Reactive strip method for mixing and reaction in two dimensions

TL;DR

This paper introduces a numerical method that simplifies the complex 2D reactive transport problem into a 1D problem, enabling efficient simulation of mixing and reactions in high Péclet number flows.

Contribution

The method generalizes the Diffusive Strip Method to handle multiple reacting species in two-dimensional flows, improving computational efficiency for high Péclet number scenarios.

Findings

Effective modeling of reactive transport in 2D flows.

Validation with shear, vortex, and chaotic flows.

Applicable to high Péclet number regimes.

Abstract

A numerical method to efficiently solve for mixing and reaction of scalars in a two-dimensional flow field at large P\'eclet numbers but otherwise arbitrary Damk\"ohler numbers is reported. We consider a strip of one reactant in a pool of another reactant, both of which are advected with the known velocity field. We first establish that the system evolution for such a system under certain conditions is described by a locally one-dimensional reaction-diffusion problem. The approximation of a locally one-dimensional dynamics is true for cases where the strip thickness is smaller than the local radius of curvature and also when the strip thickness is smaller than the distance between adjacent strips. We first demonstrate the method for the transport of a conservative scalar under a linear shear flow, point vortex and a chaotic sine flow. We then proceed to consider the situation with a simple bimolecular reaction between two reactants to yield a single product. The methodology presented herewith essentially generalizes nontrivially the Diffusive Strip Method developed by Meunier and Villermaux (J. Fluid Mech. 662, 134-172 (2010)) to address passive scalar transport, to the generalized situation with multiple reacting species. In essence, the reduction of dimensionality of the problem, which renders the 2D problem 1D, allows one to efficiently model reactive transport under high P\'eclet numbers which are otherwise prohibitively difficult to resolve from classical finite difference or finite element based methods.