Showing 1–3 of 3 results for author: Louvet, V

Task-based adaptive multiresolution for time-space multi-scale reaction-diffusion systems on multi-core architectures

Authors: Stéphane Descombes, Max Duarte, Thierry Dumont, Thomas Guillet, Violaine Louvet, Marc Massot

Abstract: A new solver featuring time-space adaptation and error control has been recently introduced to tackle the numerical solution of stiff reaction-diffusion systems. Based on operator splitting, finite volume adaptive multiresolution and high order time integrators with specific stability properties for each operator, this strategy yields high computational efficiency for large multidimensional comput… ▽ More A new solver featuring time-space adaptation and error control has been recently introduced to tackle the numerical solution of stiff reaction-diffusion systems. Based on operator splitting, finite volume adaptive multiresolution and high order time integrators with specific stability properties for each operator, this strategy yields high computational efficiency for large multidimensional computations on standard architectures such as powerful workstations. However, the data structure of the original implementation, based on trees of pointers, provides limited opportunities for efficiency enhancements, while posing serious challenges in terms of parallel programming and load balancing. The present contribution proposes a new implementation of the whole set of numerical methods including Radau5 and ROCK4, relying on a fully different data structure together with the use of a specific library, TBB, for shared-memory, task-based parallelism with work-stealing. The performance of our implementation is assessed in a series of test-cases of increasing difficulty in two and three dimensions on multi-core and many-core architectures, demonstrating high scalability. △ Less

Submitted 18 October, 2016; v1 submitted 9 June, 2015; originally announced June 2015.

Journal ref: SMAI Journal of Computational Mathematics, Vol. 3 (2017) 29-51

Analysis of operator splitting in the non-asymptotic regime for nonlinear reaction-diffusion equations. Application to the dynamics of premixed flames

Authors: Stéphane Descombes, Max Duarte, Thierry Dumont, Frédérique Laurent, Violaine Louvet, Marc Massot

Abstract: In this paper we mathematically characterize through a Lie formalism the local errors induced by operator splitting when solving nonlinear reaction-diffusion equations, especially in the non-asymptotic regime. The non-asymptotic regime is often attained in practice when the splitting time step is much larger than some of the scales associated with either source terms or the diffusion operator when… ▽ More In this paper we mathematically characterize through a Lie formalism the local errors induced by operator splitting when solving nonlinear reaction-diffusion equations, especially in the non-asymptotic regime. The non-asymptotic regime is often attained in practice when the splitting time step is much larger than some of the scales associated with either source terms or the diffusion operator when large gradients are present. In a series of previous works a reduction of the asymptotic orders for a range of large splitting time steps related to very short time scales in the nonlinear source term has been studied, as well as that associated with large gradients but for linearized equations. This study provides a key theoretical step forward since it characterizes the numerical behavior of splitting errors within a more general nonlinear framework, for which new error estimates can be derived by coupling Lie formalism and regularizing effects of the heat equation. The validity of these theoretical results is then assessed in the framework of two numerical applications, a KPP-type reaction wave where the influence of stiffness on local error estimates can be thoroughly investigated; and a much more complex problem, related to premixed flame dynamics in the low Mach number regime with complex chemistry and detailed transport, for which the present theoretical study shows to provide relevant insights. △ Less

Submitted 15 March, 2014; v1 submitted 8 February, 2014; originally announced February 2014.

Comments: SIAM Journal on Numerical Analysis (2014) 1-24

Journal ref: SIAM Journal on Numerical Analysis 52 (3), 1311-1334 (2014)

Adaptive time splitting method for multi-scale evolutionary partial differential equations

Authors: Stéphane Descombes, Max Duarte, Thierry Dumont, Violaine Louvet, Marc Massot

Abstract: This paper introduces an adaptive time splitting technique for the solution of stiff evolutionary PDEs that guarantees an effective error control of the simulation, independent of the fastest physical time scale for highly unsteady problems. The strategy considers a second order Strang method and another lower order embedded splitting scheme that takes into account potential loss of order due to t… ▽ More This paper introduces an adaptive time splitting technique for the solution of stiff evolutionary PDEs that guarantees an effective error control of the simulation, independent of the fastest physical time scale for highly unsteady problems. The strategy considers a second order Strang method and another lower order embedded splitting scheme that takes into account potential loss of order due to the stiffness featured by time-space multi-scale phenomena. The scheme is then built upon a precise numerical analysis of the method and a complementary numerical procedure, conceived to overcome classical restrictions of adaptive time stepping schemes based on lower order embedded methods, whenever asymptotic estimates fail to predict the dynamics of the problem. The performance of the method in terms of control of integration errors is evaluated by numerical simulations of stiff propagating waves coming from nonlinear chemical dynamics models as well as highly multi-scale nanosecond repetitively pulsed gas discharges, which allow to illustrate the method capabilities to consistently describe a broad spectrum of time scales and different physical scenarios for consecutive discharge/post-discharge phases. △ Less

Submitted 31 July, 2011; v1 submitted 19 April, 2011; originally announced April 2011.

Comments: Accepted to publication in Confluentes Mathematici. Dedication : Cet article est dédié à la mémoire de Michelle Schatzman. Spécialiste des méthodes de décomposition d'opérateur, sa grande clairvoyance scientifique lui a permis d'orienter plusieurs chercheurs débutants sur ce sujet à un moment où il pouvait sembler achevé. Michelle aimait dire qu'il n'y a pas de frontière entre les branches des mathématiques et que seule une grande culture permet de naviguer dans cette forêt et d'y trouver les bonnes techniques pour résoudre un problème. Ce travail est un hommage; à la croisée des mathématiques et de leurs applications effectives, il tente d'illustrer cette assertion. Michelle, ton dynamisme, ton humour et ton plaisir à parler mathématiques nous manquent

Journal ref: Confluentes Mathematici, Vol. 3, No. 3 (2011) 413-443