Towards an efficient multiscale modeling of low-dimensional reactive systems: study of numerical closure procedures

TL;DR

This paper investigates how to improve multiscale simulation methods for low-dimensional reactive systems, emphasizing the importance of cluster dynamics and microscopic state construction for accuracy and efficiency.

Contribution

It introduces enhanced procedures for microscopic state generation and highlights the necessity of including cluster information in the equation free method.

Findings

Cluster-based information improves EFM accuracy.

Careful microscopic state construction is crucial.

Some errors persist despite increasing macroscopic variables.

Abstract

In this paper, we present a study on how to develop an efficient multiscale simulation strategy for the dynamics of chemically active systems on low-dimensional supports. Such reactions are encountered in a wide variety of situations, ranging from heterogeneous catalysis to electrochemical or (membrane) biological processes, to cite a few. We analyzed in this context different techniques within the framework of an important multiscale approach known as the equation free method (EFM), which "bridges the multiscale gap" by building microscopic configurations using macroscopic-level information only. We hereby considered two simple reactive processes on a one-dimensional lattice, the simplicity of which allowed for an in-depth understanding of the parameters controlling the efficiency of this approach. We demonstrate in particular that it is not enough to base the EFM on the time evolution of the average concentrations of particles on the lattice, but that the time evolution of clusters of particles has to be included as well. We also show how important it is for the accuracy of this method to carefully choose the procedure with which microscopic states are constructed, starting from the measured macroscopic quantities. As we also demonstrate that some errors cannot be corrected by increasing the number of observed macroscopic variables, this work points towards which procedures should be used in order to generate efficient and reliable multiscale simulations of these systems.