Joseph Conroy

Numerical Simulation of a Shock-Accelerated Multiphase Fluid Interface

A Richtmyer-Meshkov Instability (RMI) [1, 2] is generated when an interface between two different fluids is impulsively accelerated. The instability develops due to misalignment of the density and pressure interfaces. This misalignment results in the deposition of vorticity, causing the formation of an instability that grows nonlinearly with time and eventually may transition to fully turbulent flow. It has been recently shown that a similar class of instability can evolve in a multi-phase flow [3], where the density gradient is caused by a second, non-fluid phase.

Analogues of Rayleigh-Taylor and Richtmyer-Meshkov instabilities in flows with nonuniform particle and droplet seeding

The well-knownRayleigh-Taylor(RT) and Richtmyer-Meshkov(RM) instabilities characterize the behavior of flows where two gases (or fluids) of different densities mix due to gravity (RT) or due to impulsive acceleration (RM). Recently, analogous instabilities have been observed in two-phase flows where the seeding density of the second phase, e.g., particles or droplets in gas, and the resulting average density, is initially non-uniform. The forcing causes the second phase to move with respect to the embedding medium. With sufficient seeding concentration, this leads to entrainment of the embedding phase. The resulting movement is similar to the movement that would evolve in a mixing flow with no secondphaseseeding,butwithnon-uniformdensity(notunlikeamixtureoflighter and heavier gases), where RT and RM instabilities develop in the case of gravityinduced and impulsive acceleration, respectively. The hydrocode SHAMRC has been used in the past to study the formation and growth of the RM instability. Here we attempt to use it to model the first order formation and growth phenomena of the new class of instability in two-phase flows first, by approximating the second phase as a continuous fluid with an averaged density, and second, by taking the relative motion of particles (droplets)into accountexplicitly. The initial conditions are varied to provide a wide range of instability growth rates. Comparison of the numerical results with experiment shows good agreement.

Vortex deposition in shock-accelerated gas with particle/droplet seeding

We present an experimental and numerical study of post-shock evolution of gas initially seeded with small droplets or particles. In two-phase media with gas being the embedding phase occupying most of the volume, shock acceleration can lead to vortex formation. The physical mechanism responsible for the vorticity deposition in this case is different from that of Richtmyer-Meshkov instability that would emerge on a gas-gas density interface. After the shock passage, the particles or droplets lag behind the surrounding gas. Momentum exchange between the embedded phase and the embedding phase leads to non-uniform local equilibrium velocity distribution, and thus to shear and vortex formation. Here we investigate shock interaction with a cylindrical particle-seeded column (with and without reshock).

Morphology of shock-accelerated multiphase flow: experiment and modeling

It was recently observed that vortices develop in multiphase media of non-uniform average density undergoing acceleration. This vortex formation is somewhat akin to vortex roll-up in gases of fluids (single-phase) due to Rayleigh‐Taylor or Richtmyer—Meshkov instabilities. Differences in the underlying physics are negligible in the case of sustained modest acceleration (Rayleigh‐Taylor instability), where conservation of momentum accounts for different velocities of volumes with different average densities. In the case of impulsive acceleration, the mechanism responsible for the multiphase analog of Richtmyer‐Meshkov instability is peculiar to multiphase flow and is explained by post-acceleration interaction of the embedded phase (e.g., droplets) with the embedding phase (e.g., gas). Impulsive acceleration of a multiphase medium can also produce spatial rearrangement of the embedded particles or droplets in accordance with their size, noticeably altering the observed flow morphology. A careful numerical simulation explicitly accounting for the embedded phase behavior is required to faithfully reproduce the experimental results. & S. Kumar