Iterated Defect Correction Research Articles

Multi-parameter wavefield reconstruction inversion for wavespeed and attenuation with bound constraints and total variation regularization

Wavefield reconstruction inversion (WRI) extends the search space of Full Waveform Inversion (FWI) by allowing for wave equation errors during wavefield reconstruction to match the data from the first iteration. Then, the wavespeeds are updated from the wavefields by minimizing the source residuals. Performing these two tasks in alternating mode breaks down the nonlinear FWI as a sequence of two linear subproblems, relaying on the bilinearity of the wave equation. We solve this biconvex optimization with the alternating-direction method of multipliers (ADMM) to cancel out efficiently the data and source residuals in iterations and stabilize the parameter estimation with appropriate regularizations.Here, we extend WRI to viscoacoustic media for attenuation imaging. Attenuation reconstruction is challenging because of the small imprint of attenuation in the data and the crosstalks with velocities. To address these issues, we recast the multivariate viscoacoustic WRI as a triconvex optimization and update wavefields, squared slowness, and attenuation factor in alternating mode at each WRI iteration. This requires to linearize the attenuation-estimation subproblem via an approximated trilinear viscoacoustic wave equation. The iterative defect correction embedded in ADMM corrects the errors generated by this linearization, while the operator splitting allows us to tailor ℓ1 regularization to each parameter class. A toy numerical example shows that these strategies mitigate crosstalk artifacts and noise from the attenuation reconstruction. A more realistic example representative of the North sea further shows the ability of the method to jointly reconstruct the wavespeed and the attenuation starting from a very crude attenuation-free initial model, the moderate strength of the crosstalk artifacts and the sensitivity of the multi-parameter reconstruction to noise.

The adapted block boundary value methods for singular initial value problems

This paper deals with the numerical methods for solving singular initial value problems. By adapting the block boundary value methods (BBVMs) for regular initial value problems, a class of adapted BBVMs are constructed for singular initial value problems. It is proved under some suitable conditions that the adapted BBVMs are uniquely solvable, stable and convergent of order p, where p is the consistence order of the methods. Several numerical examples are performed to verify the stability, efficiency and accuracy of the adapted methods. Moreover, a comparison between the adapted BBVMs and the IEM-based iterated defect correction methods is given. The numerical results show that the adapted BBVMs are comparable.

Convergence Analysis for the Iterated Defect Correction Scheme of Finite Element Methods on Rectangle Grids

This paper develops a new method to analyze convergence of the iterated defect correction scheme of finite element methods on rectangular grids in both two and three dimensions. The main idea is to formulate energy inner products and energy (semi)norms into matrix forms. Then, two constants of two key inequalities involved are min and max eigenvalues of two associated generalized eigenvalue problems, respectively. Local versions on the element level of these two generalized eigenvalue problems are exactly solved to obtain sharp (lower) upper bounds of these two constants. This and some essential observations for iterated solutions establish convergence in 2D and the monotone decreasing property in 3D. For two dimensions the results herein improve those in literature; for three dimensions the results herein are new. Numerical results are presented to examine theoretical results.

Defect correction based velocity reconstruction for physically consistent simulations of non-Newtonian flows on unstructured grids

A new algorithm to recover centroidal velocities from face-normal data on two-dimensional unstructured staggered meshes is presented. The proposed approach uses iterative defect correction in conjunction with a lower-order accurate Gauss reconstruction to obtain second-order accurate centroidal velocities. We derive the conditions that guarantee the second-order accuracy of the velocity reconstruction and demonstrate its efficacy on arbitrary polygonal mesh topologies. The necessity of the proposed algorithm for non-Newtonian flow simulations is elucidated through numerical simulations of channel flow, driven cavity and backward facing step problems with power-law and Carreau fluids. Numerical investigations show that second-order accuracy of the reconstructed velocity field is critical to obtaining physically consistent solutions of vorticity-dominated flows on non-orthogonal meshes. It is demonstrated that the spurious solutions are not linked to discrete conservation and arise solely due to the lower order accuracy of velocity reconstruction. The importance of the proposed algorithm for hemodynamic simulations is highlighted through studies of laminar flow in an idealized stenosed artery using different blood models.

High‐order finite volume schemes based on defect corrections

AbstractFor the approximation of steady state solutions, we propose an iterated defect correction approach to achieve higher‐order accuracy. The procedure starts with the steady state solution of a low‐order scheme, in general a second order one. The higher‐order reconstruction step is applied a posteriori to estimate the local discretization error of the lower‐order finite volume scheme. The defect is then used to iteratively shift the basic lower‐order scheme to the desired higher‐order accuracy given by the polynomial reconstruction. Hence, instead of solving the high‐order discrete equations the low‐order basic scheme is solved several times. This avoids that the high‐order reconstruction with a large stencil has to be implemented into an existing basic solver and can be seen as a non‐intrusive approach to higher‐order accuracy.

Convergence of Multipower Defect-Correction for spectral computations of integral operators

In this paper we propose the Multipower Defect-Correction method, a generalization of the double iteration, to compute a cluster of eigenvalues and the associated invariant subspace from discretized integral operators. It consists of an inner/outer iteration where, inside a defect-correction iteration, p power iteration steps are performed. The approximate inverse used in the defect correction is built with an approximation to the reduced resolvent operator of a coarse discretization of the integral operator. The proposed method computes eigenpairs approximations by refining initial approximations obtained from a coarser dimensional problem. It is therefore meant for large dimensional problems. Furthermore, the kernel of the integral operator may be weakly singular. We provide a proof for the convergence of this Multipower Defect-Correction method. A numerical example illustrating the theory and the behavior of the method is also presented.

IDeC( k): A new velocity reconstruction algorithm on arbitrarily polygonal staggered meshes

We propose a novel algorithm for velocity reconstruction from staggered data on arbitrary polygonal staggered meshes. The formulation of the new algorithm is based on a constant polynomial reconstruction approach in conjunction with an iterative defect correction method and is referred to as the IDeC( k) reconstruction. The algorithm is designed for second order accuracy of the reconstructed velocity field and also leads to a consistent estimate of velocity gradients. Accuracy, convergence and robustness of the new algorithm are studied on different mesh topologies and the need for higher-order reconstruction is demonstrated. Numerical experiments for several cases including incompressible viscous flows establish the IDeC( k) reconstruction as a generic, fast, robust and higher-order accurate algorithm on arbitrary polygonal meshes.

Comparison of Node-Centered and Cell-Centered Unstructured Finite-Volume Discretizations: Viscous Fluxes

Cell-centered and node-centered approaches have been compared for unstructured finite-volume discretization of inviscid fluxes. The grids range from regular grids to irregular grids, including mixed-element grids and grids with random perturbations of nodes. Accuracy, complexity, and convergence rates of defect-correction iterations are studied for eight nominally second-order accurate schemes: two node-centered schemes with weighted and unweighted least-squares (LSQ) methods for gradient reconstruction and six cell-centered schemes two node-averaging with and without clipping and four schemes that employ different stencils for LSQ gradient reconstruction. The cell-centered nearest-neighbor (CC-NN) scheme has the lowest complexity; a version of the scheme that involves smart augmentation of the LSQ stencil (CC-SA) has only marginal complexity increase. All other schemes have larger complexity; complexity of node-centered (NC) schemes are somewhat lower than complexity of cell-centered node-averaging (CC-NA) and full-augmentation (CC-FA) schemes. On highly anisotropic grids typical of those encountered in grid adaptation, discretization errors of five of the six cell-centered schemes converge with second order on all tested grids; the CC-NA scheme with clipping degrades solution accuracy to first order. The NC schemes converge with second order on regular and/or triangular grids and with first order on perturbed quadrilaterals and mixed-element grids. All schemes may produce large relative errors in gradient reconstruction on grids with perturbed nodes. Defect-correction iterations for schemes employing weighted least-square gradient reconstruction diverge on perturbed stretched grids. Overall, the CC-NN and CC-SA schemes offer the best options of the lowest complexity and secondorder discretization errors. On anisotropic grids over a curved body typical of turbulent flow simulations, the discretization errors converge with second order and are small for the CC-NN, CC-SA, and CC-FA schemes on all grids and for NC schemes on triangular grids; the discretization errors of the CC-NA scheme without clipping do not converge on irregular grids. Accurate gradient reconstruction can be achieved by introducing a local approximate mapping; without approximate mapping, only the NC scheme with weighted LSQ method provides accurate gradients. Defect correction iterations for the CC-NA scheme without clipping diverge; for the NC scheme with weighted LSQ method, the iterations either diverge or converge very slowly. The best option in curved geometries is the CC-SA scheme that offers low complexity, second-order discretization errors, and fast convergence.

Algebraic multigrid within defect correction for the linearized Euler equations

AbstractGiven the continued difficulty of developing geometric multigrid methods that provide robust convergence for unstructured discretizations of compressible flow problems in aerodynamics, we turn to algebraic multigrid (AMG) as an alternative with the potential to automatically deal with arbitrary sources of stiffness on unstructured grids. We show here that AMG methods are able to solve linear problems associated with first‐order discretizations of the compressible Euler equations extremely rapidly. In order to solve the linear problems resulting from second‐order discretizations that are of practical interest, we employ AMG applied to the first‐order system within a defect correction iteration. It is demonstrated on two‐ and three‐dimensional test cases in a range of flow regimes (sub‐, trans‐ and supersonic) that the described method converges rapidly and robustly. Copyright © 2009 John Wiley & Sons, Ltd.

Explicit and implicit FEM-FCT algorithms with flux linearization

A new approach to the design of flux-corrected transport (FCT) algorithms for continuous (linear/multilinear) finite element approximations of convection-dominated transport problems is pursued. The algebraic flux correction paradigm is revisited, and a family of nonlinear high-resolution schemes based on Zalesak’s fully multidimensional flux limiter is considered. In order to reduce the cost of flux correction, the raw antidiffusive fluxes are linearized about an auxiliary solution computed by a high- or low-order scheme. By virtue of this linearization, the costly computation of solution-dependent correction factors is to be performed just once per time step, and there is no need for iterative defect correction if the governing equation is linear. A predictor–corrector algorithm is proposed as an alternative to the hybridization of high- and low-order fluxes. Three FEM-FCT schemes based on the Runge–Kutta, Crank–Nicolson, and backward Euler time-stepping are introduced. A detailed comparative study is performed for linear convection–diffusion equations.

Stabilization of explicit coupling in fluid–structure interaction involving fluid incompressibility

In this work, we propose a stabilized explicit coupling scheme for fluid–structure interaction problems involving a viscous incompressible fluid. The coupled discrete formulation is based on Nitsche’s method with a time penalty term giving L 2 -control on the fluid pressure variations at the interface. For a linear model problem, we prove that the scheme is stable, in the energy norm, irrespectively of the so-called added- mass effect, namely, the fluid–structure density ratio and the geometry of the domain. Numerical experiments, in the linear and non-linear case, show that optimal time accuracy can be obtained by performing one defect-correction iteration.

An implicit numerical algorithm for solving the general relativistic hydrodynamical equations around accreting compact objects

An implicit algorithm for solving the equations of general relativistic hydrodynamics in conservative form in three-dimensional axi-symmetry is presented. This algorithm is a direct extension of the pseudo-Newtonian implicit radiative magnetohydrodynamical solver – IRMHD – into the general relativistic regime.We adopt the Boyer–Lindquist coordinates and formulate the hydrodynamical equations in the fixed background of a Kerr black hole. The set of equations are solved implicitly using the hierarchical solution scenario (HSS). The HSS is efficient, robust and enables the use of a variety of solution procedures that range from a purely explicit up to fully implicit schemes. The discretization of the HD-equations is based on the finite volume formulation and the defect-correction iteration strategy for recovering higher order spatial and temporal accuracies. Depending on the astrophysical problem, a variety of relaxation methods can be applied. In particular the vectorized black-white Line-Gauss–Seidel relaxation method is most suitable for modeling accretion flows with shocks, whereas the Approximate Factorization Method is for weakly compressible flows.The results of several test calculations that verify the accuracy and robustness of the algorithm are shown. Extending the algorithm to enable solving the non-ideal MHD equations in the general relativistic regime is the subject of our ongoing research.

Stabilized explicit coupling for fluid–structure interaction using Nitsche's method

In this Note we propose a stabilized explicit coupling scheme for fluid–structure interaction based on Nitsche's method. The scheme is stable irrespective of the fluid–solid density ratio. Numerical experiments show that optimal time accuracy can be obtained by performing a few defect-correction iterations. To cite this article: E. Burman, M.A. Fernández, C. R. Acad. Sci. Paris, Ser. I 345 (2007).

A trust-region strategy for manifold-mapping optimization

Studying the space-mapping technique by Bandler et al. [J. Bandler, R. Biernacki, S. Chen, P. Grobelny, R.H. Hemmers, Space mapping technique for electromagnetic optimization, IEEE Trans. Microwave Theory Tech. 42 (1994) 2536–2544] for the solution of optimization problems, we observe the possible difference between the solution of the optimization problem and the computed space-mapping solution. We repair this discrepancy by exploiting the correspondence with defect-correction iteration and we construct the manifold-mapping algorithm, which is as efficient as the space-mapping algorithm but converges to the exact solution. To increase the robustness of the algorithm we introduce a trust-region strategy (a regularization technique) based on the generalized singular value decomposition of the linearized fine and coarse manifold representations. The effect of this strategy is shown by the solution of a variety of small non-linear least squares problems. Finally we show the use of the technique for a more challenging engineering problem.

Analysis of a defect correction method for geometric integrators

We discuss a new variant of Iterated Defect Correction (IDeC), which increases the range of applicability of the method. Splitting methods are utilized in conjunction with special integration methods for Hamiltonian systems, or other initial value problems for ordinary differential equations with a particular structure, to solve the neighboring problems occurring in the course of the IDeC iteration. We demonstrate that this acceleration technique serves to rapidly increase the convergence order of the resulting numerical approximations, up to the theoretical limit given by the order of certain superconvergent collocation methods.

Modified defect correction algorithms for ODEs. Part II: Stiff initial value problems

As shown in part I of this paper and references therein, the classical method of Iterated Defect Correction (IDeC) can be modified in several nontrivial ways, extending the flexibility and range of applications of this approach. The essential point is an adequate definition of the defect, resulting in a significantly more robust convergence behavior of the IDeC iteration, in particular, for nonequidistant grids. The present part II is devoted to the efficient high-order integration of stiff initial value problems. By means of model problem investigation and systematic numerical experiments with a set of stiff test problems, our new versions of defect correction are systematically evaluated, and further algorithmic measures are proposed for the stiff case. The performance of the different variants under consideration is compared, and it is shown how strong coupling between non-stiff and stiff components can be successfully handled.

Space Mapping and Defect Correction

Abstract In this paper we show that space-mapping optimization can be understood in the framework of defect correction. Then, space-mapping algorithms can be seen as special cases of defect correction iteration. In order to analyze the properties of space mapping and the space-mapping function, we introduce the new concept of flexibility of the underlying models. The best space-mapping results are obtained for so-called equally flexible models. By introducing an a±ne operator as a left preconditioner, two models can be made equally flexible, at least in the neighborhood of a solution. This motivates an improved space-mapping (or manifold-mapping) algorithm. The left preconditioner complements traditional space mapping where only a right preconditioner is used. In the last section a few simple examples illustrate some of the phenomena analyzed in this paper.

Modified Defect Correction Algorithms for ODEs. Part I: General Theory

The well-known method of Iterated Defect Correction (IDeC) is based on the following idea: Compute a simple, basic approximation and form its defect w.r.t. the given ODE via a piecewise interpolant. This defect is used to define an auxiliary, neighboring problem whose exact solution is known. Solving the neighboring problem with the basic discretization scheme yields a global error estimate. This can be used to construct an improved approximation, and the procedure can be iterated. The fixed point of such an iterative process corresponds to a certain collocation solution. We present a variety of modifications to this algorithm. Some of these have been proposed only recently, and together they form a family of iterative techniques, each with its particular advantages. These modifications are based on techniques like defect quadrature (IQDeC), defect interpolation (IPDeC), and combinations thereof. We investigate the convergence on locally equidistant and nonequidistant grids and show how superconvergent approximations can be obtained. Numerical examples illustrate our considerations. The application to stiff initial value problems will be discussed in Part II of this paper.

Computation of low Mach thermical flows with implicit upwind methods

In order to apply a Godunov method to steady and unsteady thermical flows, we first present a truncation error analysis which proves that the Turkel preconditioned Roe splitting is applicable to flows modelized by the classical low Mach number model, including thermics. The impact of the Turkel preconditioning on the first-order/second-order pseudo-Newton algorithm (referred in the literature as the defect-correction iteration) is also analysed. Implicit methods for steady flow resolution and for accurate unsteady time advancing are proposed and studied. Numerical illustrations involve the natural convection in a square box and a flow around a perforated wall inspired by helicopter engine applications.

Analytical and numerical treatment of a singular initial value problem in avalanche modeling

We discuss a leading-edge model used in the computation of the run-out length of dry-flowing avalanches. The model has the form of a singular initial value problem for a scalar ordinary differential equation describing the avalanche dynamics. Existence, uniqueness and smoothness properties of the analytical solution are shown. We also prove the existence of a unique root of the solution. Moreover, we present a FORTRAN 90 code for the numerical computation of the run-out length. The code is based on a solver for singular initial value problems which is an implementation of the acceleration technique known as iterated defect correction based on the implicit Euler method.