Fast Solution of Fully Implicit Runge-Kutta and Discontinuous Galerkin in Time for Numerical PDEs, Part I: the Linear Setting

Fully implicit Runge-Kutta (IRK) methods have many desirable properties as time integration schemes in terms of accuracy and stability, but are rarely used in practice with numerical PDEs due to the difficulty of solving the stage equations. This paper introduces a theoretical and algorithmic preconditioning framework for solving the systems of equations that arise from IRK methods applied to linear numerical PDEs (without algebraic constraints). This framework also naturally applies to discontinuous Galerkin discretizations in time. Under quite general assumptions on the spatial discretization that yield stable time integration, the preconditioned operator is proven to have conditioning ∼ O(1), with only weak dependence on number of stages/polynomial order; for example, the preconditioned operator for 10th-order Gauss IRK has condition number less than two, independent of the spatial discretization. The new method can be used with arbitrary existing preconditioners for backward Euler-type time stepping schemes, and is amenable to the use of three-term recursion Krylov methods when the underlying spatial discretization is symmetric. The new method is demonstrated to be effective on various high-order finite-difference and finite-element discretizations of linear parabolic and hyperbolic problems, demonstrating fast, scalable solution of up to 10th order accuracy. The new method consistently outperforms existing block preconditioning approaches, and in several cases, the new method can achieve 4th-order accuracy using Gauss integration with roughly half the number of preconditioner applications and wallclock time as required using standard diagonally implicit RK methods. ∗BSS was supported by Lawrence Livermore National Laboratory under contract B639443, and as a Nicholas C. Metropolis Fellow under the Laboratory Directed Research and Development program of Los Alamos National Laboratory. OAK acknowledges the support of an Australian Government Research Training Program (RTP) Scholarship. †Theoretical Division, Los Alamos National Laboratory, U.S.A. (southworth@lanl.gov), http: //orcid.org/0000-0002-0283-4928 ‡School of Mathematics, Monash University, Australia (oliver.krzysik@monash.edu), https: //orcid.org/0000-0001-7880-6512 §Center for Applied Scientific Computing, Lawrence Livermore National Laboratory, U.S.A. (pazn er1@llnl.gov) ¶Department of Applied Mathematics, University of Waterloo, Waterloo, Canada (hdesterck@uwa terloo.ca) 1 ar X iv :2 10 1. 00 51 2v 3 [ m at h. N A ] 1 3 Ju l 2 02 1

[1]  Trygve K. Nilssen,et al.  Order optimal preconditioners for fully implicit Runge‐Kutta schemes applied to the bidomain equations , 2011 .

[2]  Will Pazner,et al.  Efficient low-order refined preconditioners for high-order matrix-free continuous and discontinuous Galerkin methods , 2019, SIAM J. Sci. Comput..

[3]  S. Orszag.  Spectral methods for problems in complex geometries , 1980 .

[4]  Martin J. Gander,et al.  Analysis of a New Space-Time Parallel Multigrid Algorithm for Parabolic Problems , 2014, SIAM J. Sci. Comput..

[5]  Thomas A. Manteuffel,et al.  Nonsymmetric Algebraic Multigrid Based on Local Approximate Ideal Restriction (ℓAIR) , 2017, SIAM J. Sci. Comput..

[6]  Laurent O. Jay,et al.  Inexact Simplified Newton Iterations for Implicit Runge-Kutta Methods , 2000, SIAM J. Numer. Anal..

[7]  Yousef Saad,et al.  A Flexible Inner-Outer Preconditioned GMRES Algorithm , 1993, SIAM J. Sci. Comput..

[8]  Bojan Orel,et al.  Real pole approximations to the exponential function , 1991 .

[9]  Dominik Schötzau,et al.  Time Discretization of Parabolic Problems by the HP-Version of the Discontinuous Galerkin Finite Element Method , 2000, SIAM J. Numer. Anal..

[10]  Kent-André Mardal,et al.  Order-Optimal Preconditioners for Implicit Runge-Kutta Schemes Applied to Parabolic PDEs , 2007, SIAM J. Sci. Comput..

[11]  Martin Kronbichler,et al.  Multigrid for Matrix-Free High-Order Finite Element Computations on Graphics Processors , 2019, ACM Trans. Parallel Comput..

[12]  John C. Butcher,et al.  A new type of singly-implicit Runge-Kutta method , 2000 .

[13]  L. Trefethen,et al.  Stability of the method of lines , 1992, Spectra and Pseudospectra.

[14]  Ricardo H. Nochetto,et al.  Galerkin and Runge–Kutta methods: unified formulation, a posteriori error estimates and nodal superconvergence , 2011, Numerische Mathematik.

[15]  Per-Olof Persson,et al.  Stage-parallel fully implicit Runge-Kutta solvers for discontinuous Galerkin fluid simulations , 2017, J. Comput. Phys..

[16]  Douglas N. Arnold,et al.  Unified Analysis of Discontinuous Galerkin Methods for Elliptic Problems , 2001, SIAM J. Numer. Anal..

[17]  E. Hairer,et al.  Solving Ordinary Differential Equations II: Stiff and Differential-Algebraic Problems , 2010 .

[18]  Iain Smears,et al.  Robust and efficient preconditioners for the discontinuous Galerkin time-stepping method , 2016, 1608.08184.

[19]  Hao Chen.  Kronecker product splitting preconditioners for implicit Runge-Kutta discretizations of viscous wave equations , 2016 .

[20]  Piet J. van der Houwen,et al.  Triangularly Implicit Iteration Methods for ODE-IVP Solvers , 1997, SIAM J. Sci. Comput..

[21]  Thomas Richter,et al.  Efficient numerical realization of discontinuous Galerkin methods for temporal discretization of parabolic problems , 2013, Numerische Mathematik.

[22]  Jens Markus Melenk,et al.  Fully discrete hp-finite elements: fast quadrature , 2001 .

[23]  Ernst Hairer,et al.  Implementation of Implicit Runge-Kutta Methods , 1996 .

[24]  Xiangmin Jiao,et al.  Optimal and Low-Memory Near-Optimal Preconditioning of Fully Implicit Runge-Kutta Schemes for Parabolic PDEs , 2020, SIAM J. Sci. Comput..

[25]  P. Raviart,et al.  On a Finite Element Method for Solving the Neutron Transport Equation , 1974 .

[26]  Benjamin Seibold,et al.  Spatial Manifestations of Order Reduction in Runge-Kutta Methods for Initial Boundary Value Problems. , 2017, 1712.00897.

[27]  Stefan Vandewalle,et al.  Multigrid Methods for Implicit Runge-Kutta and Boundary Value Method Discretizations of Parabolic PDEs , 2005, SIAM J. Sci. Comput..

[28]  Thomas A. Manteuffel,et al.  Reduction-based Algebraic Multigrid for Upwind Discretizations , 2017, 1704.05001.

[29]  Ricardo H. Nochetto,et al.  A posteriori error analysis for higher order dissipative methods for evolution problems , 2006, Numerische Mathematik.

[30]  Chi-Wang Shu,et al.  Runge–Kutta Discontinuous Galerkin Methods for Convection-Dominated Problems , 2001, J. Sci. Comput..

[31]  Hao Chen,et al.  A splitting preconditioner for the iterative solution of implicit Runge-Kutta and boundary value methods , 2014 .

[32]  Syvert P. Nørsett,et al.  Runge-Kutta methods with a multiple real eigenvalue only , 1976 .

[33]  Claudio Canuto,et al.  Finite-Element Preconditioning of G-NI Spectral Methods , 2009, SIAM J. Sci. Comput..

[34]  W. Hoffmann,et al.  Approximating Runge-Kutta matrices by triangular matrices , 1995 .

[35]  D. Arnold.  An Interior Penalty Finite Element Method with Discontinuous Elements , 1982 .

[36]  Piet J. van der Houwen,et al.  Parallel linear system solvers for Runge-Kutta methods , 1997, Adv. Comput. Math..

[37]  Khosro Shahbazi,et al.  An explicit expression for the penalty parameter of the interior penalty method , 2022 .

[38]  Theodore A. Bickart,et al.  An Efficient Solution Process for Implicit Runge–Kutta Methods , 1977 .