Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “InTe”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Modeling Single and Two-Phase Transport in Thin Porous Layers Using a Composite Continuum-Pore Network Formulation

In this work, a composite continuum-pore network formulation is presented to model single and two-phase transport thin porous layers, such as gas diffusion layers in polymer electrolyte fuel cells (PEFCs) and active electrodes in redox flow batteries (RFBs). The formulation can be integrated into CFD codes, thus combining the ease of implementation of continuum-based modeling and the computational power of pore-network modeling. The composite model includes a control volume (CV) mesh at the layer scale, which embeds a cubic pore network [1,2]. The pore-network model is used to determine analytically local anisotropic effective transport properties (local effective diffusivity and permeability), which are mapped onto the CV mesh to simulate transport in the porous transport layer [3,4]. Good agreement is found between the predicted global effective transport properties (global effective diffusivity and permeability) under dry and wet conditions and previous experimental data reported in the literature for Toray TGP-H series carbon paper. Water saturation distributions are also compared with results obtained using X-ray computed tomography [4]. [1] P.A. García-Salaberri, I.V. Zenyuk, J.T. Gostick, A.Z. Weber, Modeling Gas Diffusion Layer in Polymer Electrolyte Fuel Cells Using a Continuum-Based Pore-Network Formulation, ECS Trans. 97 (2020) 615. [2] P.A. García-Salaberri, Modeling diffusion and convection in thin porous transport layers using a composite continuum-network model: Application to gas diffusion layers in polymer electrolyte fuel cells, Int J. Heat Mass Transf. (2020), submitted. [3] P.A. García-Salaberri, J.T. Gostick, G. Hwang, A.Z. Weber, M. Vera, Effective diffusivity in partially-saturated carbon-fiber gas diffusion layers: Effect of local saturation and application to macroscopic continuum models, J. Power Sources 296 (2015) 440–453. [4] P.A. García-Salaberri, G. Hwang, M. Vera, A.Z. Weber, J.T. Gostick, Effective diffusivity in partially-saturated carbon-fiber gas diffusion layers: Effect of through-plane saturation distribution, Int. J. Heat Mass Transf. 86 (2015) 319–333.

Garcia-Salaberri, Pablo↗
DOE OSTI2025

Integrated GW Farm ABM

This Data Repository includes data used for the integrated groundwater- farm ABM model, raw model output from scenario ensemble, and processed outputs that isolate the groundwater storage depletion outcomes for the 35,000 farm cells. Model Inputs: Farm ABM Inputs: This folder contains the input data used by the integrated groundwater - farm ABM modelling script (Python file) used for the high performance computing (HPC) experiments. The sub-folder "data inputs" contains all of the farm attribute data, while the three files in the folder have the hydrogeological data lookup table (NLDAS Cost Curve Attributes.csv), a lookup table (Theis well function table.csv) for the groundwater cost curve function, and the farm indexes and corresponding NLDAS ids for all of the cells run in this experiment (nldas farms subset final.csv). NLDAS Cost curve hydrogeological data: Hydrogeological data aggregated to 1/8 degree resolution and aligned with the NLDAS grid. Parameters include: water depth below ground surface [meters], subsurface porosity [unitless], aquifer depth from ground surface to aquifer bottom [meters], annual average recharge (USGS: mm, Doll: meters), and three different hydraulic conductivity (K) values (meters/day). The three K values represent the mean value from Gleeson et al. (2018), one standard deviation above the mean from Gleeson et al. (2018), and the de Graaf et al. 2020 modifications to certain lithologies. Additional information about these datasets and their processing are documented in the supplement to Yoon et al. 2025 (in review). Output: Raw outputs: This folder contains a .zip file that has model outputs for the entire scenario ensemble. There is one csv for each farm id, using the format "farm farmid cases.csv". The relationship between the farm id and NLDAS id is defined by the "nldas farms subset final.csv" located in the Farm ABM Inputs folder. Each csv has 625 rows, corresponding to 625 combinations of different scenario parameter values. Each row (scenario) represents the outcome of a 100 year simulation. Columns define scenario settings and summary statistics for each scenario. The first four columns define the scenario settings: "hydro ratio," "econ ratio," "K scenario," and "gamma scenario." The hydro and econ ratios are values passed to the modeling script that influence multipliers for other model parameters, as documented in the supplement to Yoon et al. 2025 (in review). The gamma multiplier is a coefficient multiplier applied to the baseline gamma values (values below 1 represent lower unobserved costs compared to baseline, values above 1 represent higher costs). The K scenario names represent K values of: "low": 0.5 m/d, "int 1": 2.5 m/d, "int 2": 10 m/d, "high": 50 m/d, and "gleeson": mean Gleeson K value. "Perc vol depleted" is the fraction of groundwater depleted at the end of the 100 simulation. Processed Output: Derived depletion outcomes from raw outputs: All of the individual csv files from the Raw outputs were aggregated into a single file that has the scenario settings and fraction depletion "Perc vol depleted" for every farm cell, for every scenario. The other two files define relationships between the farm id, NLDAS id, and local and major aquifer units, used for aquifer-level depletion analysis.

Agent based modeling↗
NASA NTRS1985

The chemistry of Li/SOCl2 cells - An ESR study of carbon electrodes

Carbon electrodes from Li/SOCl2 cells were studied by electron spin resonance after various stages of discharge. Different behavior was observed in the temperature-dependent part of the ESR linewidth, defined as 'intrinsic linewidth', Delta H(int), when two different electrolytes were used. With one electrolyte, 1.5M LiAlCl4/SoCl2, the Delta H(int) value stayed constant or slightly decreased whereas with another electrolyte, 1.0M LiAlCl4/14 percent BrClin SOCl2, the value increased as discharge progressed. The carbon electrodes are modified differently during discharge with these two electrolytes, and it is speculated that this may be due to changes in the carbon matrix functional groups. This difference in the carbon electrodes may explain the claimed differences in safety performance of the cells.

Kim, S. S.↗
NASA NTRS1992

Comparison of epifluorescent viable bacterial count methods

Two methods, the 2-(4-Iodophenyl) 3-(4-nitrophenyl) 5-phenyltetrazolium chloride (INT) method and the direct viable count (DVC), were tested and compared for their efficiency for the determination of the viability of bacterial populations. Use of the INT method results in the formation of a dark spot within each respiring cell. The DVC method results in elongation or swelling of growing cells that are rendered incapable of cell division. Although both methods are subjective and can result in false positive results, the DVC method is best suited to analysis of waters in which the number of different types of organisms present in the same sample is assumed to be small, such as processed waters. The advantages and disadvantages of each method are discussed.

Rodgers, E. B.↗
NASA NTRS1995

Locations of termination shock and heliopause based on Voyager plasma and magnetic field data

The locations of the termination shock and the heliopause are studied taking into account the effects of pickup protons. The study uses available plasma and magnetic field data from Voyagers over a 14-year period (1978-1991) and Voyager observation of the 1992-93 radio emission event. Outside 30 AU, pickup protons have a significant influence on dynamical structures of the outer heliosphere. The solar wind is treated as a mixture of electrons, solar wind protons, and interstellar pickup protons. If the magnitude of the interstellar magnetic field B(sub int) is given, one can quantitatively study the motion and location of the termination shock. The location is anti-correlated with the sun spot number and the shock has an average speed of approx. 24 km/s. Because B(sub int) is poorly known, additional information is needed in studying the termination shock. Cummings, et al. have used observations of anomalous cosmic rays to estimate the location of the shock. The observations of the 1991 GMIR and GMIR shock and the 1992-93 radio emission event provide another handle for the study of the termination shock and the heliopause. After its penetration through the termination shock, the GMIR shock continued to propagate in the subsonic region of the solar wind and eventually interacted with the heliopause. This interaction produces a transmitted shock propagating outward in the interstellar medium and a reflected shock propagating inward toward the sun in the subsonic solar wind. The plasma frequencies behind the reflected and the transmitted shock can be, respectively, responsible for the 2- and 3-kHz radio emissions. Taking into account the effects of pickup protons we found that the average locations of the termination shock and the heliopause in 1991-92 are at approximately 66 AU and 150 AU, respectively.

Whang, Y. C.↗
NASA NTRS2016

CGRO/BATSE Data Support the New Paradigm For GRB Prompt Emission and the New L-i(nTh)-E-peak,i(nTh,rest) Relation

The paradigm for gamma-ray burst (GRB) prompt emission is changing. Since early in the Compton Gamma RayObservatory (CGRO) era, the empirical Band function has been considered a good description of the keV-MeV-gamma-ray prompt emission spectra despite the fact that its shape was very often inconsistent with the theoretical predictions, especially those expected in pure synchrotron emission scenarios. We have recently established a new observational model analyzing data of the NASA Fermi Gamma-ray Space Telescope. In this model, GRB prompt emission would be a combination of three main emission components: (i) a thermal-like component that we have interpreted so far as emission from the jet photosphere, (ii) a non-thermal component that we have interpreted so far as either synchrotron radiation from the propagating and accelerated charged particles within the jet or reprocessed jet photospheric emission, and (iii) an additional non-thermal (cutoff) power law (PL) extending from low to high energies in gamma-rays and most likely of inverse Compton origin. In this article we reanalyze some of the bright GRBs, namely GRBs 941017, 970111, and 990123, observed with the Burst And Transient Source Experiment (BATSE) on board CGRO with the new model. We conclude that BATSE data for these three GRBs are fully consistent with the recent results obtained with Fermi: some bright BATSE GRBs exhibit three separate components during the prompt phase with similar spectral parameters as those reported from Fermi data. In addition, the analysis of the BATSE GRBs with the new prompt emission model results in a relation between the time-resolved energy flux of the non-thermal component, F(in)(Th), and its corresponding nuFnu spectral peak energy,Epeak,inTh (i.e., FinThEpeak,inTh ), which has a similar index when fitted to a PL as the one initially derived from Fermi data. For GRBs with known redshifts (z) this results in a possible universal relation between the luminosity of the non-thermal component, LinTh, and its corresponding nuFnu spectral peak energy in the rest frame, Epeak,iNT,rest (i.e.,LinThEpeak,iNT,rest). We estimated the redshifts of GRBs 941017 and 970111 using GRB 990123 with z = 1.61as a reference. The estimated redshift for GRB 941017 is typical for long GRBs and the estimated redshift for GRB970111 is right in the range of the expected values for this burst.

acceleration of particles – black hole physics â↗
NASA NTRS2017

Dissociation and Internal Excitation of Molecular Nitrogen Due to N + N2 Collisions Using Direct Molecular Simulation

In this work we present a molecular level study of N2+N collisions, focusing on excitation of internal energy modes and non-equilibrium dissociation. The computation technique used here is the direct molecular simulation (DMS) method and the molecular interactions have been modeled using an ab−initio potential energy surface (PES) developed at NASA's Ames Research Center. We carried out vibrational excitation calculations between 5000K and 30000K and found that the characteristic vibrational excitation time for the N + N2 process was an order of magnitude lower than that predicted by the Millikan and White correlation. It is observed that during vibrational excitation the high energy tail of the vibrational energy distribution gets over populated first and the lower energy levels get populated as the system evolves. It is found that the non-equilibrium dissociation rate coefficients for the N + N2 process are larger than those for the N2 + N2 process. This is attributed to the non-equilibrium vibrational energy distributions for the N + N2 process being less depleted than that for the N2 +N2 process. For an isothermal simulation we find that the probability of dissociation goes as 1/T(sub tr) for molecules with internal energy (epsilon(sub int)) less than approximately 9.9eV, while for molecules with epsilon (sub int) greater than 9.9eV the dissociation probability was weakly dependent on translational temperature of the system. We compared non-equilibrium dissociation rate coefficients and characteristic vibrational excitation times obtained by using the ab-initio PES developed at NASA's Ames Research Center to those obtained by using an ab-initio PES developed at the University of Minnesota. Good agreement was found between the macroscopic properties and molecular level description of the system obtained by using the two PESs.

Grover, Maninder S.↗
NASA NTRSDate not supplied

Automatic Processing of Intensives at GSFC VLBI Analysis Center

VLBI Intensive (INT) sessions are conducted to determine the change in Earth rotation, which is measured as a correction to UT1. This correction varies unpredictably over time. Because UT1 is used in precise navigation, particularly GNSS, rapid turnaround is very important. The decrease of time elapsed from observations to obtained results can be shortened with the automatic data processing of new INT sessions at the analysis stage.

VLBI↗
NASA NTRS2023

VGOS VLBI Intensives Between MACGO12M and WETTZ13S for the Rapid Determination of UT1-UTC

In this work, we present a status update and preliminary results of the designated research and development VLBI Intensive program VGOS-INT-S, observed between MACGO12M and WETTZ13S for the rapid determination of the Earth’s phase of rotation, expressed via UT1-UTC. Since 2021, 27 Intensive sessions have been observed successfully utilizing a special observation strategy alternating between high- and low-elevation scans for improved determination of delays caused by the neutral atmosphere. Between the end of January and mid of March 2022, VGOS-INT-S was among the most accurate Intensive programs. During this time, eight sessions were observed with an average formal error of 3.1 µs and a bias w.r.t. IERS C04 of 1.1 µs. Later, the session performance decreased due to multiple technical difficulties.

VLBI↗
DOE OSTI2022

Impact of f-element complexation on the radiolytic robustness of separations ligands

Impact of f-element complexation on the radiolytic robustness of separations ligands Gregory P. Horne, Makayla R. Baxter, Corey D. Pilgrim, Travis S. Grimes, Center for Radiation Chemistry Research, Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID, 83415, USA Cristian Celis Barros, Department of Chemistry and Biochemistry, Florida State University, Tallahassee, FL 32306, USA. E-mail: ccelisbarros@fsu.edu Andrew R. Cook, Department of Chemistry, Brookhaven National Laboratory, Upton, New York, 11973, USA Stephen P. Mezyk, Department of Chemistry and Biochemistry, California State University Long Beach, 1250 Bellflower Boulevard, Long Beach, California, 90840-9507, USA The effects of ionizing radiation are ubiquitous throughout all aspects of a nuclear fuel cycle. However, the complexity and intensity of these effects are greatest during the management of used nuclear fuel, owing to the presence of a wide spectrum of radionuclides from neutron capture and fission processes. With regards to used nuclear fuel (UNF) reprocessing, radiation-induced processes typically promote the destruction of active compounds (e.g., complexants and additives) with the concomitant formation of potentially detrimental degradation products and corresponding changes in physical and chemical properties, which ultimately impact the effectiveness and longevity of a given reprocessing system. Concerning UNF complexants, radiation chemistry studies have historically focused on their radiation robustness in the absence of the metal ions they were designed to selectively complex. This knowledge gap is worrisome as previous studies on aqueous phase complexants have demonstrated significant changes in radiolytic behavior upon metal ion complexation.1-4 More recently, the rate of reaction of the n-dodecane radical cation—believed to be the major organic phase radiation-induced transient species responsible for complexant radiolysis in n-dodecane based solvent systems—with hexa-n-octylnitrilo-triacetamide (HONTA) was shown to increase by an order-of-magnitude upon complexation of europium or americium.5 These findings have significant implications on the projected longevity of complexants in UNF reprocessing solvent systems. Consequently, a thorough understanding of metal ion complexation effects on the radiolytic integrity of UNF complexants is essential to evaluate their potential for process application. Presented here are two recent studies from the Idaho National Laboratory Center for Radiation Chemistry Research group that demonstrate the various impacts of f-element complexation (uranium, americium, and lanthanides) on the radiolytic integrity (gamma and electron pulse) of tributyl phosphate (TBP), N,N-di-(2-ethylhexyl)butyramide (DEHBA), N,N-di-2-ethylhexylisobutryamide (DEHiBA), and 2-ethylhexylphosphonic acid mono-2-ethylhexyl ester (HEH[EHP]) under UNF reprocessing conditions. References 1) Bhattacharyya and Kundu, Int. J. Radiat. Phys. Chem., 1971, 3, 1. 2) Kundu and Matuura, Int. J. Radiat. Phys. Chem., 1975, 7, 565. 3) Ilan and Czapski, Biochimica et Biophysica Acta, 1977, 498, 386. 4) Buettner, Doherty, and Patterson, Fed. Euro. Biochem. Soc., 1983, 158 (1), 143. 5) Toigawa, Peterman, Meeker, Grimes, Zalupski, Mezyk, Cook, Yamashita, Kumagai, Matsumura, Horne, PCCP, 2021, 23, 1343.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗
DOE OSTI2021

Mathematical modeling and parameter estimation of MFI membranes for para/ortho‐xylene separation

Abstract A mathematical model of a membrane with a thin, oriented, and selective MFI layer, which also includes contributions from defects, pore blockages, support layer, and external mass transfer, was formulated based on the Maxwell‐Stefan equations. It was validated using reported (Kim et al., Angew Chemie Int Ed, 2018, 57:480–485; Jeon et al., Nature, 2017, 543:690–694) para/ortho‐xylene separation data from five MFI membranes. The diffusivities of the xylenes were considered to be the same for all membranes, while the thickness and the density of defects and pore blockage were treated as unique properties of each membrane, and their contributions were estimated by fitting the model to the corresponding separation data. The effects of these properties and the role of permeate pressure on the separation performance were subsequently analyzed. The proposed modeling, parameter estimation, and analysis framework allow one to quantitatively interpret the variation of separation performance, to understand separation bottlenecks, and to provide guidance for designing membranes with desired performance.

Liu, Jingjun↗
DOE OSTI2022

Inclusive, prompt and non-prompt J/ψ production at midrapidity in p-Pb collisions at $ \sqrt{s_{\mathrm{NN}}} $ = 5.02 TeV

A measurement of inclusive, prompt, and non-prompt J/ψ production in p-Pb collisions at a nucleon-nucleon centre-of-mass energy √ s NN = 5.02 TeV is presented. The inclusive J/ψ mesons are reconstructed in the dielectron decay channel at midrapidity down to a transverse momentum p T = o. The inclusive J/ψ nuclear modification factor R pPb is calculated by comparing the new results in p-Pb collisions to a recently measured proton-proton reference at the same centre-of-mass energy. Non-prompt J/ψ mesons, which originate from the decay of beauty hadrons, are separated from promptly produced J/ψ on a statistical basis for p T larger than 1.0 GeV/c. These results are based on the data sample collected by the ALICE detector during the 2016 LHC p-Pb run, corresponding to an integrated luminosity L int = 292 ± 11 μb –1 , which is six times larger than the previous publications. The total uncertainty on the p T -integrated inclusive J/ψ and non-prompt J/ψ cross section are reduced by a factor 1.7 and 2.2, respectively. The measured cross sections and R pPb are compared with theoretical models that include various combinations of cold nuclear matter effects. From the non-prompt J/ψ production cross section, the bb¯ production cross section at midrapidity, dσ bb¯ /dy, and the total cross section extrapolated over full phase space, σ bb¯ , are derived.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2023

Exclusive dielectron production in ultraperipheral Pb+Pb collisions at $\sqrt{s_{\textrm{NN}}}$ = 5.02 TeV with ATLAS

Exclusive production of dielectron pairs, γγ → e + e – , is studied using $\mathcal{L}_{int}$ = 1.72 nb –1 of data from ultraperipheral collisions of lead nuclei at $\sqrt{s_{NN}}$ = 5.02 TeV recorded by the ATLAS detector at the LHC. The process of interest proceeds via photon–photon interactions in the strong electromagnetic fields of relativistic lead nuclei. Dielectron production is measured in the fiducial region defined by following requirements: electron transverse momentum $p^{e}_{T}$ > 2.5 GeV, absolute electron pseudorapidity |η e | < 2.5, dielectron invariant mass m ee > 5 GeV, and dielectron transverse momentum $p^{ee}_{T}$ < 2 GeV. Differential cross-sections are measured as a function of m ee , average $p^{e}_{T}$, absolute dielectron rapidity |y ee |, and scattering angle in the dielectron rest frame, |cos θ*|, in the inclusive sample, and also with a requirement of no activity in the forward direction. The total integrated fiducial cross-section is measured to be 215 ± 1 (stat) $^{+23}_{-20}$ (syst.) ± 4 (lumi.) μ b. Within experimental uncertainties the measured integrated cross-section is in good agreement with the QED predictions from the Monte Carlo programs STARLIGHT and SUPERCHIC, confirming the broad features of the initial photon fluxes. The differential cross-sections show systematic differences from these predictions which are more pronounced at high |y ee | and |cos θ*| values.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Nuclear Structure at the Crossroads

Abstract Steven Weinberg’s seminal papers from 1990–92 initiated the use of effective field theories (EFTs) for nuclei. We summarize progress, priorities, and open questions for nuclear EFT developments based on the 2019 INT program “Nuclear Structure at the Crossroads.”

Physics↗
DOE OSTI2026

Evaluation of chromatographic resins for high efficacy separation of radium from chemical analogues

The separation of Radium (Ra) from interfering chemical analogues is of major importance for both the accurate measurement of Ra-226 in environmental samples and the preparation of radium targets destined for medical isotope production. The adsorption behavior of Ra, Barium (Ba), and Strontium (Sr) was investigated on multiple crown-ether-based proprietary extraction chromatographic resins developed by TrisKem Int. Among these, the TK101 resin, characterized by a novel room-temperature ionic liquid (RTIL) dual extractant–solvent system, demonstrated enhanced affinity and selectivity, particularly at low acid concentrations, achieving a D w of 3477 ± 73 for Ra in 0.04 M HNO₃, whereas the maximum Ra D w obtained with the traditional Sr resin in 8 M HNO 3 was only 12.33 ± 0.4. In 0.04 M HNO 3 , a Ra distribution ratio (D w ) of 776 was achieved within 1 min, indicating rapid uptake kinetics. The adsorption capacities were estimated at 19.6 ± 0.1 mg/g for Ba and 12.0 ± 0.2 mg/g for Sr. However, competing alkali metals significantly reduced alkaline earth metal uptake, indicating that the TK101 resin is best suited for radium purification following a preconcentration step to minimize matrix effects in high salinity samples. Dynamic column studies confirmed the strong chromatographic performance of the TK101 resin, with no significant Ba or Sr breakthrough in the Ra fraction and a Ra recovery of 87 ± 2.8%. These results demonstrate the potential of the TK101 resin for selective radium separation in analytical and radiochemical applications.

and nuclear chemistry↗
DOE OSTI2021

Electronic structure of reduced CeO 2 (111) surfaces interacting with hydrogen as revealed through electron energy loss spectroscopy in comparison with theoretical investigations

Based on both, ab-initio cluster calculations as well as periodic density functional based loss function calculations, we have assigned the origins of the valence electron excitation regime in electron energy loss spectra of well-ordered ceria films in (111) orientation at various states of reduction as well as after exposing the reduced films to hydrogen from the gas phase (Li et al, Angew. Chem. Int. Ed., 58 (2019) 14686–14693). The explicit calculation of intensity distributions using the dipole approximation allow us to draw conclusions about the nature of oxygen vacancies, which occur upon reduction and how those interact with hydrogen to form hydride species. Herein, we also reports a brief discussion of vibrational excitations of hydrogen loaded ceria and corroborates previous interpretations, based on inelastic neutron scattering.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Proton spin from small-x with constraints from the valence quark model

We apply the valence quark model [1] to constrain the non-perturbative initial condition for the small-x helicity evolution. The remaining free parameters are constrained by performing a global analysis akin to [2] to the available polarized small-x deep inelastic scattering data. A good description of the world data is obtained with only 8 free parameters. The model parameters are tightly constrained by the data, allowing us to predict the proton polarized structure function $g^p_1$ to be negative at small x. Furthermore, we obtain the small-x quark and gluon spins to give a contribution $\int^{0.1}_{10^{-5}} dx (\frac{1}{2}\Delta\Sigma + \Delta{G}) = 0.63 ± 0.10$ or $1.35 ± 0.16$ (or to the proton spin, depending on the applied running coupling prescription.

Color-class condensate↗
DOE OSTI2023

Deep learning closure models for large-eddy simulation of flows around bluff bodies

Near-wall flow simulation remains a central challenge in aerodynamics modelling: Reynolds-averaged Navier–Stokes predictions of separated flows are often inaccurate, and large-eddy simulation (LES) can require prohibitively small near-wall mesh sizes. A deep learning (DL) closure model for LES is developed by introducing untrained neural networks into the governing equations and training in situ for incompressible flows around rectangular prisms at moderate Reynolds numbers. The DL-LES models are trained using adjoint partial differential equation (PDE) optimization methods to match, as closely as possible, direct numerical simulation (DNS) data. They are then evaluated out-of-sample – for aspect ratios, Reynolds numbers and bluff-body geometries not included in the training data – and compared with standard LES models. The DL-LES models outperform these models and are able to achieve accurate LES predictions on a relatively coarse mesh (downsampled from the DNS mesh by factors of four or eight in each Cartesian direction). We study the accuracy of the DL-LES model for predicting the drag coefficient, near-wall and far-field mean flow, and resolved Reynolds stress. A crucial challenge is that the LES quantities of interest are the steady-state flow statistics; for example, a time-averaged velocity component $\langle {u}_i\rangle (x) = \lim _{t \rightarrow \infty } ({1}/{t}) \int _0^t u_i(s,x)\, {\rm d}s$ . Calculating the steady-state flow statistics therefore requires simulating the DL-LES equations over a large number of flow times through the domain. It is a non-trivial question whether an unsteady PDE model with a functional form defined by a deep neural network can remain stable and accurate on $t \in [0, \infty )$ , especially when trained over comparatively short time intervals. Our results demonstrate that the DL-LES models are accurate and stable over long time horizons, which enables the estimation of the steady-state mean velocity, fluctuations and drag coefficient of turbulent flows around bluff bodies relevant to aerodynamics applications.

Mechanics↗