AbstractsPhysics

Statistically steady measurements of Rayleigh-Taylor mixing in a gas channel

by Arindam Banerjee




Institution: Texas A&M University
Department:
Year: 2006
Keywords: Turbulent Flows
Record ID: 1773558
Full text PDF: http://hdl.handle.net/1969.1/4183


Abstract

A novel gas channel experiment was constructed to study the development of high Atwood number Rayleigh-Taylor mixing. Two gas streams, one containing air and the other containing helium-air mixture, flow parallel to each other separated by a thin splitter plate. The streams meet at the end of a splitter plate leading to the formation of an unstable interface and of buoyancy driven mixing. This buoyancy driven mixing experiment allows for long data collection times, short transients and was statistically steady. The facility was designed to be capable of large Atwood number studies of ABtB ~ 0.75. We describe work to measure the self similar evolution of mixing at density differences corresponding to 0.035 < ABtB < 0.25. Diagnostics include a constant temperature hot-wire anemometer, and high resolution digital image analysis. The hot-wire probe gives velocity, density and velocity-density statistics of the mixing layer. Two different multi-position single-wire techniques were used to measure the velocity fluctuations in three mutually perpendicular directions. Analysis of the measured data was used to explain the mixing as it develops to a self-similar regime in this flow. These measurements are to our knowledge, the first use of hot-wire anemometry in the Rayleigh-Taylor community. Since the measurement involved extensive calibration of the probes in a binary gas mixture of air and helium, a new convective heat transfer correlation was formulated to account for variable-density low Reynolds number flows past a heated cylinder. In addition to the hot-wire measurements, a digital image analysis procedure was used to characterize various properties of the flow and also to validate the hot-wire measurements. A test of statistical convergence was performed and the study revealed that the statistical convergence was a direct consequence of the number of different large three-dimensional structures that were averaged over the duration of the run.