Experimental adventures in variable-density mixing

Variable-density mixing occurs in the ocean and atmosphere, inertial confinement fusion capsules, stars, and many industrial processes. Our team studies mixing at the laboratory scale, with the goal of extending our understanding to larger and smaller length scales and into multiphysics regimes that include such physics as chemical and thermonuclear reactions and high-energy density plasmas. I will provide examples of studies from Los Alamos in variable-density mixing in multiple regimes, including shock-driven mixing, subsonic mixing, and high-energy density (HED) plasma mixing. In all cases, the density differences of the mixing fluids are at least double, making the mixing well outside regimes covered by the Boussinesq approximation. In these flows, we have discovered time-dependent mixing phenomena, asymmetries in mixing and distribution of turbulent kinetic energy, and interscale energy transfer driven by the mean flow that moves energy from small to large scales via eddy stretching. We have improved our physical understanding by implementing diagnostics such as particle image velocimetry and planar laser-induced fluorescence in challenging, shock-driven flow regimes. Our experimental studies and physical understanding benefit from collaborations with colleagues studying turbulence modeling and performing validation simulations, and these collaborations push us to further improve our experimental platforms and diagnostics.

This article appears in the following collection:

Figure 1

(a) Supernova 1987a remnant from the Hubble Space Telescope showing a bright ring of debris illuminated by x rays. Shocks from the explosion are heating the debris and making it glow. Image credit: National Aeronautics and Space Administration (NASA), European Space Agency (ESA), and P. Challis. (b) Cassiopeia A supernova remnant x-ray image showing shock wave (blue) and various elements, including sulfur (yellow), silicon (red), calcium (green), and iron (purple). Image credit: NASA, Chandra X-Ray Center (CXC), and Smithsonian Astrophysical Observatory (SAO).

Figure 2

(a) ICF capsule showing gold hohlraum and laser paths. (b) X-ray microscope image of a capsule imploding at NIF showing perturbations caused by the tent support structure and the gas fill tube [3].

Figure 3

Turbulence Laboratory at Los Alamos National Laboratory (photo) with Turbulent Mixing Tunnel (TMT) in the left and the Vertical Shock Tube (VST) on the right. Detailed rendering of TMT showing test section in figure on the right.

Figure 4

Velocity and density fields at various downstream locations ($x_1/d_0$) for (a) air jet and (b) ${\mathrm{SF}}_6$ jet. Density legends are included on each image because ranges are adjusted for downstream mixing conditions, and the jet centerline is marked in each figure with the red vertical lines. Images are from experiments reported by Charonko and Prestridge [11].

Figure 5

Energy transfer mechanisms in air (top) and heavy ${\mathrm{SF}}_6$ jet (bottom) showing the vortex stretching mechanism in variable-density jet that moves energy from smaller to larger scales at all measured length scales of the flow. Size of the arrows is proportional to the amount of energy transfer. Figure adapted from Lai et al. [14]

Figure 6

PLIF images of heavy gas curtain of ${\mathrm{SF}}_6$ surrounded by air (top left) accelerated by a Mach 1.36 shock wave (moves from left to right). Time series shows initial condition and compression by the shock, and extends out to 1.7 ms after the initial shock of the curtain. Figure is created from experiments reported in Orlicz et al. [23].

Figure 7

Density specific-volume covariance in gas curtain experiments determined from instantaneous realizations through the mixing region at $515\mu \mathrm{s}$ after shock and from an ensemble (thick red line), showing differences based on averaging techniques used. Turbulent mass flux in streamwise ($a_1$) and spanwise ($a_2$) directions compared to density profile in reshocked gas curtain. Figures are derived from experiments described in Ref. [30].

Figure 8

New experiments using the LANSCE pRad facility, driving a Mach 8.9 (top row) shock into xenon, with different interfaces separating the xenon from helium gas. The initial condition perturbations are (a) 14.8 mm and (b) 8.9 mm. Bottom row shows the mixing region between the xenon and helium about $18\mu \mathrm{s}$ after shock acceleration. Imprint of the initial conditions is visible in the mixing region.

Figure 9

NIF Shear platform experiments comparing evolution of two different types of initial conditions at 25 ns. The shock tube is approximately 5.7 mm long. The initial conditions are changed by modifying the surface roughness of the copper tracer foil. Kelvin-Helmholtz shear rollers can be seen in the smooth copper plan view. The rough initial conditions did not develop strong rollers, but the mixing zone width grew faster. Figures adapted from Flippo et al. with permission [37].

Figure 10

NIF Mshock platform with schematic, preshot radiograph, and curtain of dense material shown 5 ns after reshock. Reshock occurred 13 ns after first shock. Experiments described in Ref. [39].