A comparison of mix models for the Rayleigh–Taylor instability

Four mix models, implemented into an Arbitrary Lagrangian–Eulerian (ALE) multi-physics code, are compared on simulations of the Rayleigh–Taylor instability. The specific models of interest are a mass diffusion model, the k–Lk–L turbulence model, the BHR turbulence model, and a multifluid interpenetration mix model. The bubble growth rates produced by the different models are compared to experimentally determined growth rates. The diffusion model reproduces the characteristic t1/2t1/2 growth for diffusion processes and therefore does not reproduce instability growth rates, as expected. The k–Lk–L and BHR turbulence models reproduce the nominal instability growth rates at multiple Atwood numbers with a single set of model parameters. The multifluid interpenetration model exhibits diffusion-like behavior and therefore does not reproduce instability growth rates. All four models exhibit Cauchy-like convergence in the mixing layer width with decreasing mesh size, although the multifluid model exhibits both a larger error for a given mesh size and a slower convergence rate than the turbulence models.

► Four mix models have been compared on a Rayleigh–Taylor instability problem.
► The models span a range of physical processes from diffusion to turbulent mixing.
► The results suggest that multiple processes may drive mix in RT-unstable state.