Search NASA⌕ Search

SEARCH · Search NASA

Results for “Convergence criteria”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

C3F: Collaborative Container-based Model Coupling Framework

Solving complex real-world grand challenge problems requires in-depth collaboration of researchers from multiple disciplines. Such collaboration often involves harnessing multiscale and multi-dimensional data and combining models from different fields to simulate systems. However, the progress on this front has been limited mainly due to significant gaps in domain knowledge and tools that are typically employed in silos of the domains. Researchers from different fields face considerable barriers to understanding and reusing each other’s data/models in order to collaborate effectively. For example, in solving the global sustainability problems, researchers from hydrology, climate science, agriculture, and economics need to run their respective models to study different components of the global and local food, energy and water systems while, at the same time, need to interact with other researchers and integrate the results of one model with another. Developing this kind of model coupling workflow calls for (1) a large amount of data being processed and exchanged across domains and organizations, (2) identifying and processing the output of one model to make it ready for integration into another model, (3) controlling the workflow dynamically so that it runs until a certain convergence condition or other criteria is met, and (4) close collaboration among the modelers to explore, tune, and test the configuration and data transformation needed to link the models. We have developed C3F, a flexible collaborative model coupling framework to help researchers accelerate their model integration and linking efforts by leveraging advanced cyberinfrastructure such as high-performance computing and virtual containers. In this paper, we describe our experience and lessons learned in developing this cyberinfrastructure solution to support the linking of Water Balance Model (WBM) and SIMPLE-G agricultural economic model in an NSF funded INFEWS project and a DOE-funded Program on Coupled Human and Earth Systems (PCHES) to study the implications of groundwater scarcity for food-energy-water systems. The C3F model coupling framework can be extended to facilitate other model linkages as well.

containerization↗
NASA NTRS1989

Numerical algorithms for computations of feedback laws arising in control of flexible systems

Several continuous models will be examined, which describe flexible structures with boundary or point control/observation. Issues related to the computation of feedback laws are examined (particularly stabilizing feedbacks) with sensors and actuators located either on the boundary or at specific point locations of the structure. One of the main difficulties is due to the great sensitivity of the system (hyperbolic systems with unbounded control actions), with respect to perturbations caused either by uncertainty of the model or by the errors introduced in implementing numerical algorithms. Thus, special care must be taken in the choice of the appropriate numerical schemes which eventually lead to implementable finite dimensional solutions. Finite dimensional algorithms are constructed on a basis of a priority analysis of the properties of the original, continuous (infinite diversional) systems with the following criteria in mind: (1) convergence and stability of the algorithms and (2) robustness (reasonable insensitivity with respect to the unknown parameters of the systems). Examples with mixed finite element methods and spectral methods are provided.

Lasiecka, Irena↗
NASA NTRS2001

Predicting Turbulent Convective Heat Transfer in Fully Developed Duct Flows

The performance of an explicit algebraic stress model (EASM) is assessed in predicting the turbulent flow and forced heat transfer in both straight and wavy ducts, with rectangular, trapezoidal and triangular cross-sections, under fully developed conditions. A comparison of secondary flow patterns. including velocity vectors and velocity and temperature contours, are shown in order to study the effect of waviness on flow dynamics, and comparisons between the hydraulic parameters. Fanning friction factor and Nusselt number, are also presented. In all cases. isothermal conditions are imposed on the duct walls, and the turbulent heat fluxes are modeled using gradient-diffusion type models. The formulation is valid for Reynolds numbers up to 10(exp 5) and this minimizes the need for wall functions that have been used with mixed success in previous studies of complex duct flows. In addition, the present formulation imposes minimal demand on the number of grid points without any convergence or stability problems. Criteria in terms of heat transfer and friction factor needed to choose the optimal wavy duct cross-section for industrial applications among the ones considered are discussed.

Rokni, Masoud↗
NASA NTRS1984

Application of improved numerical schemes

Two approaches which accelerate the solution of the steady state Navier-Stokes equations are discussed. The SIMPLER algorithm, a revised version of SIMPLE, provides a more accurate pressure field for each iteration through the momentum equations, thereby speeding convergence. PISO (Pressure Implicit Split Operator), performs a secondary correction of the velocity and pressure fields (after the typical pressure correction) which enhances convergence. Both schemes account for terms neglected in the SIMPLE approach, but do so in slightly different ways. Two dimensional driven cavity flow and flow over a step were calculated to examine the effect of geometry on the performance of these schemes. Computations were carried out on a series of progressively finer grids. The effect of relaxation number on convergence rate was analyzed, using results from SIMPLE as criteria for performance correlation. Results show: (1) the improved schemes promoted convergence by up to 60% for the driven cavity and 40% for flow over a step; (2) for the driven cavity problem, the efficiency of PISO and SIMPLER increased as the number of nodes increased; and (3) to ensure faster convergence, higher relaxation numbers must be applied.

Neely, G. H.↗
NASA NTRS1979

Interfacial geometry and D-variation effects in two-phase systems

Numerical solutions of the governing diffusion equation for two-phase concentration dependent diffusion coefficients are examined. Solutions were also calculated for planar, cylindrical, and spherical geometries to compare the effect of interface geometries with those caused by concentration-dependent diffusion coefficients, and two methods of averaging D were considered to determine the best averaging method for different types of D-variations. The effects of interface-location criteria on mass conservation and convergence of interface location, diffusion coefficient variation in the alpha and beta-phases of a two-phase binary alloy system, effect of D(alpha) variation in a cylindrical couple on beta-phase thickness, and geometry and D-variation effects on the degree of homogenization were determined. It is concluded that typical D(alpha)-variations can have a greater influence on the kinetics of interdiffusion than the geometry.

Tenney, D. R.↗
NASA NTRS2014

Optimization of a Lunar Pallet Lander Reinforcement Structure Using a Genetic Algorithm

This paper presents a design automation process using optimization via a genetic algorithm to design the conceptual structure of a Lunar Pallet Lander. The goal is to determine a design that will have the primary natural frequencies at or above a target value as well as minimize the total mass. Several iterations of the process are presented. First, a concept optimization is performed to determine what class of structure would produce suitable candidate designs. From this a stiffened sheet metal approach was selected leading to optimization of beam placement through generating a two-dimensional mesh and varying the physical location of reinforcing beams. Finally, the design space is reformulated as a binary problem using 1-dimensional beam elements to truncate the design space to allow faster convergence and additional mechanical failure criteria to be included in the optimization responses. Results are presented for each design space configuration. The final flight design was derived from these results.

Burt, Adam O.↗
NASA NTRS1983

Structural optimization - Challenges and opportunities

A review of developments in structural optimization techniques and their interface with growing computer capabilities is presented. Structural design steps comprise functional definition of an object, an evaluation phase wherein external influences are quantified, selection of the design concept, material, object geometry, and the internal layout, and quantification of the physical characteristics. Optimization of a fully stressed design is facilitated by use of nonlinear mathematical programming which permits automated definition of the physics of a problem. Design iterations terminate when convergence is acquired between mathematical and physical criteria. A constrained minimum algorithm has been formulated using an Augmented Lagrangian approach and a generalized reduced gradient to obtain fast convergence. Various approximation techniques are mentioned. The synergistic application of all the methods surveyed requires multidisciplinary teamwork during a design effort.

Sobieszczanski-Sobieski, J.↗
NASA NTRS1984

Structural optimization: Challenges and opportunities

A review of developments in structural optimization techniques and their interface with growing computer capabilities is presented. Structural design steps comprise functional definition of an object, an evaluation phase wherein external influences are quantified, selection of the design concept, material, object geometry, and the internal layout, and quantification of the physical characteristics. Optimization of a fully stressed design is facilitated by use of nonlinear mathematical programming which permits automated definition of the physics of a problem. Design iterations terminate when convergence is acquired between mathematical and physical criteria. A constrained minimum algorithm has been formulated using an Augmented Lagrangian approach and a generalized reduced gradient to obtain fast convergence. Various approximation techniques are mentioned. The synergistic application of all the methods surveyed requires multidisciplinary teamwork during a design effort.

Sobieszczanski-Sobieski, J.↗
NASA NTRS1997

The Effect of Clouds on Water Vapor Profiling from the Millimeter-Wave Radiometric Measurements

Simultaneous measurements with the millimeter-wave imaging radiometer (MIR), cloud lidar system (CLS), and the MODIS airborne simulator (MAS) were made aboard the NASA ER-2 aircraft over the western Pacific Ocean on 17-18 January 1993. These measurements were used to study the effects of clouds on water vapor profile retrievals based on millimeter-wave radiometer measurements. The CLS backscatter measurements (at 0.532 and 1.064 am) provided information on the heights and a detailed structure of cloud layers; the types of clouds could be positively identified. All 12 MAS channels (0.6-13 Am) essentially respond to all types of clouds, while the six MIR channels (89-220 GHz) show little sensitivity to cirrus clouds. The radiances from the 12-/Am and 0.875-gm channels of the MAS and the 89-GHz channel of the MIR were used to gauge the performance of the retrieval of water vapor profiles from the MIR observations under cloudy conditions. It was found that, for cirrus and absorptive (liquid) clouds, better than 80% of the retrieval was convergent when one of the three criteria was satisfied; that is, the radiance at 0.875 Am is less than 100 W/cm.sr, or the brightness at 12 Am is greater than 260 K, or brightness at 89 GHz is less than 270 K (equivalent to cloud liquid water of less than 0.04 g/cm). The range of these radiances for convergent retrieval increases markedly when the condition for convergent retrieval was somewhat relaxed. The algorithm of water vapor profiling from the MIR measurements could not perform adequately over the areas of storm-related clouds that scatter radiation at millimeter wavelengths.

Wang, J. R.↗
DOE OSTI2023

FINITE ELEMENT MODEL MESH REFINEMENT EFFECTS ON QUALIFICATION OF NUCLEAR GRADE GRAPHITE COMPONENTS

The American Society of Mechanical Engineers (ASME) provides the full and simplified design-by-analysis probabilistic assessments for determining acceptance of nuclear grade graphite core components. The assessments can be characterized by three parts: (1) a component stress distribution, often determined by a finite element (FE) model; (2) a Weibull probability density function (pdf) that characterizes the experimental tensile strength distribution; and (3) the post-processor, which combines the FE model and the Weibull strength distribution in accordance with the full and simplified assessments to determine component acceptance. It is known that the level of mesh refinement in FE models can affect the modeled component’s calculated stress distribution. Depending on the component geometry, the stress distribution may converge with sufficient refinement. It was previously unknown whether the acceptance decision resulting from the full and simplified assessments might change even with sufficient mesh refinement. This study explores that question using experimental strength results for a dog-bone geometry for two graphite grades, IG-110 and PCEA. The simplified assessment has two criteria that must be met, the first limits the combined membrane stress by the allowable stress and the second limits the peak equivalent stress by the allowable stress scaled by the ratio of flexural to tensile strength. In the application of the simplified assessment, convergence of the peak equivalent stress required extreme mesh refinement, however, the acceptance decision was not affected. It is hypothesized that more complex geometries with stress concentrations may present mesh refinement effects on the simplified assessment acceptance decision. Mesh refinement did affect the acceptance decision in the full assessment for the applied pressure loadings in this study. This work suggests component stress distribution convergence is not a sufficient criteria for POF convergence in the full assessment and that mesh refinement should continue until the POF has converged, especially where the resulting POF is bordering the SRC acceptable POF limit.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
NASA NTRS1989

An adaptive guidance law for single stage to low earth orbit

An adaptive guidance algorithm based on a cubic spline representation of the ascent profile and imposition of a dynamic pressure constraint is studied for a single stage to low earth orbit. The flight path is divided into initial and terminal phases. In the initial phase, fully adaptive, and in the terminal phase, semi-adaptive, guidance schemes are used. The cubic spline paqrameters are determined by gradient optimization for maximum payload to orbit. In the terminal phase, a linear quadratic regulator is used to derive the optimal feedback gains to keep the vehicle close to the nominal path. The computational aspects of the guidance algorithm are examined and criteria are developed to ensure stability and convergence.

Pamadi, Bandu N.↗
DOE OSTI2026

Using Filter Methods to Guide Convergence for ADMM, with Applications to Nonnegative Matrix Factorization Problems

Nonconvex, nonlinear optimization problems arise naturally in parameter fitting and machine learning. While augmented Lagrangian methods have demonstrated robust convergence for classes of these problems, their convergence for block updates has been relatively unexplored outside of the context of the alternating direction method of multipliers (ADMM). ADMM has seen extensive use in these applications, but may exhibit uncertain convergence behavior in many practical nonconvex settings, and struggles with general nonlinear constraints. In contrast, filter methods have proved effective in enforcing convergence for sequential quadratic programming methods and interior point methods with feasibility criteria. We develop an ADMM-filter method for highly nonlinear and nonconvex problems. Here, we show convergence under mild assumptions for several types of coordinate descent schemes, and demonstrate our algorithm on nonnegative matrix factorization and completion problems in imaging and chemical spectrum analysis.

Nonconvex optimization↗
NASA NTRS1988

Eddies, streams, and convergence zones in turbulent flows

Recent studies of turbulent shear flows have shown that many of their important kinematical and dynamical properties can be more clearly understood by describing the flows in terms of individual events or streamline patterns. These events or flow regions are studied because they are associated with relatively large contributions to certain average properties of the flow, for example kinetic energy, Reynolds stress, or to particular processes in the flow, such as mixing and chemical reactions, which may be concentrated at locations where streamlines converge for fast chemical reactions (referred to as convergence or C regions), or in recirculating eddying regions for slow chemical reactions. The aim of this project was to use the numerical simulations to develop suitable criteria for defining these eddying or vortical zones. The C and streaming (S) zones were defined in order to define the whole flow field. It is concluded that homogeneous and sheared turbulent flow fields are made up of characteristic flow zones: eddy, C, and S zones. A set of objective criteria were found which describe regions in which the streamlines circulate, converge or diverge, and form high streams of high velocity flow.

Hunt, J. C. R.↗
DOE OSTI2022

Multistage mixed precision iterative refinement

Abstract Low precision arithmetic, in particular half precision (16‐bit) floating point arithmetic, is now available in commercial hardware. Using lower precision can offer significant savings in computation and communication costs with proportional savings in energy. Motivated by this, there has been a renewed interest in mixed precision iterative refinement schemes for solving linear systems , and new variants of GMRES‐based iterative refinement have been developed. Each particular variant with a given combination of precisions leads to different condition number‐based constraints for convergence of the backward and forward errors, and each has different performance costs. The constraints for convergence given in the literature are, as an artifact of the analyses, often overly strict in practice, and thus could lead a user to select a more expensive variant when a less expensive one would have sufficed. In this work, we develop a multistage mixed precision iterative refinement solver which aims to combine existing mixed precision approaches to balance performance and accuracy and improve usability. For a user‐specified initial combination of precisions, the algorithm begins with the least expensive approach and convergence is monitored via inexpensive computations with quantities produced during the iteration. If slow convergence or divergence is detected using particular stopping criteria, the algorithm switches to use a more expensive, but more reliable variant. A novel aspect of our approach is that, unlike existing implementations, our algorithm first attempts to use “stronger” GMRES‐based solvers for the solution update before resorting to increasing the precision(s). In some scenarios, this can avoid the need to refactorize the matrix in higher precision. We perform extensive numerical experiments on a variety of random dense problems and problems from real applications which confirm the benefits of the multistage approach.

Oktay, Eda↗
DOE OSTI2022

Numerical Solution of the Steady-State Network Flow Equations for a Non-Ideal Gas

Herein we formulate a steady-state network flow problem for non-ideal gas that relates injection rates and nodal pressures in the network to flows in pipes. For this problem, we present and prove a theorem on uniqueness of generalized solution for a broad class of non-ideal pressure-density relations that satisfy a monotonicity property. Further, we develop a Newton-Raphson algorithm for numerical solution of the steady-state problem, which is made possible by a systematic non-dimensionalization of the equations. The developed algorithm has been extensively tested on benchmark instances and shown to converge robustly to a generalized solution. Previous results [1]-[4], indicate that the steady-state network flow equations for an ideal gas are difficult to solve by the Newton-Raphson method because of its extreme sensitivity to the initial guess. In contrast, we find that non-dimensionalization of the steady-state problem is key to robust convergence of the Newton-Raphson method. We identify criteria based on the uniqueness of solutions under which the existence of a non-physical generalized solution found by a non-linear solver implies non-existence of a physical solution, i.e., infeasibility of the problem. Finally, we compare pressure and flow solutions based on ideal and non-ideal equations of state to demonstrate the need to apply the latter in practice. The solver developed in this article is open-source and is made available for both the academic and research communities as well as the industry.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1981

A variational approach to the question of temporal stability of equilibrium models of solar prominences. I - The formal theory

A Lagrangian approach to the equations describing small amplitude departures from equilibrium of solar prominences is presented. Seven quantities are derived which, by analogy to the concepts of energy, momentum, and angular momentum, are conserved under conditions corresponding to ignorable coordinates of classical mechanics. It is expected that these conservation laws will be useful as criteria of accuracy in obtaining eigenfrequencies for the perturbation equations when numerical techniques are used. The method is capable of handling complicated spatial dependences and has internal criteria for assessing the accuracy of any given numerical convergence scheme.

Lerche, I.↗
NASA NTRS1988

A mathematical theory of learning control for linear discrete multivariable systems

When tracking control systems are used in repetitive operations such as robots in various manufacturing processes, the controller will make the same errors repeatedly. Here consideration is given to learning controllers that look at the tracking errors in each repetition of the process and adjust the control to decrease these errors in the next repetition. A general formalism is developed for learning control of discrete-time (time-varying or time-invariant) linear multivariable systems. Methods of specifying a desired trajectory (such that the trajectory can actually be performed by the discrete system) are discussed, and learning controllers are developed. Stability criteria are obtained which are relatively easy to use to insure convergence of the learning process, and proper gain settings are discussed in light of measurement noise and system uncertainties.

Phan, Minh↗
NASA NTRS1990

Thickness and orientational design for a maximum stiff membrane

Recent results from sensitivity analysis for strain energy with anisotropic elasticity are applied to thickness and orientational design of laminated membranes. Primarily, the first order gradients of the total elastic energy are used in an optimality criteria based method. This traditional method is shown to give slow convergence with respect to design parameters, although the convergence of strain energy is very good. To get a deeper insight into this rather general characteristic, second order derivatives are included and it is shown how they can be obtained by first order sensitivity analysis. Examples of only thickness design, only orientational design, and combined thickness--orientational design are presented.

Pedersen, Pauli↗