A system for testing iterative methods for solving nonlinear equations
System of computer programs for testing iterative methods for solving nonlinear systems of equations
Mesztenyi, C. K.↗
SEARCH · Search NASA
Results for “equation solving”
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.
Optical neural engine for solving scientific partial differential equations
Abstract Solving partial differential equations (PDEs) is the cornerstone of scientific research and development. Data-driven machine learning (ML) approaches are emerging to accelerate time-consuming and computation-intensive numerical simulations of PDEs. Although optical systems offer high-throughput and energy-efficient ML hardware, their demonstration for solving PDEs is limited. Here, we present an optical neural engine (ONE) architecture combining diffractive optical neural networks for Fourier space processing and optical crossbar structures for real space processing to solve time-dependent and time-independent PDEs in diverse disciplines, including Darcy flow equation, the magnetostatic Poisson’s equation in demagnetization, the Navier-Stokes equation in incompressible fluid, Maxwell’s equations in nanophotonic metasurfaces, and coupled PDEs in a multiphysics system. We numerically and experimentally demonstrate the capability of the ONE architecture, which not only leverages the advantages of high-performance dual-space processing for outperforming traditional PDE solvers and being comparable with state-of-the-art ML models but also can be implemented using optical computing hardware with unique features of low-energy and highly parallel constant-time processing irrespective of model scales and real-time reconfigurability for tackling multiple tasks with the same architecture. The demonstrated architecture offers a versatile and powerful platform for large-scale scientific and engineering computations.
Tang, Yingheng (ORCID:0009000153622546)↗
Compressible Flow Toolbox
The Compressible Flow Toolbox is primarily a MATLAB-language implementation of a set of algorithms that solve approximately 280 linear and nonlinear classical equations for compressible flow. The toolbox is useful for analysis of one-dimensional steady flow with either constant entropy, friction, heat transfer, or Mach number greater than 1. The toolbox also contains algorithms for comparing and validating the equation-solving algorithms against solutions previously published in open literature. The classical equations solved by the Compressible Flow Toolbox are as follows: The isentropic-flow equations, The Fanno flow equations (pertaining to flow of an ideal gas in a pipe with friction), The Rayleigh flow equations (pertaining to frictionless flow of an ideal gas, with heat transfer, in a pipe of constant cross section), The normal-shock equations, The oblique-shock equations, and The expansion equations.
Melcher, Kevin J.↗
Streamwise Upwind, Moving-Grid Flow Algorithm
Extension to moving grids enables computation of transonic flows about moving bodies. Algorithm computes unsteady transonic flow on basis of nondimensionalized thin-layer Navier-Stokes equations in conservation-law form. Solves equations by use of computational grid based on curvilinear coordinates conforming to, and moving with, surface(s) of solid body or bodies in flow field. Simulates such complicated phenomena as transonic flow (including shock waves) about oscillating wing. Algorithm developed by extending prior streamwise upwind algorithm solving equations on fixed curvilinear grid described in "Streamwise Algorithm for Simulation of Flow" (ARC-12718).
Goorjian, Peter M.↗
Introduction of the Adding and Doubling Method for Solving Bateman Equations for Nuclear Fuel Depletion
This paper introduces and evaluates the Adding and Doubling Method (ADM) for solving the Bateman equations for depletion systems with varying numbers of nuclides and compares it to the Chebyshev Rational Approximation Method (CRAM), both implemented in the reactor physics analysis application Griffin. ADM, when applied to the Crank-Nicolson Finite Difference method, can produce results comparable in accuracy and precision to CRAM with comparable run times for systems with 35 or 297 nuclides. For systems with more than 300 nuclides, the matrix-matrix operations required by ADM are significantly more costly than the matrix-vector operations required by CRAM, making CRAM the more efficient method for systems with large numbers of nuclides. ADM is an accurate method that maintains other advantages over CRAM in that it does not depend on pre-generated coefficients or require complex number operations. ADM also manages to outperform CRAM by a factor of more than 250 in terms of run time for depletion systems that require multiple Bateman solves while the depletion matrix and time step size remain constant over all depletion intervals.
11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
Solving Modal Equations of Motion with Initial Conditions Using MSC/NASTRAN DMAP: Coupled Versus Uncoupled Integration - Part 2
By utilizing MSC/NASTRAN DMAP (Direct Matrix Abstraction Program) in an existing NASA Lewis Research Center coupled loads methodology, solving modal equations of motion with initial conditions is possible using either coupled (Newmark-Beta) or uncoupled (exact mode superposition) integration available within module TRD1. Both the coupled and newly developed exact mode superposition methods have been used to perform transient analyses of various space systems. However, experience has shown that in most cases, significant time savings are realized when the equations of motion are integrated using the uncoupled solver instead of the coupled solver. Through the results of a real-world engineering analysis, advantages of using the exact mode superposition methodology are illustrated.
Barnett, Alan R.↗
On the equivalence of Gaussian elimination and Gauss-Jordan reduction in solving linear equations
A novel general approach to round-off error analysis using the error complexity concepts is described. This is applied to the analysis of the Gaussian Elimination and Gauss-Jordan scheme for solving linear equations. The results show that the two algorithms are equivalent in terms of our error complexity measures. Thus the inherently parallel Gauss-Jordan scheme can be implemented with confidence if parallel computers are available.
Tsao, Nai-Kuan↗
Solving modal equations of motion with initial conditions using MSC/NASTRAN DMAP. Part 1: Implementing exact mode superposition
Within the MSC/NASTRAN DMAP (Direct Matrix Abstraction Program) module TRD1, solving physical (coupled) or modal (uncoupled) transient equations of motion is performed using the Newmark-Beta or mode superposition algorithms, respectively. For equations of motion with initial conditions, only the Newmark-Beta integration routine has been available in MSC/NASTRAN solution sequences for solving physical systems and in custom DMAP sequences or alters for solving modal systems. In some cases, one difficulty with using the Newmark-Beta method is that the process of selecting suitable integration time steps for obtaining acceptable results is lengthy. In addition, when very small step sizes are required, a large amount of time can be spent integrating the equations of motion. For certain aerospace applications, a significant time savings can be realized when the equations of motion are solved using an exact integration routine instead of the Newmark-Beta numerical algorithm. In order to solve modal equations of motion with initial conditions and take advantage of efficiencies gained when using uncoupled solution algorithms (like that within TRD1), an exact mode superposition method using MSC/NASTRAN DMAP has been developed and successfully implemented as an enhancement to an existing coupled loads methodology at the NASA Lewis Research Center.
Abdallah, Ayman A.↗
Quantum advantage for differential equation analysis
Quantum algorithms for differential equation solving, data processing, and machine learning potentially offer an exponential speedup over all known classical algorithms. However, there also exist obstacles to obtaining this potential speedup in useful problem instances. The essential obstacle for quantum differential equation solving is that outputting useful information may require difficult postprocessing, and the essential obstacle for quantum data processing and machine learning is that inputting the data is a difficult task just by itself. In this study, we demonstrate that, when combined, these difficulties solve one another. We show how the output of quantum differential equation solving can serve as the input for quantum data processing and machine learning, allowing dynamical analysis in terms of principal components, power spectra, and wavelet decompositions. To illustrate this, we consider continuous-time Markov processes on epidemiological and social networks. These quantum algorithms provide an exponential advantage over existing classical Monte Carlo methods.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
An attempt to solve differential equations symbolically
Writing programs in interactive FORMAC language to implement Picard iteration in solving systems of ordinary differential equations
Gershwin, S. B.↗
Derivation of approximate equations for solving the planar rendezvous problem
Simplified and accurate relative equations of motion for two impulse orbital rendezvous
Price, D. B.↗
Prediction of Shock Wave Structure in Weakly Ionized Gas Flow by Solving MGD Equation
This paper reports the recent research results of shockwave structure predictions using a new developed code. The modified Rankine-Hugoniot relations across a standing normal shock wave are discussed and adopted to obtain jump conditions. Coupling a electrostatic body force to the Burnett equations, the weakly ionized flow field across the shock wave was solved. Results indicated that the Modified Rankine-Hugoniot equations for shock wave are valid for a wide range of ionization fraction. However, this model breaks down with small free stream Mach number and with large ionization fraction. The jump conditions also depend on the value of free stream pressure, temperature and density. The computed shock wave structure with ionization provides results, which indicated that shock wave strength may be reduced by existence of weakly ionized gas.
Deng, Z. T.↗
Preconditioned methods for solving the incompressible and low speed compressible equations
The introduction of artificial time derivatives which permit faster convergence to the steady state is used to precondition steady state incompressible equation-solving acceleration methods. An optimal value is obtained for beta, given constant alpha, where alpha and beta are arbitrary functions in the general preconditioning stage. A symmetric hyperbolic system is formed by the resulting incompressible equations. Compressible equations in conservation form with slow flow are also considered.
Turkel, Eli↗
Separable physics-informed DeepONet: Breaking the curse of dimensionality in physics-informed machine learning
The deep operator network (DeepONet) has shown remarkable potential in solving partial differential equations (PDEs) by mapping between infinite-dimensional function spaces using labeled datasets. However, in scenarios lacking labeled data, the physics-informed DeepONet (PI-DeepONet) approach, which utilizes the residual loss of the governing PDE to optimize the network parameters, faces significant computational challenges, particularly due to the curse of dimensionality. This limitation has hindered its application to high-dimensional problems, making even standard 3D spatial with 1D temporal problems computationally prohibitive. Additionally, the computational requirement increases exponentially with the discretization density of the domain. Here, to address these challenges and enhance scalability for high-dimensional PDEs, we introduce the Separable physics-informed DeepONet (Sep-PI-DeepONet). This framework employs a factorization technique, utilizing sub-networks for individual one-dimensional coordinates, thereby reducing the number of forward passes and the size of the Jacobian matrix required for gradient computations. By incorporating forward-mode automatic differentiation (AD), we further optimize computational efficiency, achieving linear scaling of computational cost with discretization density and dimensionality, making our approach highly suitable for high-dimensional PDEs. We demonstrate the effectiveness of Sep-PI-DeepONet through three benchmark PDE models: the viscous Burgers’ equation, Biot’s consolidation theory, and a parameterized heat equation. Our framework maintains accuracy comparable to the conventional PI-DeepONet while reducing training time by two orders of magnitude. Notably, for the heat equation solved as a 4D problem, the conventional PI-DeepONet was computationally infeasible (estimated 289.35 h), while the Sep-PI-DeepONet completed training in just 2.5 h. These results underscore the potential of Sep-PI-DeepONet in efficiently solving complex, high-dimensional PDEs, marking a significant advancement in physics-informed machine learning.
Neural operator↗
Co-training of multiple neural networks for simultaneous optimization and training of physics-informed neural networks for composite curing
This paper introduces a Physics-Informed Neural Network (PINN) technique that co-trains neural networks (NNs) that represent each function in a system of equations to simultaneously solve equations representing an out-of-autoclave (OOA) cure process while conducting optimization in adherence to process requirements. Specifically, this co-training approach benefits from using NNs to represent OOA inputs (air temperature profile) and outputs (part and tool temperature profiles and degree of cure). Production requirements can then be levied on the inputs, such as maximum air temperature and minimum cure cycle, and simultaneously on the outputs, such as degree of cure, maximum part temperature, and part temperature rate limits. The technique is validated with finite element (FE) simulations and physical experiments for curing a Toray T830H-6 K/3900-2D composite panel. Furthermore, this novel approach efficiently models and optimizes the OOA cure process.
Composite curing↗
Some MACSYMA program for solving difference equations
A set of MACSYMA programs are described for finding closed form solutions to linear recurrence relations in equations having either constant or variable coefficients. In the homogenous case, a polymonial equation is obtained and the solution to the recurrence relation can be written as a linear combination of the roots of the polynomial. Exponential generating functions are used to solve variable coefficient relations. Taking successive derivatives and using the recurrence relation, an ordinary differential equation is obtained. Expanding the solution to the differential equation in a Taylor series, shows that the nth term of the series is the solution to the recurrence relation. For second order recurrences, a check is made for those that can be solved in terms of Bessel functions.
Ivie, J.↗
User Guide for Compressible Flow Toolbox Version 2.1 for Use With MATLAB(Registered Trademark); Version 7
This report provides a user guide for the Compressible Flow Toolbox, a collection of algorithms that solve almost 300 linear and nonlinear classical compressible flow relations. The algorithms, implemented in the popular MATLAB programming language, are useful for analysis of one-dimensional steady flow with constant entropy, friction, heat transfer, or shock discontinuities. The solutions do not include any gas dissociative effects. The toolbox also contains functions for comparing and validating the equation-solving algorithms against solutions previously published in the open literature. The classical equations solved by the Compressible Flow Toolbox are: isentropic-flow equations, Fanno flow equations (pertaining to flow of an ideal gas in a pipe with friction), Rayleigh flow equations (pertaining to frictionless flow of an ideal gas, with heat transfer, in a pipe of constant cross section.), normal-shock equations, oblique-shock equations, and Prandtl-Meyer expansion equations. At the time this report was published, the Compressible Flow Toolbox was available without cost from the NASA Software Repository.
Melcher, Kevin J.↗
On Spurious Numerics in Solving Reactive Equations
The objective of this study is to gain a deeper understanding of the behavior of high order shock-capturing schemes for problems with stiff source terms and discontinuities and on corresponding numerical prediction strategies. The studies by Yee et al. (2012) and Wang et al. (2012) focus only on solving the reactive system by the fractional step method using the Strang splitting (Strang 1968). It is a common practice by developers in computational physics and engineering simulations to include a cut off safeguard if densities are outside the permissible range. Here we compare the spurious behavior of the same schemes by solving the fully coupled reactive system without the Strang splitting vs. using the Strang splitting. Comparison between the two procedures and the effects of a cut off safeguard is the focus the present study. The comparison of the performance of these schemes is largely based on the degree to which each method captures the correct location of the reaction front for coarse grids. Here "coarse grids" means standard mesh density requirement for accurate simulation of typical non-reacting flows of similar problem setup. It is remarked that, in order to resolve the sharp reaction front, local refinement beyond standard mesh density is still needed.
Numerics↗