In this article we consider elliptic partial differential equations with random coefficients and/or random forcing terms. In the current treatment of such problems by stochastic Galerkin methods it is standard to assume that the random diffusion coefficient is bounded by positive deterministic constants or modeled as lognormal random field. In contrast, we make the significantly weaker assumption that the non-negative random coefficients can be bounded strictly away from zero and infinity by random variables only and may have distributions different from a lognormal one. We show that in this case the standard stochastic Galerkin approach does not necessarily produce a sequence of approximate solutions that converges in the natural norm to the exact solution even in the case of a lognormal coefficient. By using weighted test function spaces we develop an alternative stochastic Galerkin approach and prove that the associated sequence of approximate solutions converges to the exact solution in the natural norm. Hereby, ideas for the case of lognormal coefficient fields from earlier work of Galvis, Sarkis and Gittelson are used and generalized to the case of positive random coefficient fields with basically arbitrary distributions.

In this work, the error behaviour of high-order exponential operator splitting methods for the time integration of nonlinear evolutionary Schrödinger equations is investigated. The theoretical analysis utilises the framework of abstract evolution equations on Banach spaces and the formal calculus of Lie derivatives. The general approach is substantiated on the basis of a convergence result for exponential operator splitting methods of (nonstiff) order p applied to the multi-configuration time-dependent Hartree–Fock (MCTDHF) equations, which are associated with a model reduction for high-dimensional linear Schrödinger equations describing free electrons that interact by Coulomb force. Provided that the analytical solution of the MCTDHF equations constituting a system of coupled linear ordinary differential equations and low-dimensional nonlinear partial differential equations satisfies suitable regularity requirements, convergence of order p − 1 in the H1 Sobolev norm and convergence of order p in the L2 norm is proven. An analogous result follows for the cubic nonlinear Schrödinger equation, which is also illustrated by a numerical experiment.

Parallel replica dynamics is a method for accelerating the computation of processes characterized by a sequence of infrequent events. In this work, the processes are governed by the overdamped Langevin equation. Such processes spend much of their time about the minima of the underlying potential, occasionally transitioning into different basins of attraction. The essential idea of parallel replica dynamics is that the exit distribution from a given well for a single process can be approximated by the distribution of the first exit of N independent identical processes, each run for only 1 / N-th the amount of time. While promising, this leads to a series of numerical analysis questions about the accuracy of the exit distributions. Building upon the recent work in [C. Le Bris, T. Lelièvre, M. Luskin and D. Perez, Monte Carlo Methods Appl. 18 (2012) 119–146], we prove a unified error estimate on the exit distributions of the algorithm against an unaccelerated process. Furthermore, we study a dephasing mechanism, and prove that it will successfully complete.

We study preconditioning techniques for discontinuous Galerkin discretizations of isotropic linear elasticity problems in primal (displacement) formulation. We propose subspace correction methods based on a splitting of the vector valued piecewise linear discontinuous finite element space, that are optimal with respect to the mesh size and the Lamé parameters. The pure displacement, the mixed and the traction free problems are discussed in detail. We present a convergence analysis of the proposed preconditioners and include numerical examples that validate the theory and assess the performance of the preconditioners.

An extension of the local projection stabilization (LPS) finite element method for convection-diffusion-reaction equations is presented and analyzed, both in the steady-state and the transient setting. In addition to the standard LPS method, a nonlinear crosswind diffusion term is introduced that accounts for the reduction of spurious oscillations. The existence of a solution can be proved and, depending on the choice of the stabilization parameter, also its uniqueness. Error estimates are derived which are supported by numerical studies. These studies demonstrate also the reduction of the spurious oscillations.

Nous étudions l’effet d’une couche mince rugueuse périodique déposée sur une structuresemi-infinie, dans le contexte Helmholtz bi-dimensionnel. Formellement, nous obtenons desconditions de transmission équivalentes à l’ordre 1, par des techniques de typehomogénéisation. Suivent alors la résolution du problème du milieu effectif éclairé parune onde plane, et le calcul de la fonction de Green effective ; le tout par analyse deFourier. Dans un deuxième temps, nous considérons le problème de diffraction par un objetpénétrable enfoui dans la structure recouverte par la couche rugueuse. Nous le résolvonspar la méthode des éléments finis de frontière, dans le milieu effectif. Des résultatsnumériques sont présentés. Enfin, le modèle effectif est validé dans le cas d’une coucheplate, et l’approximation de Born est utilisée pour tester le code des équationsintégrales.

Cultivating oleaginous microalgae in specific culturing devices such as raceways is seenas a future way to produce biofuel. The complexity of this process coupling non linearbiological activity to hydrodynamics makes the optimization problem very delicate. Thelarge amount of parameters to be taken into account paves the way for a usefulmathematical modeling. Due to the heterogeneity of raceways along the depth dimensionregarding temperature, light intensity or nutrients availability, we adopt a multilayerapproach for hydrodynamics and biology. For free surface hydrodynamics, we use amultilayer Saint–Venant model that allows mass exchanges, forced by a simplifiedrepresentation of the paddlewheel. Then, starting from an improved Droop model thatincludes light effect on algae growth, we derive a similar multilayer system for thebiological part. A kinetic interpretation of the whole system results in an efficientnumerical scheme. We show through numerical simulations in two dimensions that ourapproach is capable of discriminating between situations of mixed water or calm andheterogeneous pond. Moreover, we exhibit that a posteriori treatment ofour velocity fields can provide lagrangian trajectories which are of great interest toassess the actual light pattern perceived by the algal cells and therefore understand itsimpact on the photosynthesis process.

We propose an unconditionally stable semi-implicit time discretization of the phase fieldcrystal evolution. It is based on splitting the underlying energy into convex and concaveparts and then performing H-1 gradient descent steps implicitly for the formerand explicitly for the latter. The splitting is effected in such a way that the resultingequations are linear in each time step and allow an extremely simple implementation andefficient solution. We provide the associated stability and error analysis as well asnumerical experiments to validate the method’s efficiency.

In this work we study a fully discrete mixed scheme, based on continuous finite elementsin space and a linear semi-implicit first-order integration in time, approximating anEricksen–Leslie nematic liquid crystal model by means of aGinzburg–Landau penalized problem. Conditional stability of this schemeis proved via a discrete version of the energy law satisfied by thecontinuous problem, and conditional convergence towards generalized Young measure-valuedsolutions to the Ericksen–Leslie problem is showed when the discreteparameters (in time and space) and the penalty parameter go to zero at the same time.Finally, we will show some numerical experiences for a phenomenon of annihilation ofsingularities.

Using Burgers’ equation with mixed Neumann–Dirichlet boundary conditions, we highlight aproblem that can arise in the numerical approximation of nonlinear dynamical systems oncomputers with a finite precision floating point number system. We describe the dynamicalsystem generated by Burgers’ equation with mixed boundary conditions, summarize some ofits properties and analyze the equilibrium states for finite dimensional dynamical systemsthat are generated by numerical approximations of this system. It is important to notethat there are two fundamental differences between Burgers’ equation with mixedNeumann–Dirichlet boundary conditions and Burgers’ equation with both Dirichlet boundaryconditions. First, Burgers’ equation with homogenous mixed boundary conditions on a finiteinterval cannot be linearized by the Cole–Hopf transformation. Thus, on finite intervalsBurgers’ equation with a homogenous Neumann boundary condition is truly nonlinear. Second,the nonlinear term in Burgers’ equation with a homogenous Neumann boundary condition isnot conservative. This structure plays a key role in understanding the complex dynamicsgenerated by Burgers’ equation with a Neumann boundary condition and how this structureimpacts numerical approximations. The key point is that, regardless of the particularnumerical scheme, finite precision arithmetic will always lead to numerically generatedequilibrium states that do not correspond to equilibrium states of the Burgers’ equation.In this paper we establish the existence and stability properties of these numericalstationary solutions and employ a bifurcation analysis to provide a detailed mathematicalexplanation of why numerical schemes fail to capture the correct asymptotic dynamics. Weextend the results in [E. Allen, J.A. Burns, D.S. Gilliam, J. Hill and V.I. Shubov,Math. Comput. Modelling 35 (2002) 1165–1195] and provethat the effect of finite precision arithmetic persists in generating a nonzero numericalfalse solution to the stationary Burgers’ problem. Thus, we show that the results obtainedin [E. Allen, J.A. Burns, D.S. Gilliam, J. Hill and V.I. Shubov, Math. Comput.Modelling 35 (2002) 1165–1195] are not dependent on a specifictime marching scheme, but are generic to all convergent numerical approximations ofBurgers’ equation.

We propose a Lagrangian approach to deriving energy-preserving finite difference schemes for the Euler–Lagrange partial differential equations. Noether’s theorem states that the symmetry of time translation of Lagrangians yields the energy conservation law. We introduce a unique viewpoint on this theorem: “the symmetry of time translation of Lagrangians derives the Euler–Lagrange equation and the energy conservation law, simultaneously.” The proposed method is a combination of a discrete counter part of this statement and the discrete gradient method. It is also shown that the symmetry of space translation derives momentum-preserving schemes. Finally, we discuss the existence of discrete local conservation laws.

Discrete-velocity approximations represent a popular way for computing the Boltzmanncollision operator. The direct numerical evaluation of such methods involve a prohibitivecost, typically O(N2d + 1)where d is the dimension of the velocity space. In this paper, followingthe ideas introduced in [C. Mouhot and L. Pareschi, C. R. Acad. Sci. Paris Sér. IMath. 339 (2004) 71–76, C. Mouhot and L. Pareschi, Math.Comput. 75 (2006) 1833–1852], we derive fast summation techniquesfor the evaluation of discrete-velocity schemes which permits to reduce the computationalcost from O(N2d + 1) to O(N̅dNd log2N), N̅ ≪ N, with almost no loss of accuracy.

In this paper, we propose a numerical method to solve stochastic elliptic interfaceproblems with random interfaces. Shape calculus is first employed to derive theshape-Taylor expansion in the framework of the asymptotic perturbation approach. Given themean field and the two-point correlation function of the random interface, we can thusquantify the mean field and the variance of the random solution in terms of certain ordersof the perturbation amplitude by solving a deterministic elliptic interface problem andits tensorized counterpart with respect to the reference interface. Error estimates arederived for the interface-resolved finite element approximation in both, the physical andthe stochastic dimension. In particular, a fast finite difference scheme is proposed tocompute the variance of random solutions by using a low-rank approximation based on thepivoted Cholesky decomposition. Numerical experiments are presented to validate andquantify the method.