Long-time behaviour of arbitrary order continuous time Galerkin schemes for some one-dimensional phase transition problems

The long-time behaviour of continuous time Galerkin (CTG) approximationsof some well-known one-dimensional non-linear evolution problemswhich model phase transitions are analyzed. These numericalschemes are fully discrete and of arbitrary order. Partially supported by the Army Research Office through grant28535-MA. Partially supported by the ONR Contract No N00014-90-J-1238.     

A numerical method using the Prfer transformation for the calculation of eigenpairs of two-parameter Sturm-Liouville problems

A shooting method for coupled Prfer equations is discussedfor numerical solution of two-parameter Sturm-Liouville problems. Research supported in part by grants from the NSERC of Canada.     

Finite-difference schemes for scalar conservation laws with source terms

Explicit and semi-implicit finite-difference schemes approximatingnon-homogeneous scalar conservation laws are analyzed. Optimalerror bounds independent of the stiffness of the underlyingequation are presented. This author has been supported by The Norwegian Research Council(NFR), program No 100284/431. e-mail: schroll{at}igpm.rwth-aachen.de This author has been supported by The Norwegian Research Council(NFR), program Nos 100284/431 and STP.29643. e-mail: ragnar{at}ifi.uio.no     

Numerical simulations of measurements of capillary contact angles

In a recent paper, Fischer and Finn have proposed a procedureto improve the accuracy in the measurement of capillary contactangles, based on the use of vessels with canonical cross-sections.We simulate numerically the behaviour of such shapes for a numberof cross-sections and fluid contact angles. Our approximationconsists of the minimization of a suitable convex functionaldiscretized by finite elements. e-mail: bellettini{at}sns.it e-mail: paolini{at}isa.mat.unimi.it     

A finite-difference algorithm for an inverse Sturm-Liouville problem

We study a method for approximating a potential q(x) in y(0)=y()=0 from finite spectral data. When the potential is symmetric,the data are the first M Dirichlet eigenvalues. In the generalcase, the first M terminal velocities are also specified. Acentred finite-difference scheme reduces the inverse Sturm-Liouvilleproblem to a matrix inverse eigenvalue problem. Our approachis motivated by the work of Paine, de Hoog and Anderssen, whoinvestigated the discrepancy between continuous and matrix eigenvaluesunder finite differences. Our modified Newton scheme is basedon choosing the number of interior mesh points in the discretizationto be 2M. The modified Newton scheme is shown to be convergentfor both the case of a symmetric and general potential. Somenumerical experiments are given. Supported in part by Institute for Scientific Computation,Texas A&M University.     

The Method of Fundamental Solutions for the Solution of Nonlinear Plane Potential Problems

The method of fundamental solutions is described for the solutionof elliptic boundary value problems governed by Laplace's equationin the plane subject to nonlinear radiation-type boundary conditions.The effectiveness of the method is demonstrated by examiningits performance on two problems from the literature, and comparisonsare made with published results obtained using boundary elementmethods. Portions of this work were conducted while this author wasa visiting assistant professor at the University of Kentucky. Partially supported by the National Science Foundation undergrant MCS-8303287 and grant RII-8610671, and by the Commonwealthof Kentucky through the Kentucky EPSCoR Program.     

Total Flux Estimates for a Finite-Element Approximation of Elliptic Equations

An elliptic boundary-value problem on a domain with prescribedDirichlet data on _I is approximated using a finite-elementspace of approximation power h^K in the L² norm. It is shownthat the total flux across _I can be approximated with an errorof O(h^K) when is a curved domain in Rⁿ (n = 2 or 3) and isoparametricelements are used. When is a polyhedron, an O(h^2K–2)approximation is given. We use these results to study the finite-elementapproximation of elliptic equations when the prescribed boundarydata on _I is the total flux. Present address: School of Mathematical and Physical Sciences,University of Sussex, Brighton, Sussex BN1 9QH.     

On approximate approximations using Gaussian kernels

This paper discusses quasi-interpolation and interpolation withGaussians. Estimates are obtained showing a high-order approximationup to some saturation error negligible in numerical applications.The construction of local high-order quasi-interpolation formulasis given. Supported in part by the International Centre for MathematicalSciences, Edinburgh.     

Finite difference discretization of the cubic Schrdinger equation

We analyze the discretization of an initial-boundary value problemfor the cubic Schrdinger equation in one space dimension bya Crank-Nicolson-type finite difference scheme. We then linearizethe corresponding equations at each time level by Newton's methodand discuss an iterative modification of the linearized schemewhich requires solving linear systems with the same tridiagonalmatrix. We prove second-order error estimates. Work supported by the Institute of Applied and ComputationalMathemati of the Research Center of Crete-FORTH.     

A new non-conforming Petrov-Galerkin finite-element method with triangular elements for a singularly perturbed advection-diffusion problem

A new non-conforming exponentially fitted Petrov-Galerkin finite-elementmethod based on Delaunay triangulation and its Dirichlet tessellationis constructed for the numerical solution of singularly perturbedstationary advectiondiffusion problems with a singular perturbationparameter . The method is analyzed mathematically and its stabilityis shown to be independent of . The error estimate in an -independentdiscrete energy norm for the approximate solution is shown todepend on first-order seminorms of the flux and the zero-orderterm of the equation, the sup norm of the exact solution, thefirst-order seminorm of the coefficient of the advection term,and the approximation error of the inhomogeneous term. Sincethe first two seminorms are not bounded uniformly in , the -uniformconvergence of the method still remains an open question. Noassumption is required that the angles in the triangulationare all acute. Since the system matrix for this method is identicalto that for the exponentially fitted box method, the theoreticalresults also provide an analysis of that box method. The newmethod also contains the central-difference and upwind methodsas two limiting cases. It can be regarded as a weighted finite-differencemethod on a triangular mesh. Numerical results are presentedto show the superior performance of the method in comparisonwith the upwind and central-difference methods for a small increasein the computation cost. Present address: School of Mathematics, The University of NewSouth Wales, Kensington, NSW 2033, Australia.     

On spurious asymptotic numerical solutions of explicit Runge-Kutta methods

The bifurcation diagram associated with the logistic equationⁿ⁺¹ = aⁿ(1 – ⁿ) is by now well known, as is its equivalenceto solving the ordinary differential equation (ODE) u' = u(1– u) by the explicit Euler difference scheme. It has alsobeen noted by Iserles that other popular difference schemesmay not only exhibit period doubling and chaotic phenomena butalso possess spurious fixed points. We investigate, both analyticallyand computationally, Runge-Kutta schemes applied to the equationu'=f(u), for f(u) = u{1 – u) and f(u) = au(1 – u)(b– u), contrasting their behaviour with the explicit Eulerscheme. We determine and provide a local analysis of bifurcationsto spurious fixed points and periodic orbits. In particularwe show that these may appear below the linearised stabilitylimit of the scheme, and may consequently lead to erroneouscomputational results. Major part of the material was published as an internal report-NASATechnical Memorandum 102919, April 1990, also as Universityof Reading Numerical Analysis Report 3/90, March 1990. This work was performed whilst a visiting scientist at NASAAmes Research Center, Moffett Field. CA 94035 USA. Staff Scientist, Fluid Dynamics Division.     

An analysis of the coupling of finite-element and Nystrm methods in acoustic scattering

In this paper we analyze the use of a combined Nystrm and finite-elementprocedure for approximating the scattered time-harmonic acousticfield produced when an incident wave interacts with a boundedinhomogeneity. The coupling technique uses a smooth artificialboundary and explicitly decouples the integral equation andfinite-element computations. We prove convergence of the method.One highlight of this analysis is that we prove Sobolev spaceerror estimates for the Nystrm scheme (which is usually analyzedin Hlder spaces). We also present some numerical results. E-mail: kirsch{at}am.uni-erlangen.de E-mail: monk{at}math.udel.edu     

Fitted and Unfitted Finite-Element Methods for Elliptic Equations with Smooth Interfaces

This paper considers the finite-element approximation of theelliptic interface problem: -?(u) + cu = f in Rⁿ (n = 2 or3), with u = 0 on , where is discontinuous across a smoothsurface in the interior of . First we show that, if the meshis isoparametrically fitted to using simplicial elements ofdegree k - 1, with k 2, then the standard Galerkin method achievesthe optimal rate of convergence in the H¹ and L² norms overthe approximations _l⁴ of _l where _{l 2}. Second, since itmay be computationally inconvenient to fit the mesh to , weanalyse a fully practical piecewise linear approximation ofa related penalized problem, as introduced by Babuska (1970),based on a mesh that is independent of . We show that, by choosingthe penalty parameter appropriately, this approximation convergesto u at the optimal rate in the H¹ norm over _l⁴ and in the L²norm over any interior domain _l^* satisfying _l^* _l^** _l⁴ for somedomain _l^**. Present address: School of Mathematical and Physical Sciences,University of Sussex, Brighton BN1 9QH     

An Example of Only Linear Convergence of Trust Region Algorithms for Non-smooth Optimization

Most superlinear convergence results about trust region algorithmsfor non-smooth optimization are dependent on the inactivityof trust region restrictions. An example is constructed to showthat it is possible that at every iteration the trust regionbound is active and the rate of convergence is only linear,though strict complementarity and second order sufficiency conditionsare satisfied. Presented at the 1983 Dundee Conference on Numerical Analysis     

Set convergence for discretizations of the attractor

We consider the discretization of a dynamical system given bya C⁰-semigroup S(t), defined on a Banach space X, possessingan attractor . Under certain weak assumptions, Hale, Lin andRaugel showed that discretizations of S(t) possess local attractors,which may be considered as approximations to . Without furtherassumptions, we show that these local attractors possess convergentsubsequences in the Hausdorff or set metric, whose limit isa compact invariant subset of . Using a new construction, wealso consider the Kloeden and Lorenz concept of attracting setsin a Banach space, and show under mild assumptions that discretizationspossess attracting sets converging to in the Hausdorff metric. ath{at}maths.bath.ac.uk Endre.Suli{at}comlab.ox.ac.uk     

Structure of solution manifolds in a strongly coupled elliptic system

In this paper we study the steady-state solutions of a reaction-diffusionmodel, the Selkov scheme for glycolysis, under homogeneous Dirichletboundary conditions. Near to thermodynamic equilibrium, thestructure and stability of solutions are fully described. Abifurcation analysis is carried out, using the size of the regionin which the reaction takes place and one diffusion coefficientas main bifurcation parameters. The analysis helps us to understandthe nature of the bifurcation points, and determines the shapesand stability of the bifurcating manifolds in the neighbourhoodof the constant state. Local convergence of spectral methodsis shown, and some global pictures are calculated using path-followingtechniques. The framework we use can be applied to a wide varietyof reaction-diffusion systems. ^*Permanent address: Dpto. de Matemtica Aplicada, Fac. de CenciasQumicas, Universidad Complutense, 28040-Madrid, Spain. Current address: Department of Mathematics, Paisley Collegeof Technology, High Street, Paisley, Renfrewshire PA1 2BE, UK. Permanent address: Dpto. de Matemtica, Fac. de Matemtca, Astronomiay Fisica, Universidad National de Crdoba, 5000-Crdoba, R.Argentina.     

Convergence of Methods for the Numerical Solution of the Korteweg--de Vries Equation

We prove that a family of methods for the numerical solutionof the Korteweg—de Vries equation is convergent. Thisfamily includes as particular cases some known finite differenceand finite element schemes. It is also found that the stabilityproperties of the methods vary significantly with the treatmentof the non-linear term. On leave from Shanghai University of Science and Technology,Shanghai, China.     

On the Convergence of Some Approximate Methods of Conformal Mapping

A method was given in Ellacott (1978) for determining approximatelythe conformal mapping of a Jordan region on to a disc. Someresults on the convergence of this method are given, which canbe used to prove the result (conjectured in Ellacott, 1978)that if the boundary curve is analytic, then convergence isuniform. The corresponding result is also proved for the Bergmanand Szeg Kernel methods with polynomial basis functions. (Theresult is already known for the Szeg? Kernel method, but a differentproof is given.) Also discussed is the use of rational basisfunctions for the Bergman Kernel method. Currently visiting Forschungsinstitut fr Mathematik, ETH-Zentrum,CH-8092, Zurich.     

Stability of the approximation of a regular solution branch

In this paper we extend the so-called inf-sup conditions tononlinear problems having regular solution branches. The inf-supconditions for nonlinear problems have been introduced, forinstance, by Caloz and Rappaz in 1994. Here we present a moregeneral approach to include turning points on a solution branch.Our abstract results are applied to model examples. E-mail: Gabriel.Caloz{at}univ-rennesl.fr     

An Algorithm with a Point-Linearized Input-Output Mapping for Integrated System Optimization and Parameter Estimation

An improved algorithm for centralized integrated system optimizationand parameter estimation based on a point-linearized I-O mappingand Newton step in updating is described. Two convergence theoremsunder different assumptions are investigated. Optimality ofthe method and the relationship between the algorithmic fixed-pointset and the optimum solution set of the system are presented.Simulation reveals that, because of the adoption of point-linearizedI-O mapping and a Newton step in updating, the proposed techniqueenjoys two benefits of easier computation and a higher convergencerate. On study leave from the Anhui Institute of Machinery & Electronics,Wuhu, Anhui Province, China On study leave from the Institute of Systems Engineering, Xi’anJiaotong University, Xi’an, China