SEARCH · Engineering Papers

Results for “direct solver”

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.

SourceResearch connectionFrom yearThrough year

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

At least 667 records · Page 37

All sources NASA NTRS DOE OSTI

NASA NTRS2007

Computational Plume Modeling of COnceptual ARES Vehicle Stage Tests

The plume-induced environment of a conceptual ARES V vehicle stage test at the NASA Stennis Space Center (NASA-SSC) was modeled using computational fluid dynamics (CFD). A full-scale multi-element grid was generated for the NASA-SSC B-2 test stand with the ARES V stage being located in a proposed off-center forward position. The plume produced by the ARES V main power plant (cluster of five RS-68 LOX/LH2 engines) was simulated using a multi-element flow solver - CRUNCH. The primary objective of this work was to obtain a fundamental understanding of the ARES V plume and its impingement characteristics on the B-2 flame-deflector. The location, size and shape of the impingement region were quantified along with the un-cooled deflector wall pressures, temperatures and incident heating rates. Issues with the proposed tests were identified and several of these addressed using the CFD methodology. The final results of this modeling effort will provide useful data and boundary conditions in upcoming engineering studies that are directed towards determining the required facility modifications for ensuring safe and reliable stage testing in support of the Constellation Program.

Allgood, Daniel C.↗

NASA NTRS2019

Investigation of Transonic Truss-Braced Wing Aircraft Transonic Wing-Strut Interference Effects Using FUN3D

This paper presents a computational study of transonic wing-strut interference effects of Transonic Truss-Braced Wing (TTBW) aircraft using the high-fidelity CFD (Computational Fluid Dynamics) code FUN3D (Fully Unstructured Navier-Stokes Three Dimensional). The study is conducted for the wing-strut and the wing-alone configurations at different Mach numbers and Reynolds numbers. The interference effects are calculated by comparing the wing aerodynamics along the spanwise direction between the wing-strut and the wing-alone configurations. The presence of the strut underneath the wing induces a suction peak on the lower surface of the wing, which causes changes in aerodynamic forces and moments, as well as the aerodynamic center location. The interference effects become more pronounced as the Mach number increases. The Reynolds number has less impact on the interference effects. A transonic wing-strut interference aerodynamic correction method is developed for use in a lower-fidelity tool, VSPAERO [a vortex lattice flow solver], coupled to a finite-element model for rapid flutter analysis.

Xiong, Juntao↗

DOE OSTI2023

Hybrid eigensolvers for nuclear configuration interaction calculations

We examine and compare several iterative methods for solving large-scale eigenvalue problems arising from nuclear structure calculations. In particular, we discuss the possibility of using block Lanczos method, a Chebyshev filtering based subspace iterations and the residual minimization method accelerated by direct inversion of iterative subspace (RMM-DIIS) and describe how these algorithms compare with the standard Lanczos algorithm and the locally optimal block preconditioned conjugate gradient (LOBPCG) algorithm. Although the RMM-DIIS method does not exhibit rapid convergence when the initial approximations to the desired eigenvectors are not sufficiently accurate, it can be effectively combined with either the block Lanczos or the LOBPCG method to yield a hybrid eigensolver that has several desirable properties. We will describe a few practical issues that need to be addressed to make the hybrid solver efficient and robust.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

Data-Driven RANS Turbulence Closures for Forced Convection Flow in Reactor Downcomer Geometry

Recent progress in data-driven turbulence modeling has shown its potential to enhance or replace traditional equation-based Reynolds-averaged Navier-Stokes (RANS) turbulence models. Here, this work utilizes invariant neural network (NN) architectures to model Reynolds stresses and turbulent heat fluxes in forced convection flows (when the models can be decoupled). As the considered flow is statistically one dimensional, the invariant NN architecture for the Reynolds stress model reduces to the linear eddy viscosity model. To develop the data-driven models, direct numerical and RANS simulations in vertical planar channel geometry mimicking a part of the reactor downcomer are performed. Different conditions and fluids relevant to advanced reactors (sodium, lead, unitary-Prandtl-number fluid, and molten salt) constitute the training database. The models enabled accurate predictions of velocity and temperature, and compared to the baseline k–τ turbulence model with the simple gradient diffusion hypothesis, do not require tuning of the turbulent Prandtl number. The data-driven framework is implemented in the open-source graphics processing unit–accelerated spectral element solver nekRS and has shown the potential for future developments and consideration of more complex mixed convection flows.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

DOE OSTI2026

Excitation energy of fission fragments within nuclear time-dependent density functional theory

The number and properties of the neutrons and photons emitted in nuclear fission are directly related to the excitation energy of the fission fragments when they are formed at scission. Though not observable experimentally because of the extremely short timescales, the excitation energy of fission fragments can be predicted by microscopic theory based on time-dependent density functional theory (TDDFT). Initial results on the value of the total kinetic energy of fission reactions were very promising, but could not probe all possible fragmentations. Here, in this work, we perform large-scale TDDFT calculations in 240 Pu enabled by the development of a new TDDFT solver. We obtain TDDFT trajectories covering nearly all possible fragmentations. We find that the total kinetic energy is close to experimental values only for the most likely fission while it is severely underestimated at both small and large asymmetries. This conclusion seems rather independent of the parametrization of the energy functional, both in its particle-hole and particle-particle channels.

Physics - Nuclear physics and radiation physics↗

DOE OSTI2022

New modelling capabilities in IDT

This work concerns the enhanced modelling capabilities of the discrete ordinates transport solver IDT. The novelties introduced allow for modelling unstructured geometries composed by a collection of X/Y segments and circles, and the use of reciprocity and conservation relations reduce the memory imprint as well as the computational cost of the method. IDT decomposes geometries in modular Cartesian patterns, which are the so-called Heterogeneous Cartesian Cells (HCCs), containing a chunk of the original unstructured geometries. Each HCC can be then discretized by superimposing a XY grid to refine locally the HCC. Unlike the most popular MOC, IDT performs the spatial sweeping by directional collision probabilities instead of trajectories. The sources and interface angular fluxes are expanded up to linear order. The accuracy of ray-tracing, the memory imprints together with the novel mesh refinement capabilities have been verified. A first set of preliminary results on PWR lattice problems will be presented in this paper.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2021

Pulse Detonation Engine for Advanced Oxy-Combustion of Coal-Based Fuel for Direct Power Extraction Applications

Pressure gain combustion and magnetohydrodynamic (MHD) systems have the potential to provide a step increase in the efficiency of combined-cycle power plants. Specific advantages include a net pressure increase to the system instead of a pressure drop, the high temperature of the detonation waves can increase the efficiency of power extraction compared to other processes, significant thermal energy can be released in a compact region, and the high velocities of the flow increase extraction of electrical power. In summary, a pressure gain combustor coupled with a MHD has the potential to be transformative. Despite the potential advantages, relatively little research has been conducted considering coupled pressure gain combustion systems with MHD systems. With this background and motivation, the overall goal of this effort was to advance the knowledge, technology, and computational tools associated with coupled detonation and MHD systems. A joint experimental and computational approach was used while seeking to accomplish the goals of this work. Specifically, two pulse detonation engines were developed and used for the experiments to produce detonations. Detonation speeds were measured for a variety of flow and fuel conditions (e.g., methane, propane, with coal particles). Preliminary electrical conductivity measurements were collected. An extensive amount of research was performed to identify sensitivities of detonation behaviors to the presence of combustion products. Computationally, a twofold approach was used in this work. First, a solver was developed for solving the governing equations for a reactive flow with coupled detonation and ionization chemistry. The solver was applied to study the impacts of seed material ionization on detonation. Second, a conservation element-solution element (CE-SE) based numerical solver for detonation studies with a reduced reaction mechanism for oxy-methane combustion was developed and verified on standard test cases. Key findings and contributions from this work are as follows. A system was developed for injecting powderized coal, or other seeding material, into a pulse-detonation system. The influence of a combustion product (i.e., CO 2 ) on detonation behavior was identified. Knowledge gained from this work is applicable to devices such as rotating detonation engines, where combustion products mix with fresh reactants. A system for measuring the electrical conductivity of the exhaust from a pulse-detonation engine was developed. The open-source solver, Clawpack, was extended to solve the reactive Euler equations for simulating detonations. A coupled combustion and ionization chemistry was developed in a single chemical kinetic model for methane oxidation. This model can be used to solve coupled MHD and detonation simulations. It was found that parasitic interactions from ionization chemistry with the magnetic field can reduce the detonation velocity by up to 8%, with a potential impact on power extraction of 15%. It is recommended that interactions between the detonation front and MHD field be considered. Higher gas temperatures and velocities were achieved owing to oxy-fuel detonations. Use of radical dissociation reactions in the reduced reaction mechanism, was found to be critical in predicting detonation temperature and velocity accurately.

01 COAL, LIGNITE, AND PEAT↗

DOE OSTI2024

Initial Demonstration of New Griffin Technologies for Simulating the Running-In Phase of Pebble Bed Reactors

Griffin is a reactor multiphysics modeling application based on MOOSE (Multiphysics Object-Oriented Simulation Environment) and specifically targeting transient modeling of advanced reactors. Griffin has been used recently to model pebble-bed reactors for the Nuclear Regulatory Commission (NRC) Office of Nuclear Regulatory Research and the Advanced Reactor Technology program. This modeling work has focused thus far on the direct calculation of equilibrium cores. This report documents an initial demonstration of a new running-in simulation capability. The new running-in capability is verified using the existing direct equilibrium core calculation capability. A simplified pebble-bed reactor model is then used to demonstrate the running-in simulation capability. This demonstration shows that Griffin is able to simulate years of operation during the running-in phase efficiently with each depletion step taking only several seconds. Two new technologies are also presented in this report which have been developed in Griffin that will be essential for improved accuracy both of the direct equilibrium core computation and the new running-in simulation capability. The first technology is an online cross section generation capability specifically targeted for pebble-bed reactors. This will improve the accuracy of the depletion calculation as the cross sections are generated at the exact core status. This also avoids the difficult step of pre-generating a separate standalone multigroup cross section set. Secondly, a newly implemented discretization for discontinuous finite element method (DFEM) SN transport in cylindrical (RZ) coordinates, which can be solved efficiently using the existing SN sweep solver, is discussed and some results are shown demonstrating the usefulness of the additional accuracy transport provides over a diffusion approximation.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

NASA NTRS2014

Validation of CFD Codes for Parawing Geometries in Subsonic to Supersonic Flows

Computational Fluid Dynamic studies of a rigid parawing at Mach numbers from 0.8 to 4.65 were carried out using three established inviscid, viscous and independent panel method codes. Pressure distributions along four chordwise sections of the wing were compared to experimental wind tunnel data gathered from NASA technical reports. Results show good prediction of the overall trends and magnitudes of the pressure distributions for the inviscid and viscous solvers. Pressure results for the panel method code diverge from test data at large angles of attack due to shock interaction phenomena. Trends in the flow behavior and their effect on the integrated force and moments on this type of wing are examined in detail using the inviscid CFD code results.

Direct Entry Trajectory↗

NASA NTRSDate not supplied

Turbulence Simulations of Transonic Flows over an NACA-0012 Airfoil

Three different simulation approaches, namely unsteady Reynolds-averaged Navier-Stokes (URANS), delayed detached-eddy simulation (DDES), and wall-modeled large-eddy simulation (WMLES) are employed to simulate transonic flow over an NACA-0012 airfoil at different angles of attack covering pre-and post-buffet-onset regimes. The freestream Mach number is 0.75,and the Reynolds number based on the chord length is 10million. These conditions are the same as the wind-tunnel experimental conditions of McDevitt and Okuno (1985). The NASA FUN3D solver is used for the simulations, which is an unstructured, compressible flow solver. The URANS simulations are performed using the Spalart-Allmaras (SA) model with the compressibility correction, the DDES predictions are based on the SA model, and the WMLES are performed using an equilibrium wall-model. The unsteady RANS simulations, only with the compressibility correction, predict the pre-and post-buffet characteristics, which compare well with the experimental results. DDES results predicted a lower buffet onset angle compared to experiment. The predicted shock locations are upstream of the locations predicted by URANS. Using a fine grid in the span wise direction, WMLES predictions show buffeting consistent with the experiment.

Transonic Buffet↗

NASA NTRSDate not supplied

Computational Modeling of Two Mars Powered Descent Vehicle Concepts Tested in the Langley Unitary Plan Wind Tunnel

Future human Mars missions will require powered descent starting at supersonic conditions, something which has never been done before at Mars. Computational powered descent flowfield simulations have been completed at full-scale Mars conditions, but the available ground test data are not suitable for calibrating computational uncertainties for aerodynamic interference on proposed Mars descent vehicles. Testing was conducted in the NASA Langley Unitary Plan Wind Tunnel in order to investigate the aerodynamic interference of sub-scale versions of two Mars powered descent vehicle concepts at supersonic Mach numbers (2.4 and 3.5): a model based on a blunt hypersonic inflatable aerodynamic decelerator (HIAD) and the second representing a more slender rigid vehicle with body flaps (CobraMRV). This paper covers computational flowfield predictions completed at wind tunnel conditions and comparisons to the test data. On the blowing HIAD models, the time-averaged pressure inboard of the nozzles was generally well-predicted, especially if the nozzles are canted outward, when the nozzles are located further from the nose. At intermediate CobraMRV thrust coefficients, CFD pressures are more accurately predicted than they are for the HIAD models, largely due to the nozzle locations and pointing directions. Overall, the CFD pressure coefficients were predicted within 0.2 of the steady pressure measurements for all blowing models, with smaller discrepancies at higher HIAD thrust, and larger discrepancies at lower HIAD CobraMRV thrust. All HIAD models were predicted to have a gradually decreasing axial force coefficient as the total thrust increases, in agreement with available pressure sensitive paint data. On models with canted nozzles or with nozzles further from the nose, the force coefficient was slightly higher for a given thrust. On the CobraMRV model, the CFD also shows consistent results between solvers and follows trends revealed in the data; the aerodynamic force coefficient remains near the non-blowing value at a tunnel Mach number of 2.4 regardless of thrust, and increases above that level at a Mach number of 3.5, consistent with the discrete pressure data. CFD analysis at tunnel and flight conditions will continue as flight system designs concepts mature.

Supersonic Retropropulsion↗

NASA NTRSDate not supplied

Computational Modeling of Two Mars Powered Descent Vehicle Concepts Tested in the Langley Unitary Plan Wind Tunnel

Supersonic Retropropulsion↗

DOE OSTI2022

Towards Compact Neural Networks via End-to-End Training: A Bayesian Tensor Approach with Automatic Rank Determination

Post-training model compression can reduce the inference costs of deep neural networks, but uncompressed training still consumes enormous hardware resources and energy. To enable low-energy training on edge devices, it is highly desirable to directly train a compact neural network from scratch with a low memory cost. Low-rank tensor decomposition is an effective approach to reduce the memory and computing costs of large neural networks. However, directly training low-rank tensorized neural networks is a very challenging task because it is hard to determine a proper tensor rank a priori, and the tensor rank controls both model complexity and accuracy. Here, this paper presents a novel end-to-end framework for low-rank tensorized training. We first develop a Bayesian model that supports various low-rank tensor formats (e.g., CANDECOMP/PARAFAC, Tucker, tensor-train, and tensor-train matrix) and reduces neural network parameters with automatic rank determination during training. Then we develop a customized Bayesian solver to train large-scale tensorized neural networks. Our training methods shows orders-of-magnitude parameter reduction and little accuracy loss (or even better accuracy) in the experiments. On a very large deep learning recommendation system with over 4.2 ×10 9 model parameters, our method can reduce the parameter number to 1.6 ×10 5 automatically in the training process (i.e., by 2.6 ×10 4 times) while achieving almost the same accuracy. Code is available at https://github.com/colehawkins/bayesian-tensor-rank-determination.

compact neural networks↗

DOE OSTI2025

Implementation of the D1S Methodology for Shutdown Dose Rate Calculations in the OpenMC Monte Carlo Particle Transport Code

We present an implementation of the direct one-step (D1S) methodology for shutdown dose rate (SDR) calculations in the OpenMC Monte Carlo particle transport code. In addition to being the first fully open-source D1S implementation, it is also the first to require no ad hoc source code or nuclear data library modifications. The code can seamlessly switch between production of prompt and decay photons based on a user input parameter, and the decay data needed for decay photon generation are made available through a depletion chain file, which is already used for OpenMC’s built-in depletion/activation solver. A set of Python functions significantly eases the burden of computing and applying time correction factors needed to properly account for the time dependence of radionuclide activity. To assess the accuracy of the D1S implementation, SDR calculations have been carried out for three problems: a prism of iron irradiated by 14-MeV neutrons, the ITER port plug computational benchmark, and the Frascati Neutron Generator (FNG) ITER dose rate benchmark problem from the Shielding INtegral Benchmark Archive and Database (SINBAD). For each of these problems, comparisons were made to calculations using the rigorous two-step (R2S) method. The results on the iron prism problem illustrate how the D1S method achieves superior spatial resolution compared to the R2S method without the need for spatial discretization of the activation regions. The D1S and R2S results for the ITER port plug benchmark agree well with previously reported results in the literature. While the D1S results are 10% to 15% lower than the R2S results, this may be due to stochastic uncertainty and/or spatial discretization in the R2S calculations. On the FNG dose rate benchmark problem, the D1S method produces dose rate estimates that are within 4% of the dose rates predicted using a cell-based R2S workflow. The D1S estimates of the SDR are also in reasonable agreement with the experimental measurements and show the same basic trends that have been observed in previous works. A qualitative analysis of the execution time and uncertainty for the R2S and D1S workflows suggests that the D1S method would attain a higher figure of merit.

D1S method↗

NASA NTRS2004

Evaluation of a Multigrid Scheme for the Incompressible Navier-Stokes Equations

A fast multigrid solver for the steady, incompressible Navier-Stokes equations is presented. The multigrid solver is based upon a factorizable discrete scheme for the velocity-pressure form of the Navier-Stokes equations. This scheme correctly distinguishes between the advection-diffusion and elliptic parts of the operator, allowing efficient smoothers to be constructed. To evaluate the multigrid algorithm, solutions are computed for flow over a flat plate, parabola, and a Karman-Trefftz airfoil. Both nonlifting and lifting airfoil flows are considered, with a Reynolds number range of 200 to 800. Convergence and accuracy of the algorithm are discussed. Using Gauss-Seidel line relaxation in alternating directions, multigrid convergence behavior approaching that of O(N) methods is achieved. The computational efficiency of the numerical scheme is compared with that of Runge-Kutta and implicit upwind based multigrid methods.

Swanson, R. C.↗

NASA NTRS2008

Aeroacoustic Simulations of Tandem Cylinders with Subcritical Spacing

Tandem cylinders are being studied because they model a variety of component level interactions of landing gear. The present effort is directed at the case of two identical cylinders with their centroids separated in the streamwise direction by 1.435 diameters. Experiments in the Basic Aerodynamic Research Tunnel and Quiet Flow Facility at NASA Langley Research Center have provided an extensive experimental database of the nearfield flow and radiated noise. The measurements were conducted at a Mach number of 0.1285 and Reynolds number of 1.66x10(exp 5) based on the cylinder diameter. A trip was used on the upstream cylinder to insure a fully turbulent flow separation and, hence, to simulate a major aspect of high Reynolds number flow. The parallel computational effort uses the three-dimensional Navier-Stokes solver CFL3D with a hybrid, zonal turbulence model that turns off the turbulence production term everywhere except in a narrow ring surrounding solid surfaces. The experiments exhibited an asymmetry in the surface pressure that was persistent despite attempts to eliminate it through small changes in the configuration. To model the asymmetry, the simulations were run with the cylinder configuration at a nonzero but small angle of attack. The computed results and experiments are in general agreement that vortex shedding for the spacing studied herein is weak relative to that observed at supercritical spacings. Although the shedding was subdued in the simulations, it was still more prominent than in the experiments. Overall, the simulation comparisons with measured near-field data and the radiated acoustics are reasonable, especially if one is concerned with capturing the trends relative to larger cylinder spacings. However, the flow details of the 1.435 diameter spacing have not been captured in full even though very fine grid computations have been performed. Some of the discrepancy may be associated with the simulation s inexact representation of the experimental configuration, but numerical and flow modeling errors are also likely contributors to the observed differences.

Lockard, David P.↗

DOE OSTI2026

Hierarchical-embedding autoencoder with a predictor as efficient architecture for learning time-evolution in multi-scale turbulent flows

We introduce a scale-aware, data-driven deep learning modeling framework for accurately predicting the time evolution of multi-scale turbulent plasma and liquid flows. The approach is motivated by the idea of scale separation. Structures of vastly different length scales emerge in these systems, and interactions between these structures occur only locally. To exploit this structure, the flow state is transformed by a hierarchical, fully convolutional autoencoder, not into a single embedding layer as in conventional convolutional surrogate models, but into a series of embedding layers. A stepwise training strategy ensures that fine-scale features are encoded on a high-resolution grid, while larger structures are represented on progressively coarser layers. The time evolution predictor advances all embedding layers in sync, capturing local interactions between features at the same scale as well as between all scales. This approach enables efficient modeling of multi-scale systems since negligible interactions between distant, small-scale structures do not need to be directly modeled. Our hierarchical-embedding autoencoder with a predictor framework is evaluated on canonical examples of multi-scale turbulence: two-dimensional Kolmogorov flow and Hasegawa–Wakatani plasma turbulence. In both cases, the proposed framework significantly improves predictive accuracy relative to conventional convolutional network architectures. A significant improvement in prediction accuracy was observed for crucial statistical characteristics of the Hasegawa–Wakatani plasma as well as for individual trajectories of the Kolmogorov flow turbulence. Importantly, the model's rollout for the Hasegawa–Wakatani problem demonstrates a four-order-of-magnitude speedup compared to traditional numerical solvers.

Khrabry, Alexander I. [Princeton Univ., NJ (United↗

NASA NTRS1999

Time-Shifted Boundary Conditions Used for Navier-Stokes Aeroelastic Solver

Under the Advanced Subsonic Technology (AST) Program, an aeroelastic analysis code (TURBO-AE) based on Navier-Stokes equations is currently under development at NASA Lewis Research Center s Machine Dynamics Branch. For a blade row, aeroelastic instability can occur in any of the possible interblade phase angles (IBPA s). Analyzing small IBPA s is very computationally expensive because a large number of blade passages must be simulated. To reduce the computational cost of these analyses, we used time shifted, or phase-lagged, boundary conditions in the TURBO-AE code. These conditions can be used to reduce the computational domain to a single blade passage by requiring the boundary conditions across the passage to be lagged depending on the IBPA being analyzed. The time-shifted boundary conditions currently implemented are based on the direct-store method. This method requires large amounts of data to be stored over a period of the oscillation cycle. On CRAY computers this is not a major problem because solid-state devices can be used for fast input and output to read and write the data onto a disk instead of storing it in core memory.

Srivastava, Rakesh↗