Numerical investigations of Rayleigh-Taylor instability with a density gradient layer

The Rayleigh-Taylor instability (RTI) with a density gradient layer was investigated by implicit large eddy simulations (ILES) in a framework of three dimensional compressible multiphase flow with the single fluid approximation. The simulations are firstly initialized by multi-mode random perturbations in classic RTI and the associated case with a premixed layer. It is found that the late time behaviours of the turbulent mixing layer in classic and premixed cases are similar although an initial premixed layer slows down the growth of the turbulent mixing layer significantly. A variety of flow quantities for different cases are compared and studied at different time. In addition, the turbulent mixing caused by RTI with a single-mode dominated perturbation is investigated in classic and premixed cases. Different stages are identified for the time evolution of turbulent mixing layer. The dynamics of bubble-spike structures at early time, and the production and transfer of turbulent kinetic energy at late time are expatiated. Typical quantities of turbulent mixing are also summarized. It is found that there is an inactive stage before the fast growth of bubble-spike structures with an initial premixed layer and the spectra need more time to adjust until the flow turns to a self-similar regime. (C) 2021 Elsevier Ltd. All rights reserved.

Automated summary

This study investigated the Rayleigh-Taylor instability with a density gradient layer through implicit large eddy simulations, comparing and studying the behaviors of turbulent mixing layer in classic and premixed cases. Different flow quantities for various cases at different times were compared, and the dynamics of turbulent mixing caused by RTI with single-mode perturbations were examined. It was found that an initial premixed layer slows down the growth of turbulent mixing layer significantly and requires more time for spectra adjustment before turning to a self-similar regime.