Statistical comparison between experiments and numerical simulations of shock-accelerated gas cylinders

We present detailed spatial analysis comparing experimental data and numerical simulation results for Richtmyer-Meshkov instability experiments of Prestridge et al. and Tomkins et al. These experiments consist, respectively, of one and two diffuse cylinders of sulphur hexafluoride (SF{sub 6}) impulsively accelerated by a Mach 1.2 shockwave in air. The subsequent fluid evolution and mixing is driven by the deposition of baroclinic vorticity at the interface between the two fluids. Numerical simulations of these experiments are performed with three different versions of high resolution finite volume Godunov methods, including a new weighted adaptive Runge-Kutta (WARK) scheme. We quantify the nature of the mixing using using integral measures as well as fractal analysis and continuous wavelet transforms. Our investigation of the gas cylinder configurations follows the path of our earlier studies of the geometrically and dynamically more complex gas 'curtain' experiment. In those studies, we found significant discrepancies in the details of the experimentally measured mixing and the details of the numerical simulations. Here we evaluate the effects of these hydrodynamic integration techniques on the diffuse gas cylinder simulations, which we quantitatively compare with experimental data