Multiparticle Collision Dynamics for Ferrofluids

TL;DR

This paper introduces a hybrid mesoscopic simulation model combining multi-particle collision dynamics with stochastic magnetization dynamics to study ferrofluid flow and magnetic field effects.

Contribution

A novel hybrid simulation method for ferrofluids that captures fluctuating hydrodynamics and magnetization dynamics, validated against theoretical and benchmark flow problems.

Findings

Reproduces the magnetoviscous effect in Poiseuille flow.

Quantitative agreement with theoretical predictions for rotational viscosity.

Flow around a square cylinder shows increased recirculation length and decreased drag under magnetic fields.

Abstract

Detailed studies of the intriguing field-dependent dynamics and transport properties of confined flowing ferrofluids require efficient mesoscopic simulation methods that account for fluctuating ferrohydrodynamics. Here, we propose such a new mesoscopic model for the dynamics and flow of ferrofluids, where we couple the multi-particle collision dynamics method as a solver for the fluctuating hydrodynamics equations to the stochastic magnetization dynamics of suspended magnetic nanoparticles. This hybrid model is validated by reproducing the magnetoviscous effect in Poiseuille flow, obtaining the rotational viscosity in quantitative agreement with theoretical predictions. We also illustrate the new method for the benchmark problem of flow around a square cylinder. Interestingly, we observe that the length of the recirculation region is increased whereas the drag coefficient is decreased in ferrofluids when an external magnetic field is applied, compared with the field-free case at the same effective Reynolds number. The presence of thermal fluctuations and the flexibility of this particle-based mesoscopic method provides a promising tool to investigate a broad range of flow phenomena of magnetic fluids and could also serve as an efficient way to simulate solvent effects when colloidal particles are immersed in ferrofluids.