Search NASA⌕ Search

SEARCH · Search NASA

Results for “implicit solvers”

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 181 records · Page 10

An implicit Navier-Stokes code for turbulent flow modeling

This paper presents a numerical approach to calculating turbulent flows employing advanced turbulence models. The main features include a line-by-line Gauss-Seidel algorithm using Roe's approximate Riemann solver, TVD numerical schemes, implicit boundary conditions and a decoupled turbulence-model solver. Based on the problems tested so far, the method has consistently demonstrated its ability in offering accuracy, boundedness and a fast rate of convergence to steady-state solution.

Huang, P. G.↗

Efficient discontinuous Galerkin implementations and preconditioners for implicit unsteady compressible flow simulations

This work presents and compares efficient implementations of high-order discontinuous Galerkin methods: a modal matrix-free discontinuous Galerkin (DG) method, a hybridizable discontinuous Galerkin (HDG) method, and a primal formulation of HDG, applied to the implicit solution of unsteady compressible flows. The matrix-free implementation allows for a reduction of the memory footprint of the solver when dealing with implicit time-accurate discretizations. HDG reduces the number of globally-coupled degrees of freedom relative to DG, at high order, by statically condensing element-interior degrees of freedom from the system in favor of face unknowns. The primal formulation further reduces the element-interior degrees of freedom by eliminating the gradient as a separate unknown. This paper introduces a p-multigrid preconditioner implementation for these discretizations and presents results for various flow problems. Benefits of the p-multigrid strategy relative to simpler, less expensive, preconditioners are observed for stiff systems, such as those arising from low-Mach number flows at high-order approximation. The p-multigrid preconditioner also shows excellent scalability for parallel computations. Finally, additional savings in both speed and memory occur with a matrix-free/reduced version of the preconditioner.

97 MATHEMATICS AND COMPUTING↗

Parallel performance investigations of an unstructured mesh Navier-Stokes solver

A Reynolds-averaged Navier-Stokes solver based on unstructured mesh techniques for analysis of high-lift configurations is described. The method makes use of an agglomeration multigrid solver for convergence acceleration. Implicit line-smoothing is employed to relieve the stiffness associated with highly stretched meshes. A GMRES technique is also implemented to speed convergence at the expense of additional memory usage. The solver is cache efficient and fully vectorizable, and is parallelized using a two-level hybrid MPI-OpenMP implementation suitable for shared and/or distributed memory architectures, as well as clusters of shared memory machines. Convergence and scalability results are illustrated for various high-lift cases.

Mavriplis, Dimitri J.↗

Recent Progress in OVERFLOW Convergence Improvements

Improvements have been made to the implicit symmetric successive overrelaxation algorithm in the OVERFLOW 2.3 structured, overset grid, computational fluid dynamics flow solver. These improvements, consisting of implicit boundary conditions, improved flux Jacobian linearizations, and CFL number ramping, are a series of evolutionary changes to the linear solver that have resulted in increased nonlinear convergence rates and faster time to solution. A series of test cases are presented that demonstrate the effect of the changes through comparison with the original SSOR path and other linear solver implementations within OVERFLOW.

Joseph M Derlaga↗

A Simplified FANG Cell-Centered Finite-Volume Method and Comparison with Other Methods for Trouble-Prone Grids

We propose a simplication of the face-averaged nodal-gradient (FANG) method for a cell-centered finite-volume Euler/Navier-Stokes solver on arbitrary grids, and compare it with other gradient methods for trouble-prone grids in two dimensions. The implementation of the FANG method is simplified by adding the face-neighbor cells of the cells around a node to a least-squares gradient stencil. The resulting method is stable for both triangular and quadrilateral grids. Although it increases the residual stencil for triangular grids, it allows the solver to work seamlessly for mixed grids and greatly simplifies the implementation, especially in three dimensions. For comparison, only explicit weighted/unweighted least-squares cell-centered and nodal gradient methods are considered. These gradients are used in both inviscid and viscous schemes, and we investigate their impact on the iterative convergence of an implicit defect-correction solver on difficult grids such as highly-curved-and-thin grids and highly distorted anisotropic grids. Finally, we will also consider a face-stencil-based limiter and compare it with a conventional cell-stencil-based limiter.

Hiroaki Nishikawa↗

Aerodynamic optimization studies on advanced architecture computers

The approach to carrying out multi-discipline aerospace design studies in the future, especially in massively parallel computing environments, comprises of choosing (1) suitable solvers to compute solutions to equations characterizing a discipline, and (2) efficient optimization methods. In addition, for aerodynamic optimization problems, (3) smart methodologies must be selected to modify the surface shape. In this research effort, a 'direct' optimization method is implemented on the Cray C-90 to improve aerodynamic design. It is coupled with an existing implicit Navier-Stokes solver, OVERFLOW, to compute flow solutions. The optimization method is chosen such that it can accomodate multi-discipline optimization in future computations. In the work , however, only single discipline aerodynamic optimization will be included.

Chawla, Kalpana↗

Implementation and (Inverse Modified) Error Analysis for Implicitly Templated ODE-Nets

We focus on learning unknown dynamics from data using ODE-nets templated on implicit numerical initial value problem solvers. First, we perform inverse modified error analysis of the ODE-nets using unrolled implicit schemes for ease of interpretation. It is shown that training an ODE-net using an unrolled implicit scheme returns a close approximation of an inverse modified differential equation (IMDE). In addition, we establish a theoretical basis for hyperparameter selection when training such ODE-nets, whereas current strategies usually treat numerical integration of ODE-nets as a black box. We thus formulate an adaptive algorithm which monitors the level of error and adapts the number of (unrolled) implicit solution iterations during the training process, so that the error of the unrolled approximation is less than the current learning loss. This helps accelerate training while maintaining accuracy. Several numerical experiments are performed to demonstrate the advantages of the proposed algorithm compared to nonadaptive unrollings and validate the theoretical analysis. Here, we also note that this approach naturally allows for incorporating partially known physical terms in the equations, giving rise to what is termed “gray box” identification.

ODE-nets↗

Demonstration of a new unstructured mesh IMC x-ray transport capability in LAP codes

In this document, the Advanced Simulation and Computing (ASC) Transport Project’s Jayenne team presents evidence that the Los Alamos National Laboratory (LANL) Level 2 milestone statement, Demonstrate a new unstructured mesh IMC x-ray transport capability in LAP codes, due Q4 FY20, has been satisfied with the release and integration of the Jayenne team’s Implicit Monte Carlo (IMC) solver libraries into the lumos multiphysics solver. The Jayenne project’s software includes the algorithms and features prescribed by the milestone description. The lumos software is maintained by the Lagrangian Applications Project (LAP) along with the hydrodynamics code flag and operates under the same ASC program as Jayenne. A set of demonstration problems has been specified, executed, and analyzed.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Implementation of Implicit Adaptive Mesh Refinement in an Unstructured Finite-Volume Flow Solver

This paper explores the implementation of adaptive mesh refinement in an unstructured, finite-volume solver. Unsteady and steady problems are considered. The effect on the recovery of high-order numerics is explored and the results are favorable. Important to this work is the ability to provide a path for efficient, implicit time advancement. A method using a simple refinement sensor based on undivided differences is discussed and applied to a practical problem: a shock-shock interaction on a hypersonic, inviscid double-wedge. Cases are compared to uniform grids without the use of adapted meshes in order to assess error and computational expense. Discussion of difficulties, advances, and future work prepare this method for additional research. The potential for this method in more complicated flows is described.

Schwing, Alan M.↗

Speedup of UEDGE Parameter Scans Using Machine-Learning Optimized OpenMP Parallelization and a Continuation Solver

This article presents the OpenMP parallelization of the preconditioning Jacobian assembly and right‐hand side residual evaluation in UEDGE. A continuation algorithm, utilizing the internal NKSOL implicit Jacobian‐Free Newton‐Krylov solver to efficiently scan physical parameters, is also presented. The implemented parallelization reduces the computational time for a benchmark scan run on 32 threads by compared to the serial version when using trained random forest regression models to identify the optimal decomposition of the system of equations. Random forest regression models applied to the UEDGE time‐dependent and continuation solver algorithms did not yield meaningful improvement in computational performance. A benchmark DIII‐D gas injection rate scan in the 0.35–0.75 kA interval, performed on a test cluster using the parallelized code and continuation solver, produced 1066 steady‐state solutions with a 22 s average wall‐clock computational time per steady‐state solution.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A high-order numerical algorithm for DNS of low-Mach-number reactive flows with detailed chemistry and quasi-spectral accuracy

A novel and efficient algorithm is presented in this paper to deal with DNS of turbulent reacting flows under the low-Mach-number assumption, with detailed chemistry and a quasi-spectral accuracy. The temporal integration of the equations relies on an operating-split strategy, where chemical reactions are solved implicitly with a stiff solver and the convection-diffusion operators are solved with a Runge-Kutta-Chebyshev method. The spatial discretisation is performed with high-order compact schemes, and a FFT based constant-coefficient spectral solver is employed to solve a variable-coefficient Poisson equation. The numerical implementation takes advantage of the 2DECOMP&FFT libraries developed by, which are based on a pencil decomposition method of the domain and are proven to be computationally very efficient. An enhanced pressure-correction method is proposed to speed up the achievement of machine precision accuracy. It is demonstrated that a second-order accuracy is reached in time, while the spatial accuracy ranges from fourth-order to sixth-order depending on the set of imposed boundary conditions. In conclusion, the software developed to implement the present algorithm is called HOLOMAC, and its numerical efficiency opens the way to deal with DNS of reacting flows to understand complex turbulent and chemical phenomena in flames.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Simulation and fitting of complex reaction network TPR: The key is the objective function

In this research, a method has been developed for finding improved fits during simulation and fitting of data from complex reaction network temperature programmed reactions (CRN-TPR). It was found that simulation and fitting of CRN-TPR presents additional challenges relative to simulation and fitting of simpler TPR systems. The method used here can enable checking the plausibility of proposed chemical mechanisms and kinetic models. The most important finding was that when choosing an objective function, use of an objective function that is based on integrated production provides more utility in finding improved fits when compared to an objective function based on the rate of production. The response surface produced by using the integrated production is monotonic, suppresses effects from experimental noise, requires fewer points to capture the response behavior, and can be simulated numerically with smaller errors. For CRN-TPR, there is increased importance (relative to simple reaction network TPR) in resolving of peaks prior to fitting, as well as from weighting of experimental data points. Using an implicit ordinary differential equation solver was found to be inadequate for simulating CRN-TPR. Lastly, the method employed here was capable of attaining improved fits in simulation and fitting of CRN-TPR when starting with a postulated mechanism and physically realistic initial guesses for the kinetic parameters.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Diffusion-driven Fluid Dynamics in Ideal Gases and Plasmas

The classical transport theory based on Chapman-Enskog methods provides self-consistent approximations for the kinetic flux of mass, heat, and momentum in a fluid limit characterized with a small Knudsen number. The species mass fluxes relative to the center of mass, or “diffusive fluxes,” are expressed as functions of known gradient quantities with kinetic coefficients evaluated using similar analyses for mixtures of gases or plasma components. The sum over species of the diffusive mass fluxes is constrained to be zero in the Lagrange frame, and thus results in a non-zero molar flux leading to a pressure perturbation. At an interface between two species initially in pressure equilibrium, the pressure perturbation driven by the diffusive molar flux induces a center of mass velocity directed from the species of greater atomic mass towards the lighter atomic mass species. As the ratio of the species particle masses increases, this center of mass velocity carries an increasingly greater portion of the mass across the interface and for a particle mass ratio greater than about two, the center of mass velocity carries more mass than the gradient driven diffusion flux. Early time transients across an interface between two species in a 1D plasma regime and initially in equilibrium are compared using three methods; a fluid code with closure in a classical transport approximation, a particle in cell simulation, and an implicit Fokker-Planck solver for the particle distribution functions. The early time transient phenomenology is shown to be similar in each of the computational simulation methods, including a pressure perturbation associated with the stationary “induced” component of the center of mass velocity which decays to pressure equilibrium during diffusion. At early times, the diffusive process generates pressure and velocity waves which propagate outward from the interface and are required to maintain momentum conservation. The energy in the outgoing waves dissipates as heat in viscous regions, and it is hypothesized that these diffusion driven waves may sustain fluctuations in less viscid finite domains after reflections from the boundaries. Finally, these fluid dynamic phenomena are similar in gases or plasmas and occur in flow transients with a moderate Knudsen number. The analysis and simulation results show how the kinetic flux, represented in the fluid transport closure, directly modifies the mass averaged flow described with the Euler equations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

scikit-SUNDAE ((SUN)DIALS Differential Algebraic Equations) [SWR-24-137]

Scikit-SUNDAE provides Python bindings to SUNDIALS integrators. The implicit differential algebraic (IDA) solver and C-based variable-coefficient ordinary differential equations (CVODE) solver are both included. The name SUNDAE combines (SUN)DIALS and DAE, which stands for differential algebraic equations. Solvers specific to DAE problems are not frequently available in Python. An ordinary differential equation (ODE) solver is also included for completeness. ODEs can be categorized as a subset of DAEs (i.e., DAEs with no algebraic constraints). https://pypi.org/project/scikit-sundae

Randall, Corey↗

The General Atomics Center for Magnetohydrodynamic Modeling (CEMM) Final Report

The Center for Extended Magnetohydrodynamic Modeling (CEMM) is built around the further development, verification, validation, and other application of two major extended magnetohydrodynamics (MHD) codes NIMROD and M3D-C1. The General Atomics (GA) CEMM Program focuses on three main research areas: extension of the M3D-C1 MHD code plasma response calculations to new regimes, coupling of the DK4D drift kinetic equation solver to a temperature evolution equation, and simulations of disruption mitigation for unstable plasmas with pre-existing magnetic islands using massive gas injection (MGI) with the NIMROD MHD code. A systematic scan of the zero-crossing of the rotation profile in a DIII-D ITER-similar shape equilibrium using linear, time-independent modeling with M3D-C1 was performed. A new Python utility, autoC1, has been developed to automate the running of M3D-C1 for linear simulations. A number of ELM suppression experiments were analyzed including a validation and cross-code verification with the MARS-F MHD code of magnetic signal phase shifts across ELM-suppression bifurcations. A linearized, time-dependent temperature equation was tightly coupled to the electron and ion drift-kinetic equations (DKEs) in the DK4D drift-kinetic equation solver. Various explicit and implicit implementations were attempted, but all methods proved numerically unstable, even with the addition of numerical dissipation. Several simulations were completed investigating the effects of pre-existing islands on disruption mitigation by MGI in DIII-D using NIMROD.

3D MHD Simulation↗

Application of an axisymmetric implicit blunt body procedure - Computation of solar wind flows past terrestrial planets

A recently developed axisymmetric implicit unsteady Euler equation solver has been applied to the calculation of solar wind flows past terrestrial planets. The primary emphasis of this paper is on the application of that computational technique to the calculation of the flow field properties of steady, axisymmetric, supersonic/hypersonic flows past blunt-nosed shapes characteristic of interplanetary solar wind flows past both magnetic and nonmagnetic planets. The purpose is to demonstrate both the accuracy and flexibility of this new technique as well as to present a novel application of such supersonic blunt-body procedures.

Chaussee, D. S.↗