**Titre/Title :**

Rayleigh-Taylor instability - a century of open questions

**Contact :**

Achim Wirth (équipe Meige)

**Résumé/Abstract :**

Rayleigh-Taylor instability is the transient buoyancy/inertia-driven process that occurs when more dense fluid overlies lighter fluid. Conceptually, the configuration could hardly be simpler, yet it has provided enough puzzlement for scientists to motivate over a century of research. Lord Rayleigh, in 1883, was one of the first to investigate the problem, but his experiment mislead him to the wrong conclusion. He had inadvertently observed a diffusion-driven instability which has turned out to be of equal significance. Independently, G.I.Taylor in 1950, constructed the first theoretical analysis of the buoyancy/inertia problem that we now call after both Rayleigh and Taylor. However, the linear model proposed by Taylor is only valid for very short time-scales immediately after the instability is initiated, since the flow quickly becomes turbulent. For the following half-century, scientists have argued about the development of the non-linear phase of Rayleigh-Taylor instability, and have obtained remarkably few firm conclusions. This talk will instead concentrate on the late-time decay phase of Rayleigh-Taylor instability in finite domains, since in this regime the separation of length- and time-scales between the turbulence and the instability growth is sufficient to permit analysis of the growth without needing full knowledge of the turbulent field. A combination of experiments and simulation provide validation for some robust conclusions.