Abstract
We report implicit large-eddy simulations of (Mach 1.2) shocked SF6 gas curtain (GC) experiments addressing the impact of initial conditions (ICs) on material mixing and transition to turbulence. Initial GCs with realistic three-dimensional characteristics are generated with separate Navier–Stokes–Boussinesq simulations of a mixture of SF6 and air falling through the shock-tube test section. SF6 concentration fluctuations present in the laboratory experiments are emulated to address their potential effects. The predicted evolution of shocked GC widths is fairly robust and insensitive to ICs and grid resolution before reshock and in good agreement with laboratory experiments. After reshock, predicted results are sensitive to IC spectral content and its consequences on the morphology of the thicker more-complex mixing layers at reshock time. The presence of small-scale material-concentration fluctuations in ICs can promote late-time features traditionally associated with transition to turbulence, i.e., faster GC width growth, more isotropic features, and self-similar spectra. An effective data reduction procedure is found useful in improving comparisons with the laboratory data and characterizing the instability behaviors. As in our recent planar Richtmyer–Meshkov studies, we find that a single IC parameter can be usefully identified as relevant in determining whether the shock-driven flow is in linear ballistic or nonlinear mode-coupling regimes; we are thus able to demonstrate the recently reported bipolar behavior of the planar Richtmyer–Meshkov instability also in the case of the shocked GC configuration.
Acknowledgements
The authors thank K. Prestridge, S. Balasubramanian, B.J. Balakumar, and C.A. Zoldi-Sood for stimulating discussions and sharing information and data from their experiments and simulations. Authors also thank J.R. Ristorcelli and M.J. Andrews for very helpful discussions. LANL is operated by the Los Alamos National Security, LLC for the US Department of Energy NNSA under contract number DE-AC52–06NA25396. This work was made possible by funding from the LANL Laboratory Directed Research and Development Program on “Turbulence by Design” at LANL through directed research project number 20090058DR.