Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Model”

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 1,531 records · Page 85

Fast and accurate reduced-order modeling of a MOOSE-based additive manufacturing model with operator learning

One predominant challenge in additive manufacturing (AM) is to achieve specific material properties by manipulating manufacturing process parameters during the runtime. Such manipulation tends to increase the computational load imposed on existing simulation tools employed in AM. The goal of the present work is to construct a fast and accurate reduced-order model (ROM) for an AM model developed within the Multiphysics Object-Oriented Simulation Environment (MOOSE) framework, ultimately reducing the time/cost of AM control and optimization processes. Our adoption of the operator learning (OL) approach enabled us to learn a family of differential equations produced by altering process variables in the laser’s Gaussian point heat source. More specifically, we used the Fourier neural operator (FNO) and deep operator network (DeepONet) to develop ROMs for time-dependent responses. Furthermore, we benchmarked the performance of these OL methods against a conventional deep neural network (DNN)-based ROM. Ultimately, we found that OL methods offer comparable performance and, in terms of accuracy and generalizability, even outperform DNN at predicting scalar model responses. The DNN-based ROM afforded the fastest training time. Furthermore, all the ROMs were faster than the original MOOSE model yet still provided accurate predictions. FNO had a smaller mean prediction error than DeepONet, with a larger variance for time-dependent responses. Unlike DNN, both FNO and DeepONet were able to simulate time series data without the need for dimensionality reduction techniques. Finally, the present work can help facilitate the AM optimization process by enabling faster execution of simulation tools while still preserving evaluation accuracy.

36 MATERIALS SCIENCE↗

A holographic model for pseudogap in BCS–BEC crossover (I): Pairing fluctuations, double-trace deformation and dynamics of bulk bosonic fluid

We build a holographic model for the pairing fluctuation pseudogap phase in fermionic high temperature superconductivity/superfluidity based on the BCS–BEC crossover scenario. The pseudogap originates from incoherent Cooper pairing and has been observed in recent cold atom experiments. The strength of Cooper pairing and hence the BCS–BEC crossover is controlled by an effective 4-Fermi interaction and we argue that the double-trace deformation for charged scalar operator is a close analog in large N field theories. We employ the double-trace deformed Abelian Higgs model of holographic superconductors and propose that the incoherent fluctuations of the charged scalar in the bulk is the holographic dual of the fluctuating Cooper pairs. Using a Madelung transformation and the velocity-potential formalism, we develop a quantum fluid dynamics as an effective theory for these bulk fluctuations. The new fluid dynamics takes care of the boundary conditions required by AdS/CFT and encodes the vacuum polarization effect in curved spacetime. The pseudogap in conductivity can be related to the plasma oscillation of this bulk fluid.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Modeling the adsorbate coverage distribution over a multi-faceted catalytic grain in the presence of an electric field: O/Fe from first principles

The impact of an external electric field on the concerted behavior of oxygen over a multi-faceted catalytic Fe grain is determined via the interpolation of ab initio models of oxygen adsorption on Fe(100), Fe(110), and Fe(111) in the presence of an external electric field. The application of both negative and positive electric fields weaken the adsorption strength for oxygen on all three surface facets, with Fe(110) experiencing the greatest effect. Kinetic models of a multi-faceted catalytic Fe grain show that the average oxygen coverage over the grain surface is reduced under the influence of both a negative and positive electric field, which are consistent with phase diagram results at comparable pressures. Furthermore, we show that there is a weak synergistic effect between a Pd promoter and a positive electric field on the oxygen adsorption energy, i.e. the Pd promoter and electric field combination weaken the oxygen adsorption energy to a greater degree than the simple addition of both components separately. Overall, the work shows that the application of an applied external electric field may be a useful tool in fine-tuning chemical properties of Fe-based catalysts in hydrodeoxygenation applications.

Bray, Jacob E.↗

Train small, model big: Scalable physics simulators via reduced order modeling and domain decomposition

Numerous cutting-edge scientific technologies originate at the laboratory scale, but transitioning them to practical industry applications is a formidable challenge. Traditional pilot projects at intermediate scales are costly and time-consuming. An alternative, the pilot-scale model, relies on high-fidelity numerical simulations, but even these simulations can be computationally prohibitive at larger scales. To overcome these limitations, we propose a scalable, physics-constrained reduced order model (ROM) method. The ROM identifies critical physics modes from small-scale unit components, projecting governing equations onto these modes to create a reduced model that retains essential physics details. We also employ Discontinuous Galerkin Domain Decomposition (DG-DD) to apply ROM to unit components and interfaces, enabling the construction of large-scale global systems without data at such large scales. Here this method is demonstrated on the Poisson and Stokes flow equations, showing that it can solve equations about 15–40 times faster with only ~1% relative error. Furthermore, ROM takes one order of magnitude less memory than the full order model, enabling larger scale predictions at a given memory limitation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Two-tier model reduction of viscoelastically damped finite element models

Transient simulations of linear viscoelastically damped structures necessitate critical computational resources to directly integrate the full-order finite element model with time-stepping algorithms. Traditional modal reduction techniques are not directly applicable to these systems since viscoelastic materials depend on time and frequency. A more appropriate reduction basis is obtained from the nonlinear, complex eigenvalue problem, whose eigenvectors capture the appropriate kinematics and enable frequency-based mode selection; unfortunately, the computational cost is prohibitive for computing these modes from large-scale engineering models. To address this shortcoming, this work proposes a novel two-tier reduction procedure to reduce the upfront cost of solving the complex, nonlinear eigenvalue problem. The first reduction step reduces the full-order model with real mode shapes linearized about various centering frequencies to capture the kinematics over a full range of viscoelastic material behavior (glassy, rubbery, and glass-transition zones). This tier-one reduction preserves time-temperature superposition and allows the equations to depend parametrically on operating temperature. The second-level reduction then solves the complex, nonlinear eigenmode solutions in the tier-one reduced space about a fixed temperature to further reduce the equations-of-motion. The method is demonstrated on a cantilevered sandwich plate to display its accuracy and efficiency in comparison to full-order model predictions.

97 MATHEMATICS AND COMPUTING↗

Comparing cropland net primary production estimates from inventory, a satellite-based model, and a process-based model in the Midwest of the United States

Accurately quantifying the spatial and temporal variability of net primary production (NPP) for croplands is essential to understand regional cropland carbon dynamics. We compared three NPP estimates for croplands in the Midwestern United States: inventory-based estimates using crop yield data from the U.S. Department of Agriculture (USDA) National Agricultural Statistics Service (NASS); estimates from the satellite-based Moderate Resolution Imaging Spectroradiometer (MODIS) NPP product; and estimates from the General Ensemble biogeochemical Modeling System (GEMS) process-based model. The three methods estimated mean NPP in the range of 469–687 g C m -2 yr -1 and total NPP in the range of 318–490 Tg C yr -1 for croplands in the Midwest in 2007 and 2008. The NPP estimates from crop yield data and the GEMS model showed the mean NPP for croplands was over 650 g C m -2 yr -1 while the MODIS NPP product estimated the mean NPP was less than 500 g C m -2 yr -1 . MODIS NPP also showed very different spatial variability of the cropland NPP from the other two methods. We found these differences were mainly caused by the difference in the land cover data and the crop specific information used in the methods. Our study demonstrated that the detailed mapping of the temporal and spatial change of crop species is critical for estimating the spatial and temporal variability of cropland NPP. Finally, we suggest that high resolution land cover data with species–specific crop information should be used in satellite-based and process-based models to improve carbon estimates for croplands.

Biogeochemical model↗

Tidally averaged circulation in Puget Sound sub-basins: Comparison of historical data, analytical model, and numerical model

Through extensive field data collection and analysis efforts conducted since the 1950s, researchers have established an understanding of the characteristic features of circulation in Puget Sound. The pattern ranges from the classic fjordal behavior in some basins, with shallow brackish outflow and compensating inflow immediately below, to the typical two-layer flow observed in many partially mixed estuaries with saline inflow at depth. An attempt at reproducing this behavior by fitting an analytical formulation to past data is presented, followed by the application of a three-dimensional circulation and transport numerical model. The analytical treatment helped identify key physical processes and parameters, but quickly reconfirmed that response is complex and would require site-specific parameterization to include effects of sills and interconnected basins. The numerical model of Puget Sound, developed using unstructured-grid finite volume method, allowed resolution of the sub-basin geometric features, including presence of major islands, and site-specific strong advective vertical mixing created by bathymetry and multiple sills. The model was calibrated using available recent short-term oceanographic time series data sets from different parts of the Puget Sound basin. The results are compared against (1) recent velocity and salinity data collected in Puget Sound from 2006 and (2) a composite data set from previously analyzed historical records, mostly from the 1970s. The results highlight the ability of the model to reproduce velocity and salinity profile characteristics, their variations among Puget Sound subbasins, and tidally averaged circulation. Sensitivity of residual circulation to variations in freshwater inflow and resulting salinity gradient in fjordal sub-basins of Puget Sound is examined.

58 GEOSCIENCES↗

Modeling Analysis of the Swell and Wind-Sea Climate in the Salish Sea

Accurate model hindcast of wave climate in complex estuarine systems is challenging because high-resolution wave model and wind forcing are required. In this study, a modeling approach using the unstructured-grid Simulating WAves Nearshore and wind product from a high-resolution regional Weather Research and Forecasting hindcast was used to simulate the swell and wind-sea climate in the Salish Sea, a large estuary with many interconnected waterways on the Pacific Northwest coast of North America. The model hindcast was validated with observed data at four wave buoys in the Salish Sea. Spatial distribution and seasonal variations of wave climate in the Salish Sea were analyzed. Of the three major basins in the Salish Sea, the Strait of Juan de Fuca has the largest waves and is dominated by swells propagated from the Pacific Ocean. Significant wave heights in the Strait of Georgia have spatial and seasonal distribution patterns similar to those found in the Strait of Juan de Fuca. Waves in Puget Sound are small and primarily dominated by the wind-sea climate. Strong seasonal variations are observed in the Strait of Juan de Fuca and Strait of Georgia, but there is little seasonality of wave climate in Puget Sound. The high-resolution wave hindcast conducted in this study provides a comprehensive and important data set for better understanding the role of wave climate in coastal processes and natural hazards assessment in the Salish Sea.

wave climate, salish sea, modeling, swell, sea sta↗

Modeling Extreme Deformations in Lithium Ion Batteries

A simultaneously coupled modeling approach to study the electrochemical and thermal behavior of lithium-ion batteries under large mechanical deformation has been developed. The thermo-electrochemical pseudo-2D (P2D) battery model is coupled with a mechanical material model. Mechanical, thermal, and electrochemical models are implemented as user-defined sub-routines in the commercial multi-physics code LS-DYNA. The mechanical strain experienced by anode, cathode and separator results in thickness and porosity changes in each layer which in turn influences electrochemical behavior. The evolution of concentration profiles and cell potential are studied under different mechanical loading conditions. Internal short-circuits caused by mechanical deformation and corresponding physical behaviors are also elucidated. We discuss the competing effects of improved transport at higher temperature due to the internal short-circuit versus a drop in the effective ionic conductivity and electrolyte diffusivity due to mechanical deformation.

47 OTHER INSTRUMENTATION↗

An investigation of a model of the flow pattern transition mechanism in relation to the identification of annular flow of R134a in a vertical tube using various void fraction models and flow regime maps

In the present study, new experimental data are presented for literature on the prediction of film thickness and identification of flow regime during the co-current downward condensation in a vertical smooth copper tube having an inner diameter of 8.1 mm and a length of 500 mm. R134a and water are used as working fluids in the tube side and annular side of a double tube heat exchanger, respectively. Condensation experiments are done at mass fluxes of 300 and 515 kg m{sup -2} s{sup -1}. The condensing temperatures are between 40 and 50 C; heat fluxes are between 12.65 and 66.61 kW m{sup -2}. The average experimental heat transfer coefficient of the refrigerant HFC-134a is calculated by applying an energy balance based on the energy transferred from the test section. A mathematical model by Barnea et al. based on the momentum balance of liquid and vapor phases is used to determine the condensation film thickness of R134a. The comparative film thickness values are determined indirectly using relevant measured data together with various void fraction models and correlations reported in the open literature. The effects of heat flux, mass flux, and condensation temperature on the film thickness and condensation heat transfer coefficient are also discussed for the laminar and turbulent flow conditions. There is a good agreement between the film thickness results obtained from the theoretical model and those obtained from six of 35 void fraction models in the high mass flux region of R134a. In spite of their different valid conditions, six well-known flow regime maps from the literature are found to be predictive for the annular flow conditions in the test tube in spite of their different operating conditions. (author)

42 ENGINEERING↗

Spatial representation of organic carbon and active-layer thickness of high latitude soils in CMIP5 earth system models

Soil properties such as soil organic carbon (SOC) stocks and active-layer thickness are used in earth system models (F.SMs) to predict anthropogenic and climatic impacts on soil carbon dynamics, future changes in atmospheric greenhouse gas concentrations, and associated climate changes in the permafrost regions. Accurate representation of spatial and vertical distribution of these soil properties in ESMs is a prerequisite for redudng existing uncertainty in predicting carbon-climate feedbacks. We compared the spatial representation of SOC stocks and active-layer thicknesses predicted by the coupled Modellntercomparison Project Phase 5 { CMIP5) ESMs with those predicted from geospatial predictions, based on observation data for the state of Alaska, USA. For the geospatial modeling. we used soil profile observations {585 for SOC stocks and 153 for active-layer thickness) and environmental variables (climate, topography, land cover, and surficial geology types) and generated fine-resolution (50-m spatial resolution) predictions of SOC stocks (to 1-m depth) and active-layer thickness across Alaska. We found large inter-quartile range (2.5-5.5 m) in predicted active-layer thickness of CMIP5 modeled results and small inter-quartile range (11.5-22 kg m-2) in predicted SOC stocks. The spatial coefficient of variability of active-layer thickness and SOC stocks were lower in CMIP5 predictions compared to our geospatial estimates when gridded at similar spatial resolutions (24.7 compared to 30% and 29 compared to 38%, respectively). However, prediction errors. when calculated for independent validation sites, were several times larger in ESM predictions compared to geospatial predictions. Primaly factors leading to observed differences were ( 1) lack of spatial heterogeneity in ESM predictions, (2) differences in assumptions concerning environmental controls, and (3) the absence of pedogenic processes in ESM model structures. Our results suggest that efforts to incorporate these factors in F.SMs should reduce current uncertainties associated with ESM predictions of carbon-climate feedbacks.

Active-layer thickness↗

Fast Biological Modeling for Voxel-based Heavy Ion Treatment Planning Using the Mechanistic Repair-Misrepair-Fixation Model and Nuclear Fragment Spectra

Purpose: The physical and biological differences between heavy ions and photons have not been fully exploited and could improve treatment outcomes. In carbon ion therapy, treatment planning must account for physical properties, such as the absorbed dose and nuclear fragmentation, and for differences in the relative biological effectiveness (RBE) of ions compared with photons. We combined the mechanistic repair-misrepair-fixation (RMF) model with Monte Carlo-generated fragmentation spectra for biological optimization of carbon ion treatment plans. Methods and Materials: Relative changes in double-strand break yields and radiosensitivity parameters with particle type and energy were determined using the independently benchmarked Monte Carlo damage simulation and the RMF model to estimate the RBE values for primary carbon ions and secondary fragments. Depth-dependent energy spectra were generated with the Monte Carlo code FLUKA for clinically relevant initial carbon ion energies. The predicted trends in RBE were compared with the published experimental data. Biological optimization for carbon ions was implemented in a 3-dimensional research treatment planning tool. Results: We compared the RBE and RBE-weighted dose (RWD) distributions of different carbon ion treatment scenarios with and without nuclear fragments. The inclusion of fragments in the simulations led to smaller RBE predictions. A validation of RMF against measured cell survival data reported in published studies showed reasonable agreement. We calculated and optimized the RWD distributions on patient data and compared the RMF predictions with those from other biological models. The RBE values in an astrocytoma tumor ranged from 2.2 to 4.9 (mean 2.8) for a RWD of 3 Gy(RBE) assuming (α/β){sub X} = 2 Gy. Conclusions: These studies provide new information to quantify and assess uncertainties in the clinically relevant RBE values for carbon ion therapy based on biophysical mechanisms. We present results from the first biological optimization of carbon ion radiation therapy beams on patient data using a combined RMF and Monte Carlo damage simulation modeling approach. The presented method is advantageous for fast biological optimization.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Approximate Bayesian Model Inversion for PDEs with Heterogeneous and State-Dependent Coefficients

We present two approximate Bayesian inference methods for parameter estimation in partial differential equation (PDE) models with space-dependent and state-dependent parameters. We demonstrate that these methods provide accurate and cost-effective alternatives to Markov Chain Monte Carlo simulation. We assume a parameterized Gaussian prior on the unknown functions, and approximate the posterior density by a parameterized multivariate Gaussian density. The parameters of the prior and posterior are estimated from sparse observations of the PDE model's states and the unknown functions themselves by maximizing the evidence lower bound (ELBO), a lower bound on the log marginal likelihood of the observations. The first method, Laplace-EM, employs the expectation maximization algorithm to maximize the ELBO, with a Laplace approximation of the posterior on the E-step, and minimization of a Kullback-Leibler divergence on the M-step. The second method, DSVI-EB, employs the doubly stochastic variational inference (DSVI) algorithm, in which the ELBO is maximized via gradient-based stochastic optimization, with nosiy gradients computed via simple Monte Carlo sampling and Gaussian backpropagation. We apply these methods to identifying diffusion coefficients in linear and nonlinear diffusion equations, and we find that both methods provide accurate estimates of posterior densities and the hyperparameters of Gaussian priors. While the Laplace-EM method is more accurate, it requires computing Hessians of the physics model. The DSVI-EB method is found to be less accurate but only requires gradients of the physics model.

approximate Bayesian inference, model inversion, v↗

Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders

Nearly all model-reduction techniques project the governing equations onto a linear subspace of the original state space. Such subspaces are typically computed using methods such as balanced truncation, rational interpolation, the reduced-basis method, and (balanced) proper orthogonal decomposition (POD). Unfortunately, restricting the state to evolve in a linear subspace imposes a fundamental limitation to the accuracy of the resulting reduced-order model (ROM). In particular, linear-subspace ROMs can be expected to produce low-dimensional models with high accuracy only if the problem admits a fast decaying Kolmogorov n -width (e.g., diffusion-dominated problems). Unfortunately, many problems of interest exhibit a slowly decaying Kolmogorov n-width (e.g., advection-dominated problems). To address this, we propose a novel framework for projecting dynamical systems onto nonlinear manifolds using minimum-residual formulations at the time-continuous and time-discrete levels; the former leads to manifold Galerkin projection, while the latter leads to manifold least-squares Petrov–Galerkin (LSPG) projection. We perform analyses that provide insight into the relationship between these proposed approaches and classical linear-subspace reduced-order models; we also derive a posteriori discrete-time error bounds for the proposed approaches. In addition, we propose a computationally practical approach for computing the nonlinear manifold, which is based on convolutional autoencoders from deep learning. Lastly, we demonstrate the ability of the method to significantly outperform even the optimal linear-subspace ROM on benchmark advection-dominated problems, thereby demonstrating the method's ability to overcome the intrinsic n -width limitations of linear subspaces.

97 MATHEMATICS AND COMPUTING↗