SEARCH · Search NASA
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.
This viewgraph presentation addresses the problem of efficiently simulating the evolution of a quantum system. The contents include: 1) Quantum Simulation; 2) Extracting Answers from Quantum Simulations; 3) Quantum Fourier Transform; 4) Eigenvalue Estimation; 5) Fermionic Simulations.
Oak Ridge National Laboratory and the University of Michigan are jointly developing the MPACTcode to be the primary neutron transport code for the Virtual Environment for Reactor Applications (VERA). To solve the transport equation, MPACT uses the 2D/1D method, which decomposes the problem into a stack of 2D planes that are then coupled with a 1D axial calculation. MPACT uses the Method of Characteristics for the 2D transport calculations and P3 for the 1D axial calculations, then accelerates the solution using the 3D Coarse mesh Finite Dierence (CMFD) method. Increasing the number of 2D MOC planes will increase the accuracy of the alculation, but will increase the computational burden of the calculations and can cause slow convergence or instability. To prevent these problems while maintaining accuracy, the sub-plane scheme has been implemented in MPACT. This method sub-divides the MOC planes into sub-planes, refining the 1D P3 and 3D CMFD calculations without increasing the number of 2D MOC planes. To test the sub-plane scheme, three of the VERA Progression Problems were selected: Problem 3, a single assembly problem; Problem 4, a 3x3 assembly problem with control rods and pyrex burnable poisons; and Problem 5, a quarter core problem. These three problems demonstrated that the sub-plane scheme can accurately produce intra-plane axial flux profiles that preserve the accuracy of the fine mesh solution. The eigenvalue dierences are negligibly small, and dierences in 3D power distributions are less than 0.1% for realistic axial meshes. Furthermore, the convergence behavior with the sub-plane scheme compares favorably with the conventional 2D/1D method, and the computational expense is decreased for all calculations due to the reduction in expensive MOC calculations.
PyARC was recently developed at Argonne National Laboratory to automate many of the tasks required in the ARC (Argonne Reactor Computation) fast reactor simulation workflow, from input file generation, code execution, data transfer between ARC codes, and output postprocessing. PyARC will likely be the path forward to train new users of the ARC codes with the goal of wide adoption by the national laboratories, academia, and industry. In particular, for the ANL-JAEA collaboration under the Civil Nuclear Working Group (CNWG) project agreement NE-01, PyARC will be used to model the Joyo and EBR-II reactors for comparisons with measured data and calculated results from JAEA (Task 3: Fast Reactor Fuel and Core). As an additional avenue for verification and validation, this report investigates the use of PyARC towards a variety of existing ARC-based reactor models, in order to understand its efficacy in replicating the behavior of base ARC codes and better understand any limitations within modeling realistic fast reactor problems. To this end, PyARC was used to model the Joyo MKI, RBEC Benchmark-M, PRISM Mod-B, and EBR-II Run 138B cores, and its results were compared to those from existing ARC-based models. It was found that for hexagonal-based geometries PyARC was able to replicate the behavior of ARC codes to within 10 pcm for small reactor cores, and ~150pcm difference in eigenvalue for larger cores. These discrepancies are attributed primarily to differences in local mesh refinement options between ARC and PyARC, which currently cannot be resolved with PyARC’s latest version (1.6.0). In some of these cases, PyARC was used to model steady-state problems with initial core compositions originating from a prior REBUS depletion calculation. While PyARC was not designed to support such steady-state calculations, workarounds were applied to replicate the behavior of ARC-based calculations as closely as possible. Thus, these results demonstrate the wide extent to which they can be applied to fast reactor problems while still providing immense benefit to the user in terms of automating and standardizing common routines within the fast reactor analysis workflow. This study concluded that PyARC will be suitable for modeling the steady-state conditions of the EBR-II and Joyo fast reactors as part of the CNWG project agreement.
This Letter introduces a method for determining the energy spectrum of lattice quantum chromodynamics by applying the Lanczos algorithm to the transfer matrix and using a bootstrap generalization of the Cullum-Willoughby method to filter out spurious eigenvalues. Proof-of-principle analyses of the simple harmonic oscillator and the lattice quantum chromodynamics proton mass demonstrate that this method provides faster ground-state convergence than the “effective mass,” which is related to the power-iteration algorithm. Lanczos provides more accurate energy estimates than multistate fits to correlation functions with small imaginary times while achieving comparable statistical precision. Two-sided error bounds are computed for Lanczos results and guarantee that excited-state effects cannot shift Lanczos results far outside their statistical uncertainties.
The laminar flame speed s l is an important reference quantity for characterising and modelling combustion. Experimental measurements of laminar flame speed require the residence time of the fuel/air mixture (τ f ) to be shorter than the autoignition delay time (τ). This presents a considerable challenge for conditions where autoignition occurs rapidly, such as in compression ignition engines. As a result, experimental measurements in typical compression ignition engine conditions do not exist. Simulations of freely propagating premixed flames, where the burning velocity is found as an eigenvalue of the solution, are also not well posed in such conditions, since the mixture ahead of the flame can autoignite, leading to the so called “cold boundary problem”. In this paper, a numerical method for estimating a reference flame speed, s R , is proposed that is valid for laminar flame propagation at autoignitive conditions. Two isomer fuels are considered to test this method: ethanol, which in the considered conditions is a single-stage ignition fuel; and dimethyl ether, which has a temperature-dependent single- or two-stage ignition and a negative temperature coefficient regime for τ. Calculations are performed for the flame position in a one-dimensional computational domain with inflow-outflow boundary conditions, as a function of the inlet velocity U I and for stoichiometric fuel–air premixtures. The response of the flame position, L F , to U I shows distinct stabilisation regimes. For single-stage ignition fuels, at low U I the flame speed exceeds U I and the flame becomes attached to the inlet. Above a critical U I value, the flame detaches from the inlet and L f becomes extremely sensitive to U I until, for sufficiently high U I , the sensitivity decreases and L f corresponds to the location expected from a purely autoignition stabilised flame. The transition from the attached to the autoignition regimes has a corresponding peak dL f /dU I value which is proposed to be a unique reference flame speed s R for single-stage ignition fuels. For two-stage ignition fuels, there is an additional stable regime where a high-temperature flame propagates into a pool of combustion intermediates generated by the first stage of autoignition. This results in two peaks in dL f /dU I and therefore two reference flame speed values. The lower value corresponds to the definition of s R for single-stage ignition fuels, while the higher value exists only for two-stage ignition fuels and corresponds to a high temperature flame propagating into the first stage of autoignition and is denoted s R ' . Finally, a transport budget analysis for low- and high-temperature radical species is also performed, which confirms that the flame structures at U I = s R and U I = s R ' do indeed correspond to premixed flames (deflagrations), as opposed to spontaneous ignition fronts which do not have a unique propagation speed.
The problem of estimating linear functions of a state vector in a multi-input/output system is considered. A simple lower bound on the observer order with arbitrary eigenvalues is established. Algorithms to construct minimal-order stable (or arbitrary dynamics) observers are also outlined.
New direct and implicit algorithms for optical matrix-vector and systolic array processors are considered. Direct rather than indirect algorithms to solve linear systems and implicit rather than explicit solutions to solve second-order partial differential equations are discussed. In many cases, such approaches more properly utilize the advantageous features of optical systolic array processors. The matrix-decomposition operation (rather than solution of the simplified matrix-vector equation that results) is recognized as the computationally burdensome aspect of such problems that should be computed on an optical system. The Householder QR matrix-decomposition algorithm is considered as a specific example of a direct solution. Extensions to eigenvalue computation and formation of matrices of special structure are also noted.
The problem of optimal placement of active members which are used for vibration control in adaptive truss structures is investigated. The control scheme is based on the method of eigenvalue assignment as a means of shaping the transient response of the controlled adaptive structures, and the minimization of required control action is considered as the optimization criterion. To this end, a performance index which measures the control strokes of active members is formulated in an efficient way. In order to reduce the computation burden, particularly for the case where the locations of active members have to be selected from a large set of available sites, several heuristic searching schemes are proposed for obtaining the near-optimal locations. The proposed schemes significantly reduce the computational complexity of placing multiple active members to the order of that when a single active member is placed.
The eigenvalue sensitivity for hydrodynamic stability operators is investigated. Classical matrix perturbation techniques as well as the concept of epsilon-pseudospectra are applied to show that parts of the spectrum are highly sensitive to small perturbations. Applications are drawn from incompressible plane Couette flow, trailing line vortex flow, and compressible Blasius boundary-layer flow. Parameter studies indicate a monotonically increasing effect of the Reynolds number on the sensitivity. The phenomenon of eigenvalue sensitivity is due to the nonnormality of the operators and their discrete matrix analogs and may be associated with large transient growth of the corresponding initial value problem.
Comparisons of responses and first-order sensitivities of responses to the neutron fission spectrum ( χ ) have been done using different transport methods. Kodeli and Slavič compared sensitivities from Monte Carlo, deterministic, and method of characteristics of k eff in the SNEAK-7 benchmark, and the agreement was reasonable. Yamamoto compared deterministic and Monte Carlo sensitivities of k - and α -eigenvalues to the fission χ , but only for material (i.e. single-isotope) values of the fission χ . Haeck et al. compared Monte Carlo and deterministic sensitivities of k eff to the fission χ of individual isotopes for several benchmarks, and the agreement was excellent (except, strangely, for U-236). Kiedrowski and Brown compared Monte Carlo and analytic sensitivities of k ∞ to the fission χ in analytic problems, but only for material values of the fission χ .
The continuum-scale electrokinetic porous-media flow and excess charge redistribution equations are uncoupled using eigenvalue decomposition. The uncoupling results in a pair of independent diffusion equations for “intermediate” potentials subject to modified material properties and boundary conditions. The fluid pressure and electrostatic potential are then found by recombining the solutions to the two intermediate uncoupled problems in a matrix-vector multiplication. Expressions for the material properties or source terms in the intermediate uncoupled problem may require extended precision or careful rewriting to avoid numerical cancellation, but the solutions themselves can typically be computed in double precision. The approach works with analytical or gridded numerical solutions and is illustrated through two examples. The solution for flow to a pumping well is manipulated to predict streaming potential and electroosmosis, and a periodic one-dimensional analytical solution is derived and used to predict electroosmosis and streaming potential in a laboratory flow cell subjected to low frequency alternating current and pressure excitation. The examples illustrate the utility of the eigenvalue decoupling approach, repurposing existing analytical solutions or numerical models and leveraging solutions that are simpler to derive for coupled physics.
Matrix elements of potential energy are examined in detail. We consider a model problem — a particle in a central potential. The most popular forms of central potential are taken up, namely, square-well potential, Gaussian, Yukawa and exponential potentials. We study eigenvalues and eigenfunctions of the potential energy matrix constructed with oscillator functions. It is demonstrated that eigenvalues coincide with the potential energy in coordinate space at some specific discrete points. We establish approximate values for these points. It is also shown that the eigenfunctions of the potential energy matrix are the expansion coefficients of the spherical Bessel functions in a harmonic oscillator basis. We also demonstrate a close relation between the separable approximation and L{sup 2} basis (J-matrix) method for the quantum theory of scattering.
For the past 40 years, it has been known from the theoretical work of Gelbard [1] and Brissenden [2] that the results from Monte Carlo (MC) criticality calculations are biased if the number of neutrons per generation is not sufficiently large. The computed k-effective eigenvalue will be lower than the actual eigenvalue, and the computed eigenfunction will be too low in high-importance regions and too high in low-importance regions. This difficulty is commonly called the undersampling problem for criticality calculations. The biases for undersampling were confirmed for practical, realistic problems with modern MC codes by Brown [3]. The solution to undersampling is simple - use a larger number of neutrons per generation to reduce the biases to levels that are negligible. Over the past few years, the OECD-NEA-WPNCS Expert Group on Advanced Monte Carlo Techniques (EGAMCT) [4,5,6] investigated an extreme form of undersampling called clustering. When the number of neutrons per generation is small, neutrons in successive generations tend to group together (cluster) due to correlations, and the clusters tend to migrate through the problem space. While the problems of undersampling and clustering are well known and well understood theoretically, there were no existing diagnostics for MC codes to determine whether a given calculation is using sufficient neutrons per generation to eliminate these problems. That is, no tools were available to diagnose biases due to undersampling or clustering. The tools described herein resolve this longstanding problem by providing 2 robust methods for detecting undersampling and clustering in MC criticality calculations. The diagnosis of undersampling and clustering is part of a larger effort to thoroughly overhaul the calculational methods for criticality problems, providing for automated sampling of the initial fission distribution, acceleration of the iteration convergence process, automated detection of convergence in the iterations and starting of the active tally cycles, and the diagnosis of undersampling. All of these new methods have been prototyped in a local modified version of mcnp6.2 [7] and tested on a variety of critical systems.
The major goals of the project were to develop and analyze new ensemble simulation techniques, including trajectory stratification and preconditioned MCMC techniques, as well as develop fast numerical linear algebra techniques closely related to ensemble simulation ideas. The trajectory stratification techniques involve simulating in parallel short trajectory fragments of a Markov process confined to a specific region of space‐time and then patching together the statistics gathered to assemble estimates of very general dynamical properties. We have also developed this approach for rare event simulation and extended the techniques to applications requiring a more general framework (such as electronic structure calculations). The preconditioned MCMC techniques involve simulating multiple Markov chains in parallel and then using information from the ensemble to speed the mixing of each individual chain. The fast randomized linear algebra methods are motivated by the diffusion Monte Carlo technique, but are applicable to finding the dominant eigenvalue of (almost) general matrices. For most non‐negative matrices, the schemes result in an error (compared to the power method) that is constant in the dimension of the problem. For more general matrices, we see a very clear sublinear cost trend in computational tests.
The feasibility of using reduced order models and reduced order observers with eigenvalue/eigenvector assignment procedures is investigated. A review of spectral assignment synthesis procedures is presented. Then, a reduced order model which retains essential system characteristics is formulated. A constant state feedback matrix which assigns desired closed loop eigenvalues and approximates specified closed loop eigenvectors is calculated for the reduced order model. It is shown that the eigenvalue and eigenvector assignments made in the reduced order system are retained when the feedback matrix is implemented about the full order system. In addition, those modes and associated eigenvectors which are not included in the reduced order model remain unchanged in the closed loop full order system. The full state feedback design is then implemented by using a reduced order observer. It is shown that the eigenvalue and eigenvector assignments of the closed loop full order system rmain unchanged when a reduced order observer is used. The design procedure is illustrated by an actual design problem.
The feasibility of using reduced order models and reduced order observers with eigenvalue/eigenvector assignment procedures is investigated. A review of spectral assignment synthesis procedures is presented. Then, a reduced order model which retains essential system characteristics is formulated. A constant state feedback matrix which assigns desired closed loop eigenvalues and approximates specified closed loop eigenvectors is calculated for the reduced order model. It is shown that the eigenvalue and eigenvector assignments made in the reduced order system are retained when the feedback matrix is implemented about the full order system. In addition, those modes and associated eigenvectors which are not included in the reduced order model remain unchanged in the closed loop full order system. The fulll state feedback design is then implemented by using a reduced order observer. It is shown that the eigenvalue and eigenvector assignments of the closed loop full order system remain unchanged when a reduced order observer is used. The design procedure is illustrated by an actual design problem.
The conjugate gradient algorithm for solving Hermitian positive definite linear systems is usually combined with preconditioning in order to speed up convergence. In recent years, there has been a revival of polynomial preconditioning, motivated by the attractive features of the method on modern architectures. Standard techniques for choosing the preconditioning polynomial are based only on bounds for the extreme eigenvalues. Here a different approach is proposed, which aims at adapting the preconditioner to the eigenvalue distribution of the coefficient matrix. The technique is based on the observation that good estimates for the eigenvalue distribution can be derived after only a few steps of the Lanczos process. This information is then used to construct a weight function for a suitable Chebyshev approximation problem. The solution of this problem yields the polynomial preconditioner. In particular, we investigate the use of Bernstein-Szego weights.
Abstract An algorithm for solving smooth nonconvex optimization problems is proposed that, in the worst-case, takes ${\mathscr O}(\varepsilon ^{-3/2})$ iterations to drive the norm of the gradient of the objective function below a prescribed positive real number $\varepsilon $ and can take ${\mathscr O}(\varepsilon ^{-3})$ iterations to drive the leftmost eigenvalue of the Hessian of the objective above $-\varepsilon $. The proposed algorithm is a general framework that covers a wide range of techniques including quadratically and cubically regularized Newton methods, such as the Adaptive Regularization using Cubics (arc) method and the recently proposed Trust-Region Algorithm with Contractions and Expansions (trace). The generality of our method is achieved through the introduction of generic conditions that each trial step is required to satisfy, which in particular allows for inexact regularized Newton steps to be used. These conditions center around a new subproblem that can be approximately solved to obtain trial steps that satisfy the conditions. A new instance of the framework, distinct from arc and trace, is described that may be viewed as a hybrid between quadratically and cubically regularized Newton methods. Numerical results demonstrate that our hybrid algorithm outperforms a cubically regularized Newton method.