Atwood number effects in statistically stationary Rayleigh Taylor turbulence

The effect of Atwood number, A, on Rayleigh-Taylor (RT) turbulence is studied using a statistically stationary Rayleigh-Taylor (SRT) flow configuration. The SRT flow configuration represents the flow dynamics of late-time self-similar RT growth in the limit of small wavelength initial perturbations, and has been validated with the direct numerical simulation results of Cabot & Cook (2006) at A=0.5. Due to stationarity, SRT flow can be simulated over a long time at the same flow condition, leading to excellent statistical convergence. This is particularly useful for higher Atwood number cases that are challenging to simulate in the traditional temporally evolving RT configuration. We present justifications for the validity of the flow configuration beyond A = 0.5, and apply it to Atwood numbers in the range of 0.01 – 0.8. Results are validated against others in the literature. Scalings are proposed for ensemble-averaged quantities to obtain normalized profiles that are independent of the Atwood number, and analytical fits to these normalized profiles are examined. Finally, the consistency of the scaling analysis and model profiles are verified using the ensemble-averaged transport equations.