Spectral measures and iterative bounds for effective diffusivity of steady and space-time periodic flows

TL;DR

This paper develops an iterative method to analytically compute moments of spectral measures, enabling accurate bounds on effective diffusivity in periodic flows, extending previous spectral approaches to more complex fluid flow models.

Contribution

It introduces a new iterative approach to calculate spectral measure moments analytically, broadening the applicability of spectral bounds for effective diffusivity in periodic flows.

Findings

High-order bounds accurately capture asymptotic behavior

Method applies to both steady and space-time periodic flows

Demonstrated effectiveness in two-dimensional flows

Abstract

Over three decades ago the advection-diffusion equation for a steady fluid velocity field was homogenized, leading to a Stieltjes integral representation for the effective diffusivity, which is given in terms of a spectral measure of a compact self-adjoint operator and the P\'eclet number of the fluid flow. This result was recently extended to space-time periodic flows, instead involving an unbounded self-adjoint operator. Pad\'e approximants provide rigorous upper and lower bounds for Stieltjes functions in terms of the moments of the spectral measure. However, with the lack of a method for calculating the moments of the spectral measure for general fluid velocity fields, the utility of this powerful mathematical framework for calculating bounds for the effective diffusivity has not been fully realized. Here we significantly expand the applicability of this framework by providing an iterative method that enables an arbitrary number of moments, hence bounds, to be calculated analytically in closed form for both spatially and space-time periodic flows. The method is demonstrated for periodic flows in two spatial dimensions. The known asymptotic behavior of the effective diffusivity for a steady flow is accurately captured by high order upper and lower bounds, demonstrating the ability of the method to provide accurate estimates for the effective diffusivity for a broad range of parameter values.