Search results for "Numerical stability"
showing 10 items of 29 documents
Numerically solving the relativistic Grad–Shafranov equation in Kerr spacetimes: numerical techniques
2018
The study of the electrodynamics of static, axisymmetric and force-free Kerr magnetospheres relies vastly on solutions of the so called relativistic Grad-Shafranov equation (GSE). Different numerical approaches to the solution of the GSE have been introduced in the literature, but none of them has been fully assessed from the numerical point of view in terms of efficiency and quality of the solutions found. We present a generalization of these algorithms and give detailed background on the algorithmic implementation. We assess the numerical stability of the implemented algorithms and quantify the convergence of the presented methodology for the most established setups (split-monopole, parab…
Numerical model of macro-segregation during directional crystallization process
1998
Abstract In the paper the mathematical model of macro-segregation proceeding during the directional crystallization process is presented. The boundary-initial problem considered is discussed. Next the numerical approximation constructed on the basis of the boundary element method supplemented by a procedure called the artificial heat source method is described. The boundary condition on the solidification front resulting from the alloy component balance is introduced, while in finally the practical aspects of computations concerning the course of the process are discussed.
Entropy dissipation of moving mesh adaptation
2014
Non-uniform grids and mesh adaptation have become an important part of numerical approximations of differential equations over the past decades. It has been experimentally noted that mesh adaptation leads not only to locally improved solution but also to numerical stability of the underlying method. In this paper we consider nonlinear conservation laws and provide a method to perform the analysis of the moving mesh adaptation method, including both the mesh reconstruction and evolution of the solution. We moreover employ this method to extract sufficient conditions — on the adaptation of the mesh — that stabilize a numerical scheme in the sense of the entropy dissipation.
Mirror and triplet displacement energies within nuclear DFT: : numerical stability
2017
Isospin-symmetry-violating class II and III contact terms are introduced into the Skyrme energy density functional to account for charge dependence of the strong nuclear interaction. The two new coupling constants are adjusted to available experimental data on triplet and mirror displacement energies, respectively. We present preliminary results of the fit, focusing on its numerical stability with respect to the basis size.
Numerically stable computation of step-sizes for descent methods. The nonconvex case
1977
The computation of step-sizes which guarantee convergence in unconstrained minimization by descent methods is considered. The use of a “control” or “range” function is highly attractive for this purpose because of its simplicity. Since the Armijo-Goldstein test may fail prematurely due to numerical instability near the minimizer, we consider a range function based on gradient values alone as has been done forg convex in [8]. Numerical algorithms are given for the computation of step-sizes whose behaviour under roundoff is shown to be benign in the sense of F. L. Bauer [5].
DLPNO-MP2 second derivatives for the computation of polarizabilities and NMR shieldings
2021
We present a derivation and efficient implementation of the formally complete analytic second derivatives for the domain-based local pair natural orbital second order Møller–Plesset perturbation theory (MP2) method, applicable to electric or magnetic field-response properties but not yet to harmonic frequencies. We also discuss the occurrence and avoidance of numerical instability issues related to singular linear equation systems and near linear dependences in the projected atomic orbital domains. A series of benchmark calculations on medium-sized systems is performed to assess the effect of the local approximation on calculated nuclear magnetic resonance shieldings and the static dipole …
Efficient numerical integration of neutrino oscillations in matter
2016
A special purpose solver, based on the Magnus expansion, well suited for the integration of the linear three neutrino oscillations equations in matter is proposed. The computations are speeded up to two orders of magnitude with respect to a general numerical integrator, a fact that could smooth the way for massive numerical integration concomitant with experimental data analyses. Detailed illustrations about numerical procedure and computer time costs are provided.
Discrete-ring vortex solitons
2010
We study analytically and numerically the existence and stability of discrete vortex solitons in the circular arrays of nonlinear optical waveguides, governed by the discrete nonlinear Schrodinger equation. Stable vortex breathers with periodically oscillating topological charge are identified and a continuous interpolating map is constructed which allows to recover trajectories of individual phase dislocations in the form of hyperbolic avoided crossings.
A model study of Hartree-Fock and Linear Response in coordinate space
1979
A fast procedure for spherical Hartree-Fock is obtained by coordinate space representation and a modification of gradient iteration. Along similar lines, the corresponding Linear Response equations are derived and solved, in order to achieve a fully consistent treatment. The Linear Response equations are applied to a change in particle numbers, i.e. to the description of isotopic differences. In a model study we look for their physical and numerical properties, i.e. linearity of the response, numerical stability and consistency requirements for the Hartree-Fock basis.
Linear response strength functions with iterative Arnoldi diagonalization
2009
We report on an implementation of a new method to calculate RPA strength functions with iterative non-hermitian Arnoldi diagonalization method, which does not explicitly calculate and store the RPA matrix. We discuss the treatment of spurious modes, numerical stability, and how the method scales as the used model space is enlarged. We perform the particle-hole RPA benchmark calculations for double magic nucleus 132Sn and compare the resulting electromagnetic strength functions against those obtained within the standard RPA.