates.

Antonov, Stoichko↗

Post-Deployment Characterization of Glass Fiber-Reinforced Thermoset and Thermoplastic Composite Tidal Turbine Blades

In 2021, the National Renewable Energy Laboratory (NREL) supported Verdant Power with the most successful tidal energy deployment in U.S. history. Three of their Gen5d 5 m turbines were deployed as part of the Roosevelt Island Tidal Energy project. Initially, the three rotors initially deployed were manufactured from glass fiber-reinforced epoxy composites. Midway through the deployment, one rotor was replaced with one manufactured at NREL. The new rotor utilized a novel infusible thermoplastic resin system. Since the deployment, one epoxy rotor and one thermoplastic rotor were returned to NREL for continued materials and manufacturing research. The two rotors underwent full-scale structural testing before being sectioned and cut into specimens for a variety of manufacturing quality tests, thermomechanical characterization, and evaluation of material performance in marine environments to understand the key differences between the fiberglass-reinforced epoxy and Elium composites used for the respective rotors. Matrix burn-off tests showed that the Elium blades had a considerably higher fiber volume fraction compared to the epoxy blades (61% vs. 49%). Environmental aging of the specimens showed that the epoxy laminates absorbed more water over the conditioning period; however, it was determined that the Elium laminates had higher diffusion coefficients, so they initially absorbed water faster. Finally, one full epoxy blade and one full Elium blade were conditioned at ambient temperatures for up to 11 months, while periodic mass measurements were taken. The datasets were extrapolated to assume a full 20-year operational life span, and it was determined that the blades would not reach full saturation during that time span.

composite manufacturing↗

Post-Deployment Characterization of Glass Fiber-Reinforced Thermoset and Thermoplastic Composite Tidal Turbine Blades

In 2021, the National Renewable Energy Laboratory (NREL) supported Verdant Power with the most successful tidal energy deployment in U.S. history. Three of their Gen5d 5 m turbines were deployed as part of the Roosevelt Island Tidal Energy project. Initially, the three deployed rotors were manufactured from glass fiber-reinforced epoxy composites. Midway through the deployment, one rotor was replaced with one manufactured at NREL. The new rotor utilized a novel infusible thermoplastic resin system (Elium from Arkema). Since the deployment, one epoxy rotor and one thermoplastic rotor were returned to NREL for continued materials and manufacturing research. The two rotors underwent full-scale structural testing before being sectioned and cut into specimens for a variety of manufacturing quality tests, thermomechanical characterization, and evaluation of material performance in marine environments to understand the key differences between the fiberglass-reinforced epoxy and Elium composites used for the respective rotors. Matrix burn-off tests showed that the Elium blades had a considerably higher fiber volume fraction compared to the epoxy blades (61% vs. 49%). Environmental aging of the specimens showed that the epoxy laminates absorbed more water over the conditioning period; however, it was determined that the Elium laminates had higher diffusion coefficients, so they initially absorbed water faster. Finally, one full epoxy blade and one full Elium blade were conditioned at ambient temperatures for up to 11 months, while periodic mass measurements were taken. The datasets were extrapolated to assume a full 20-year operational life span, and it was determined that the blades would not reach full saturation during that time span.

composite manufacturing↗

Valuing Maintenance Strategies for Fusion Plants as Part of a Future Electricity Grid

Scheduled maintenance is likely to be lengthy and therefore consequential for the economics of fusion power plants. The maintenance strategy that maximizes the economic value of a plant depends on internal factors such as the cost and durability of the replaceable components, the frequency and duration of the maintenance blocks, and the external factors of the electricity system in which the plant operates. Here, this paper examines the value of fusion power plants with various maintenance properties in a decarbonized United States Eastern Interconnection circa 2050. Seasonal variations in electricity supply and demand mean that certain times of year, particularly spring to early summer, are best for scheduled maintenance. Seasonality has two important consequences. First, the value of a plant can be 15% higher than what one would naively expect if value were directly proportional to its availability. Second, in some cases, replacing fractions of a component in shorter maintenance blocks spread over multiple years is better than replacing it all at once during a longer outage, even through the overall availability of the plant is lower in the former scenario.

power plant, maintenance, seasonality, electricity↗

Detecting fractionalization in critical spin liquids using color centers

Quantum spin liquids are highly entangled ground states of insulating spin systems, in which magnetic ordering is prevented down to the lowest temperatures due to quantum fluctuations. One of the most extraordinary characteristics of quantum spin liquid phases is their ability to support fractionalized, low-energy quasiparticles known as spinons, which carry spin-1/2 but bear no charge. Relaxometry based on color centers in crystalline materials—of which nitrogen-vacancy (NV) centers in diamond are a well-explored example—provides an exciting new platform to probe the spin spectral functions of magnetic materials with both energy and momentum resolution and to search for signatures of these elusive, fractionalized excitations. In this work, we theoretically investigate the color-center relaxometry of two archetypal quantum spin liquids: the two-dimensional U(1) quantum spin liquid with a spinon Fermi surface and the spin-1/2 antiferromagnetic spin chain. The former is characterized by a metallic, spin-split ground state of mobile, interacting spinons, which closely resembles a spin-polarized Fermi liquid ground state but with neutral quasiparticles. We show that the observation of the Stoner continuum and the collective spin wave mode in the spin spectral function would provide a strong evidence for the existence of spinons and fractionalization. In one dimension, mobile spinons form a Luttinger liquid ground state. We show that the spin spectral function exhibits strong features representing the collective density and spin-wave modes, which are broadened in an algebraic fashion with an exponent characterized by the Luttinger parameter. The possibilities of measuring these collective modes and detecting the power-law decay of the spectral weight using NV relaxometry are discussed. We also examine how the transition rates are modified by marginally irrelevant operators in the Heisenberg limit. Published by the American Physical Society 2024

Takei, So (ORCID:0000000191771895)↗

Physics Guided Simulation of Electrostatic Discharge: Technical Report

Triboelectrically-charged objects may create threshold sparks, electrostatic discharge (ESD) events, to equilibrate charge between themselves and other relatively charged objects. ESD events exhibit many complex physical phenomena. They are a nexus of several fields of physics with disparate characteristic scales: plasma physics, chemical kinetics, hydrodynamics, circuit models, etc. These scales can span many orders of magnitude from the varied collisions thermalizing information within a plasma on the $\mathcal{O}(fs/ps)$ to the physical size of the plasma channel on the $\mathcal{O}(100µm)$, to the speed of a nonlinear hydrodynamic wave propagating at $\mathcal{O}(µm, ns)$. These threshold ESD events may occur in situations of programmatic importance, delivering energy and power profiles to a “victim load” generating deleterious consequences. To predict and mitigate these consequences we must answer questions about the spark’s energy budget: how much energy goes into producing the spark channel; how much gets radiated away; how much energy is advected away into the hydrodynamics; and how much energy is delivered to a victim load. An ESD simulation toolset has been created and evolved in order to answer these questions. An appropriate, physics guided implementation for simulation can be done by gaining insight into its constituent physics and leveraging that intuition to choose a suitable numerical operator. We examine in detail the chemical kinetics, circuit discharge, and hydrodynamics to deter mine dominant regimes, values, timescales, and interactions to uncover the underlying physical dynamics. We also examine and propose model reduction schemes for high-dimensional chemical kinetics. We use past and current work with experimentally validated and theoretically-verified hydrodynamics to calculate applicability limits of the non-ionizing strong shock limit. We quantify the energy budget from a hydrodynamic perspective and demonstrate that a significant fraction of the stored energy is “earmarked” for hydrodynamic advection as an energy terminus. Lastly, we combine the constituent physics of an ESD event (chemical kinetics, circuit model, and hydrodynamics) into a cohesive, actionable toolset and obtain promising results from an isothermal test case. We then propose a viable, modular evolution of the ESD toolset based upon the performed examination of the physics uncovering dominant physical scales and the stiffness of the compositional differential system.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

First study of polarized proton-proton scattering with small-𝑥 helicity evolution

We perform a phenomenological study of helicity-dependent parton distribution functions (PDFs) using small-𝑥 helicity evolution equations, incorporating for the first time single-inclusive jet production data in polarized proton-proton (𝑝⁢𝑝) scattering at parton momentum fractions 𝑥 <0.1. We also simultaneously include double-longitudinal spin asymmetries in inclusive and semi-inclusive deep-inelastic scattering probing 𝑥 <0.1. Employing the polarized small-𝑥 pure-glue calculation of 𝑝⁢𝑝 → 𝑔⁡𝑋 for the jet production cross section, we modify the large-𝑁 𝑐 &𝑁 𝑓 KPS-CTT evolution equations by setting 𝑁 𝑓 = 0 to replicate the large-𝑁𝑐 (pure-glue) limit, while retaining external quark flavors for the spinor field operators. We find that the 𝑝⁢𝑝 data have a considerable impact on the helicity PDFs at small 𝑥, reducing their uncertainties and leading to a total quark and gluon helicity in the proton for 𝑥 < 0.1 of −0.04 ±0.23. Combining our analysis with a recent JAM helicity PDF analysis of the world polarized data, which includes 𝑥 > 0.1, we find a total quark and gluon helicity contribution for 𝑥 >10 −7 of between 0.02 and 0.51.

Adamiak, Daniel [Michigan State University] (ORCID↗

Experimental Investigation on Separation Behavior of Heavy-Oil Emulsion for Polymer Flooding on Alaska North Slope

The first-ever field pilot on Alaska North Slope (ANS) to validate using polymer floods for heavy-oil enhanced oil recovery is currently ongoing. One of the major concerns of the operator is the effect of polymer on oil/water-separation efficiency after polymer breakthrough. This work investigates the influence of polymer on the separation behavior of heavy-oil emulsions and evaluates the performance of emulsion breakers (EBs). In this study, two types of heavy-oil emulsions were prepared and tested at 20 and 50% water cut (WC), respectively. The bottle test method was used in the experiments, in which the separated water volume with time, the separated water quality, and the volume fraction of phases were recorded. Results showed that polymer accelerated the oil/water separation acting as an emulsion inhibitor at 20% WC but tended to impede the water separation at 50% WC. Regardless of WC, polymer resulted in poor water quality and the formation of a stable intermediate oil in water (o/w) emulsion, because of the increased viscosity of the water phase. The performance of EBs showed a complex dependency on the WC, the type of demulsifier and dosage, and the polymer concentration. Despite the varied conditions encountered in the heavy-oil/water/polymer/demulsifier system, a compound EB achieved satisfactory demulsification performance, showing the highest potential for deployment in the current ANS polymer flooding pilot. Here, we systematically studied the potential influence of polymer breakthrough on the separation behavior of heavy-oil emulsion on ANS for the first time. The findings of this study will provide practical guidance in advance for produced fluid treatment of the ongoing first-ever polymer flooding pilot on ANS.

54 ENVIRONMENTAL SCIENCES↗

Lithium-Ion Battery Diagnostics Using Electrochemical Impedance via Machine-Learning

Diagnosing battery states such as health, state-of-charge, or temperature is crucial for ensuring the safety and reliability of electrochemical energy storage systems. While some states, such as temperature, may be measured using cheap sensors, accurate diagnosis of battery health metrics usually requires time-consuming performance measurements, making them infeasible for use in real-world operation. These health metrics can be measured during lab-testing and then estimated on-line using predictive life models or via state observer algorithms such as Kalman filters, but these predictive methods should be supplemented by actual measurement of battery health whenever possible to ensure reliability. Rapid measurement of battery health may be done by various types of fast diagnostic techniques such as electrochemical impedance spectroscopy (EIS), which can be performed in only a few minutes and require only a fraction of the energy and power needed for a full charge and discharge measurement. But there is a substantial challenge for estimating battery health using EIS data, as EIS is sensitive to cell temperature, state-of-charge, current, and resting time in addition to health. Thus, utilizing EIS data to predict battery capacity requires correcting for all these additional variables, a task that is extremely difficult to handle analytically. This talk utilizes machine-learning methods to estimate the effectiveness of battery capacity prediction from EIS data, leveraging a data set of hundreds of EIS measurements recorded at varying temperature and state-of-charge throughout a 500-day aging study of 32 commercial, large-format NMC-Graphite lithium-ion batteries. Using EIS as input to machine-learning models is complicated by the nonlinear response of impedance to battery health, temperature, and state-of-charge, as well as the collinearity between the impedance response at neighboring frequencies, which can easily lead to overfit models. To train robust models, features from EIS data need to be extracted from the data or some subset of critical frequencies selected. Many approaches for extracting and selecting features from EIS data from electrochemical analysis and machine-learning fields were identified for analysis: using the entire raw spectra; selection of one, two, or many frequencies from the entire spectra; selecting interesting points from the EIS measurement using domain knowledge; fitting EIS with an equivalent-circuit model; calculating statistics on the raw impedance values; and reducing the dimensionality of the data using unsupervised linear (principal component analysis) and non-linear (uniform manifold approximation and projection) methods. These approaches were rigorously compared using a machine-learning pipeline approach, training linear, Gaussian process, and random forest regression models and quantifying performance using cross-validation as well as a held-out test set. An artificial neural network model trained on the raw spectra was also tested. Promising pipelines were fine-tuned via Bayesian hyperparameter optimization using cross-validation loss and training with class-specific weights to counter data set imbalance. The most reliable method for utilizing impedance in this work was the selection of two optimal frequencies through an exhaustive search, resulting in about 2% mean absolute error on test data for both Gaussian process and random forest model architectures. Interrogation of a variety of models reveals critical frequencies of 100 Hz and 103 Hz for this data set, though the optimal set of frequencies is not necessarily intuitive, i.e., the best performing models are not simply those that use impedance at frequencies that have the highest correlation to the relative discharge capacity. The best performing model is an ensemble model, which is able to predict battery capacity with 1.9% mean absolute error for unseen cells using impedance recorded at a variety of temperatures and states-of-charge.

battery↗

A seamless multiscale operator neural network for inferring bubble dynamics

Modelling multiscale systems from nanoscale to macroscale requires the use of atomistic and continuum methods and, correspondingly, different computer codes. Here, we develop a seamless method based on DeepONet, which is a composite deep neural network (a branch and a trunk network) for regressing operators. In particular, we consider bubble growth dynamics, and we model tiny bubbles of initial size from 100 nm to 10 $\mathrm {\mu }\textrm {m}$ , modelled by the Rayleigh–Plesset equation in the continuum regime above 1 $\mathrm {\mu }\textrm {m}$ and the dissipative particle dynamics method for bubbles below 1 $\mathrm {\mu }\textrm {m}$ in the atomistic regime. After an offline training based on data from both regimes, DeepONet can make accurate predictions of bubble growth on-the-fly (within a fraction of a second) across four orders of magnitude difference in spatial scales and two orders of magnitude in temporal scales. The framework of DeepONet is general and can be used for unifying physical models of different scales in diverse multiscale applications.

Mechanics↗