A. J. Nonaka

A. Donev, A. J. Nonaka, Y. Sun et al.

We formulate low Mach number fluctuating hydrodynamic equations appropriate for modeling diffusive mixing in isothermal mixtures of fluids with different density and transport coefficients. These equations eliminate the fluctuations in pressure associated with the propagation of sound waves by replacing the equation of state with a local thermodynamic constraint. We demonstrate that the low Mach number model preserves the spatio-temporal spectrum of the slower diffusive fluctuations. We develop a strictly conservative finite-volume spatial discretization of the low Mach number fluctuating equations in both two and three dimensions and construct several explicit Runge-Kutta temporal integrators that strictly maintain the equation of state constraint. The resulting spatio-temporal discretization is second-order accurate deterministically and maintains fluctuation-dissipation balance in the linearized stochastic equations. We apply our algorithms to model the development of giant concentration fluctuations in the presence of concentration gradients, and investigate the validity of common simplifications such as neglecting the spatial non-homogeneity of density and transport properties. We perform simulations of diffusive mixing of two fluids of different densities in two dimensions and compare the results of low Mach number continuum simulations to hard-disk molecular dynamics simulations. Excellent agreement is observed between the particle and continuum simulations of giant fluctuations during time-dependent diffusive mixing.

A. J. Nonaka, Y. Sun, J. B. Bell et al.

Continuing on our previous work [ArXiv:1212.2644], we develop semi-implicit numerical methods for solving low Mach number fluctuating hydrodynamic equations appropriate for modeling diffusive mixing in isothermal mixtures of fluids with different densities and transport coefficients. We treat viscous dissipation implicitly using a recently-developed variable-coefficient Stokes solver [ArXiv:1308.4605]. This allows us to increase the time step size significantly compared to the earlier explicit temporal integrator. For viscous-dominated flows, such as flows at small scales, we develop a scheme for integrating the overdamped limit of the low Mach equations, in which inertia vanishes and the fluid motion can be described by a steady Stokes equation. We also describe how to incorporate advanced higher-order Godunov advection schemes in the numerical method, allowing for the treatment of fluids with high Schmidt number including the vanishing mass diffusion coefficient limit. We incorporate thermal fluctuations in the description in both the inertial and overdamped regimes. We apply our algorithms to model the development of giant concentration fluctuations during the diffusive mixing of water and glycerol, and compare numerical results with experimental measurements. We find good agreement between the two, and observe propagative (non-diffusive) modes at small wavenumbers (large spatial scales), not reported in published experimental measurements of concentration fluctuations in fluid mixtures. Our work forms the foundation for developing low Mach number fluctuating hydrodynamics methods for miscible multi-species mixtures of chemically reacting fluids.