Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “numerical 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.

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

At least 1,639 records · Page 91

A fast solver for the gyrokinetic field equation with adiabatic electrons

Describing turbulence and microinstabilities in fusion devices is often modelled with the gyrokinetic equation. During the time evolution of the distribution function a field equation for the electrostatic potential needs to be solved. In the case of adiabatic electrons it contains a flux-surface-average term resulting in an integro-differential equation. Its numerical solution is time and memory intensive for three-dimensional configurations. Here a new algorithm is presented which only requires the numerical inversion of a simpler differential operator and a subsequent addition of a correction term. This new procedure is as fast as solving the equation without the surface average.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A study of the efficiency of various Navier-Stokes solvers

A comparative study of the efficiency of some finite difference methods for the solution of the Navier-Stokes equations was conducted. The study was restricted to the two-dimensional steady, uniform property vorticity-stream function equations. The comparisons were drawn by recording the CPU time required to obtain a solution as well as the accuracy of this solution using five numerical methods: central differences, first order upwind differences, second order upwind differences, exponential differences, and an ADI solution of the central difference equations. Solutions were obtained for two test cases: a recirculating eddy inside a square cavity with a moving top, and an impinging jet flow. The results show that whenever the central difference method is stable it generates results with a given accuracy for less CPU time than any other method.

Atias, M.↗

Solution of elliptic partial differential equations by fast Poisson solvers using a local relaxation factor. 2: Two-step method

A two-step semidirect procedure is developed to accelerate the one-step procedure described in NASA TP-2529. For a set of constant coefficient model problems, the acceleration factor increases from 1 to 2 as the one-step procedure convergence rate decreases from + infinity to 0. It is also shown numerically that the two-step procedure can substantially accelerate the convergence of the numerical solution of many partial differential equations (PDE's) with variable coefficients.

Chang, S. C.↗

Implicit Formulation of Muscle Dynamics in OpenSim

Astronauts lose bone and muscle mass during spaceflight. Exercise countermeasure is the primary method for counteracting bone and muscle mass loss in space. New spacecraft exercise device concepts are currently being developed for the NASAs new crew exploration vehicle. The NASA Digital Astronaut Project (DAP) uses computational modeling to help determine if the new exercise devices will be effective as countermeasures. The NASA Digital Astronaut Project is developing the ability to utilize predictive simulation to provide insight into the change in kinematics and kinetics with a change in device and gravitational environment (1-g versus 0-g). For example, in space exercise the subject's body weight is applied in addition to the loads prescribed for musculoskeletal maintenance. How and where these loads are applied obviously directly impacts bone and tissue loads. Additionally, due to space vehicle structural requirements, exercise devices are often placed on vibration isolation systems. This changes the apparent impedance or stiffness of the device as seen by the user. Data collection under these conditions is often impractical and limited. Predictive modeling provides a means to have a virtual subject to test hypotheses. Predictive simulation provides a virtual subject for which we are able to perform studies such as sensitivity to device loading and vibration isolation without the need for laboratory kinematic or kinetic test data.Direct Collocation optimization provides an efficient means to perform task based optimization and predictive modeling. It is relatively straight forward to structure a physical exercise task in a Direct Collocation mathematical formulation: perform a motion such that you start at an initial pose, achieve a given amount of deflection i.e a squat, return to the initial pose, and minimize muscle activation cost. Direct Collocation is advantageous in that it does not require numerical integration to evaluate the objective function. Instead, the system dynamics are transformed to discrete time and the optimizer is constrained such that the solution is not considered to be a valid unless the dynamic equations are satisfied at all time points. The simulation and optimization are effectively done simultaneously. Due to the implicit integration, time steps can be more coarse than in a differential equation solver. In a gait scenario this means that that the model constraints and cost function are evaluated at 100 nodes in the gait cycle versus 10,000 integration steps in a variable-step forward dynamic simulation. Furthermore, no time is wasted on accurate simulations of movements that are far from the optimum. Constrained optimization algorithms require a Jacobian matrix that contains the partial derivatives of each of the dynamic constraints with respect to of each of the state and control variables at all time points. This is a large but sparse matrix. An implicit dynamics formulation requires computation of the dynamic residuals f as a function of the states x and their derivatives, and controls u:f(x, dxdt, u) 0If the dynamics of musculoskeletal system are formulated implicitly, the Jacobian elements are often available analytically, eliminating the need for numerical differentiation; this is obviously computationally advantageous. Additionally, implicit formulation of musculoskeletal dynamics do not suffer from singularities from low mass bodies, zero muscle activation, or other stiff system or

physical exercise↗

An integrated boundary approach for colloidal suspensions simulated using smoothed dissipative particle dynamics

We report that in particle-based continuum solvers such as smoothed particle hydrodynamics (SPH) and smoothed dissipative particle dynamics (SDPD), one of the most significant challenges is the treatment of solid boundaries like walls and colloidal particles, whose presence leads to a truncation of the integral approximation, and hence, error in the numerical solution. In this work, we describe an integrated boundary framework for modeling colloidal suspensions composed of rigid spherical particles. The integral corresponding to the colloid's contribution is analytically evaluated, giving a simple and computationally inexpensive approach relative to conventional boundary particle techniques. We formulate a thermodynamically-consistent version of this top-down method for mesoscale simulations, in which the fluid exchanges momentum with the suspended particles due to thermal fluctuations, giving a framework for modeling the dynamics of colloids at arbitrary Reynolds and Péclet numbers. The resulting evolution equations are validated for a single colloidal particle in a fluid at constant temperature. This simple approach requires ~ N c (p/m) R$2\atop{c}$ fewer pair force calculations relative to traditional boundary particle strategies, where N c is the number of colloids in the system, R c is the colloid radius, ρ is the colloid mass density, and m is the mass of the SDPD particles. Additionally, the use of integrated boundaries removes the need for rigid body constraint dynamics, giving an elegant and efficient basis for large-scale simulations of colloidal suspensions that is general and does not make any physical assumptions about the flow.

42 ENGINEERING↗

A multidimensional unified gas-kinetic scheme for radiative transfer equations on unstructured mesh

In order to extend the unified gas kinetic scheme (UGKS) to solve radiative transfer equations in a complex geometry, a multidimensional asymptotic preserving implicit method on unstructured mesh is constructed in this paper. With an implicit formulation, the CFL condition for the determination of the time step in UGKS can be much relaxed, and a large time step is used in simulations. Differently from previous direction-by-direction UGKS on orthogonal structured mesh, on unstructured mesh the interface flux transport takes into account multi-dimensional effect, where gradients of radiation intensity and material temperature in both normal and tangential directions of a cell interface are included in the flux evaluation. The multiple scale nature makes the UGKS be able to capture the solutions in both optically thin and thick regions seamlessly. In the optically thick region the condition of cell size being less than photon's mean free path is fully removed, and the UGKS recovers a solver for diffusion equation in such a limit on unstructured mesh. For a distorted quadrilateral mesh, the UGKS goes to a nine-point scheme for the diffusion equation, and it naturally reduces to the standard five-point scheme for a orthogonal quadrilateral mesh. Numerical computations covering a wide range of transport regimes on unstructured and distorted quadrilateral meshes will be presented to validate the current approach. - Highlights: • A unified gas-kinetic scheme (UGKS) on unstructured mesh has been developed in all flow regimes. • The scheme solves diffusion equation in optical thick region with large cell size. • The implicit UGKS removes the CFD condition. The efficiency of the scheme has been much improved.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Application of the Parareal Algorithm for Acoustic Wave Propagation

We present an application of the parareal algorithm to solve wave propagation problems in the time domain. The parareal algorithm is based on a decomposition of the integration time interval in time slices. It involves a serial prediction step based on a coarse approximation, and a correction step (computed in parallel) based on a fine approximation within each time slice. In our case, the spatial discretization is based on a spectral element approximation which allows flexible and accurate wave simulations in complex geological media. Fully explicit time advancing schemes are classically used for both coarse and fine solvers.In a first stage, we solve the 1D acoustic wave equation in an homogeneous medium in order to test stability and convergence properties of the parareal algorithm. We confirmed the stability problems outlined by Bal and Farhat et al. for hyperbolic problems. These stability issues are mitigated by a time-discontinuous Galerkin discretization of the coarse solver. It may also involve a coarser spatial discretization (hp-refinement) which helps to preserve stability and allows more significant computer savings. Besides, we explore the contribution of elastodynamic homogenization to build consistent coarse grid solvers. Extension to 2D/3D realistic geological media is an ongoing work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Analysis of hydrogen plasma in a microwave plasma chemical vapor deposition reactor

The aim of this work is to build a numerical model of hydrogen plasma inside a microwave plasma chemical vapor deposition system. This model will help in understanding and optimizing the conditions for the growth of carbon nanostructures. A 2D axisymmetric model of the system is implemented using the finite element high frequency Maxwell solver and the heat transfer solver in COMSOL Multiphysics. The system is modeled to study variation in parameters with reactor geometry, microwave power, and gas pressure. The results are compared with experimental measurements from the Q-branch of the H{sub 2} Fulcher band of hydrogen using an optical emission spectroscopy technique. The parameter γ in Füner's model is calibrated to match experimental observations at a power of 500 W and 30 Torr. Good agreement is found between the modeling and experimental results for a wide range of powers and pressures. The gas temperature exhibits a weak dependence on power and a strong dependence on gas pressure. The inclusion of a vertical dielectric pillar that concentrates the plasma increases the maximum electron temperature by 70%, the maximum gas temperature by 50%, and the maximum electron number density by 70% when compared to conditions without the pillar at 500 W and 30 Torr. Experimental observations also indicate intensified plasma with the inclusion of a pillar.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Linearised Fokker–Planck collision model for gyrokinetic simulations

We introduce a gyrokinetic, linearised Fokker–Planck collision model that satisfies conservation laws and is accurate at arbitrary collisionalities. The differential test-particle component of the operator is exact; the integral field-particle component is approximated using a spherical harmonic and a modified Laguerre polynomial expansion developed by Hirshman and Sigmar (1976 Phys. Fluids 19 1532). The numerical methods of the implementation in the δf-gyrokinetic code stella (Barnes et al 2019 J. Comput. Phys. 391 365–80) are discussed, and conservation properties of the operator are demonstrated. The collision model is then benchmarked against the collision model of the gyrokinetic solver GS2 in the limiting cases of a reduced test-particle collision operator and energy- and momentum-conserving operator. The accuracy of the full collision model is investigated by solving the parallel Spitzer-Härm problem for the transport coefficients. It is shown that retaining collisional energy flux and higher-order terms in the field-particle operator reduces errors in the transport coefficients from 10%–25% for a simple momentum- and energy-conserving model to under 1%.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

On-the-fly response function generation method for composite coarse mesh

The hybrid stochastic deterministic transport code COMET, based on the incident response expansion theory, is used to model reactor cores with high fidelity and formidable computational speed. COMET models a reactor core using a library of incident flux response expansion coefficients that are pre computed for all the unique lattice cells (e.g., fuel assemblies, reflector blocks, etc.) in the core. In order to further improve its computational efficiency in pre-calculating the response library a new response function generation method is developed to compute the response functions for the composite coarse meshes made of a smaller set of unique lattices on the fly within the COMET's deterministic transport core sweep. The efficiency is achieved by eliminating a number of unique lattices that can be made up from the reduced set of unique meshes on the fly. The numerical process consists of the following steps. First, the boundary condition on composite coarse mesh boundaries is projected onto the expansion basis to compute the incident flux moments on external surfaces of all the basic (reduced set of unique) coarse meshes. Secondly, the deterministic sweeping solver in COMET is used to converge on the outgoing/incoming flux expansion moments crossing interfaces between the basic coarse meshes. Thirdly, the response functions for the composite coarse meshes are constructed as a superposition on the fly. The new response function generation method was tested on 88 composite coarse meshes consisting of CANDU fuel bundles and moderator blocks. It was found that response functions generated by the new method agree very well with those generated by direct Monte Carlo calculations. The average and maximum relative differences in the surface-to-surface response coefficients computed by the two methods are 0.10% and 0.20%, respectively. Similarly, the average and maximum relative differences in the response fission densities are 0.13% and 0.43%, respectively. These discrepancies are within one standard deviation of the stochastic uncertainties. The new method is five times faster than the original direct Monte Carlo method. The size of the response function library for the new method is five times smaller than that for the original method, leading to significantly less requirement for the computer hard drive space and memory. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Solution method for a hovering helicopter rotor using the Euler equations

A method for the calculation of the flow field of a helicopter rotor blade in hover is presented. The approach uses a finite volume solution of the three dimensional Euler equations for the blade near field. In the Euler solver the velocity field is decomposed into two parts. One is the induced velocity of the vortex wake extending below the blade, found from a free wake calculation procedure. The other part is the unknown additional velocity field of the rotor blade. This approach eliminates numerical diffusion of the rolled up wake vorticity due to truncation error and artificial viscosity. Also, the effects of the far wake are included in the limited computational domain. Solutions are presented for an isolated wing and a model helicopter rotor and compared to experiment.

Roberts, T. W.↗

Prediction of blade-vortex interaction noise using airloads generated by a finite-difference technique

The present numerical finite-difference scheme for helicopter blade-load prediction during realistic, self-generated three-dimensional blade-vortex interactions (BVI) derives the velocity field through a nonlinear superposition of the rotor flow-field yielded by the full potential rotor flow solver RFS2 for BVI, on the one hand, over the rotational vortex flow field computed with the Biot-Savart law. Despite the accurate prediction of the acoustic waveforms, peak amplitudes are found to have been persistently underpredicted. The inclusion of BVI noise source in the acoustic analysis significantly improved the perceived noise level-corrected tone prediction.

Tadghighi, Hormoz↗

Navier-Stokes computation of compressible turbulent flows with a second order closure, part 1

A second order closure turbulence model for compressible flows is developed and implemented in a 2D Reynolds-averaged Navier-Stokes solver. From the beginning where a kappa-epsilon turbulence model was implemented in the bidiagonal implicit method of MACCORMACK (referred to as the MAC3 code) to the final stage of implementing a full second order closure in the efficient line Gauss-Seidel algorithm, numerous work was done, individually and collectively. Besides the collaboration itself, the final product of this work is a second order closure derived from the Launder, Reece, and Rodi model to account for near wall effects, which has been called FRAME model, which stands for FRench-AMerican-Effort. During the reporting period, two different problems were worked out. The first was to provide Ames researchers with a reliable compressible boundary layer code including a wide collection of turbulence models for quick testing of new terms, both in two equations and in second order closure (LRR and FRAME). The second topic was to complete the implementation of the FRAME model in the MAC5 code. The work related to these two different contributions is reported. dilatation in presence of stron shocks. This work, which has been conducted during a work at the Center for Turbulence Research with Zeman aimed also to cros-check earlier assumptions by Rubesin and Vandromme.

Haminh, Hieu↗

Implementation of the Jacobian-free Newton-Krylov method for solving the for solving the first-order ice sheet momentum balance

We have implemented the Jacobian-free Newton-Krylov (JFNK) method for solving the rst-order ice sheet momentum equation in order to improve the numerical performance of the Community Ice Sheet Model (CISM), the land ice component of the Community Earth System Model (CESM). Our JFNK implementation is based on signicant re-use of existing code. For example, our physics-based preconditioner uses the original Picard linear solver in CISM. For several test cases spanning a range of geometries and boundary conditions, our JFNK implementation is 1.84-3.62 times more efficient than the standard Picard solver in CISM. Importantly, this computational gain of JFNK over the Picard solver increases when rening the grid. Global convergence of the JFNK solver has been signicantly improved by rescaling the equation for the basal boundary condition and through the use of an inexact Newton method. While a diverse set of test cases show that our JFNK implementation is usually robust, for some problems it may fail to converge with increasing resolution (as does the Picard solver). Globalization through parameter continuation did not remedy this problem and future work to improve robustness will explore a combination of Picard and JFNK and the use of homotopy methods.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Numerical Evidence of Multiple Solutions for the Reynolds-Averaged Navier–Stokes Equations

In this paper, evidence is presented for the existence of multiple solutions of Reynolds-averaged Navier–Stokes equations with the one-equation Spalart–Allmaras and two-equation Wilcox $k-ω$ turbulence models on fixed grids in three dimensions and how they were obtained is described. The two major configurations considered are an “academic” extruded two-dimensional airfoil geometry and the trap wing test case. The observed appearance of the multiple solutions seems to be closely related to smooth body separation (sometimes massive) routinely observed in flows over high-lift configurations, especially near stall angles of attack. The results are obtained and cross-verified with two stabilized finite-element codes (streamwise upwind Petrov–Galerkin), which provide residual converged results for complex flows with second-order discretizations. In the paper, the ways multiple solutions have been obtained are described, including such obvious ones as providing a different initial guess to the steady-state solver as well as somewhat unexpected (in this context) techniques of using implicit residual smoothing while time marching to steady state. The phenomenon of the so-called pseudosolutions is also discussed, which is loosely defined as solutions to the discrete system of equations having “sharp” convergence behavior (sometimes up to six to seven orders of the relative residual reduction), which fail to achieve the stronger, machine-zero convergence criterion. Some numerical observations are also presented on the sensitivity of the obtained multiple solutions to the discretization and grid perturbations.

Engineering↗

Hybrid programming-model strategies for GPU offloading of electronic structure calculation kernels

To address the challenge of performance portability and facilitate the implementation of electronic structure solvers, we developed the basic matrix library (BML) and Parallel, Rapid O(N), and Graph-based Recursive Electronic Structure Solver (PROGRESS) library. The BML implements linear algebra operations necessary for electronic structure kernels using a unified user interface for various matrix formats (dense and sparse) and architectures (CPUs and GPUs). Focusing on density functional theory and tight-binding models, PROGRESS implements several solvers for computing the single-particle density matrix and relies on BML. In this paper, we describe the general strategies used for these implementations on various computer architectures, using OpenMP target functionalities on GPUs, in conjunction with third-party libraries to handle performance critical numerical kernels. In this study, we demonstrate the portability of this approach and its performance in benchmark problems.

36 MATERIALS SCIENCE↗

A Low-Rank Solver for the Navier--Stokes Equations with Uncertain Viscosity

In this work, we study an iterative low-rank approximation method for the solution of the steady-state stochastic Navier--Stokes equations with uncertain viscosity. The method is based on linearization schemes using Picard and Newton iterations and stochastic finite element discretizations of the linearized problems. For computing the low-rank approximate solution, we adapt the nonlinear iterations to an inexact and low-rank variant, where the solution of the linear system at each nonlinear step is approximated by a quantity of low rank. This is achieved by using a tensor variant of the GMRES method as a solver for the linear systems. We explore the inexact low-rank nonlinear iteration with a set of benchmark problems, using a model of flow over an obstacle, under various configurations characterizing the statistical features of the uncertain viscosity, and we demonstrate its effectiveness by extensive numerical experiments.

97 MATHEMATICS AND COMPUTING↗

Development of a fractional-step method for the unsteady incompressible Navier-Stokes equations in generalized coordinate systems

A fractional step method is developed for solving the time-dependent three-dimensional incompressible Navier-Stokes equations in generalized coordinate systems. The primitive variable formulation uses the pressure, defined at the center of the computational cell, and the volume fluxes across the faces of the cells as the dependent variables, instead of the Cartesian components of the velocity. This choice is equivalent to using the contravariant velocity components in a staggered grid multiplied by the volume of the computational cell. The governing equations are discretized by finite volumes using a staggered mesh system. The solution of the continuity equation is decoupled from the momentum equations by a fractional step method which enforces mass conservation by solving a Poisson equation. This procedure, combined with the consistent approximations of the geometric quantities, is done to satisfy the discretized mass conservation equation to machine accuracy, as well as to gain the favorable convergence properties of the Poisson solver. The momentum equations are solved by an approximate factorization method, and a novel ZEBRA scheme with four-color ordering is devised for the efficient solution of the Poisson equation. Several two- and three-dimensional laminar test cases are computed and compared with other numerical and experimental results to validate the solution method. Good agreement is obtained in all cases.

Rosenfeld, Moshe↗