Mixing between miscible fluids of different densities is often encountered in practical applications such as astrophysics, atmospheric and oceanic flows, combustion, aerodynamics and Inertial Confinement Fusion. In these cases, large density variations arise due to compositional changes during mixing of fluids with very different molar masses, from temperature, or acoustic fluctuations. In general, flows with turbulent density fluctuations have been studied in conjunction with compressibility effects (e.g. in aerodynamics) or as a result of temperature changes (e.g. in combustion). Such flows have been the subject of numerous fundamental turbulence studies and many modeling strategies now exist. Nevertheless, large density variations can also occur due to mixing of different molar mass fluids. In this case, fundamental turbulence studies as well as specific engineering models are scarce. Even when such effects were considered, they were nevertheless coupled with additional acoustic or thermal effects, such that their role was difficult to ascertain.

In recent years, there has been a renewed interest to isolate the variable-density effects due to compositional changes in flows with different molar mass fluids. Thus, it is now known that variable-density mixing is asymmetric, with pure light fluid mixing faster than the pure heavy fluid, such that mixing models based on the Boussinesq approximation are not valid. Additional asymmetries can also be seen on the two sides of mixing layers generated by the basic fluid instabilities. In this talk, I will survey recent high resolution Direct Numerical Simulations results concerning canonical turbulent flows with large density variations as seen in Rayleigh-Taylor, Richtmyer-Meshkov, and Kelvin-Helmholtz instabilities, shock – variable density turbulence interaction, and shock bubble-interaction, by trying to emphasize common variable density effects. I will also discuss implications for turbulence modeling strategies and the significance for several important applications.