On the correction to Einstein's formula for the effective viscosity

Abstract

This paper is a follow-up of article [6], on the derivation of accurate effective models for viscous dilute suspensions. The goal is to identify an effective Stokes equation providing a $o(\lambda^2)$ approximation of the exact fluid-particle system, with $\lambda$ the solid volume fraction of the particles. This means that we look for an improvement of Einstein's formula for the effective viscosity in the form $\mu_{eff}(x) = \mu + \frac{5}{2} \mu \rho(x) \lambda + \mu_2(x) \lambda^2$. Under a separation assumption on the particles, we proved in [6] that if a $o(\lambda)^2$ Stokes effective approximation exists, the correction $\mu_2$ is necessarily given by a mean field limit, that can then be studied and computed under further assumptions on the particle configurations. Roughly, we go here from the conditional result of [6] to an unconditional result: we show that such a $o(\lambda^2)$ Stokes approximation indeed exists, as soon as the mean field limit exists. This includes the case of periodic and random stationary particle configurations.