Gas Mixing and Final Mixture Composition Control in Simple Geometry Micro-mixers via DSMC Analysis

Abstract

The mixing process of two pressure driven steady-state rarefied gas streams flowing between two parallel plates was investigated via DSMC (Direct Simulation Monte Carlo) for different combinations of gases. The distance from the inlet, where the associated relative density difference of each species is minimized and the associated mixture homogeneity is optimized, is the so-called mixing length. In general, gas mixing progressed very rapidly. The type of gas surface interaction was clearly the most important parameter affecting gas mixing. As the reflection became more specular, the mixing length significantly increased. The mixing lengths of the HS (hard sphere) and VHS (variable hard sphere) collision models were higher than those of the VSS (variable soft sphere) model, while the corresponding relative density differences were negligible. In addition, the molecular mass ratio of the two components had a minor effect on the mixing length and a more important effect on the relative density difference. The mixture became less homogenous as the molecular mass ratio reduced. Finally, varying the channel length and/or the wall temperature had a minor effect. Furthermore, it was proposed to control the output mixture composition by adding in the mixing zone, the so-called splitter, separating the downstream flow into two outlet mainstreams. Based on intensive simulation data with the splitter, simple approximate expressions were derived, capable of providing, once the desired outlet mixture composition was specified, the correct position of the splitter, without performing time consuming simulations. The mixing analysis performed and the proposed approach for controlling gas mixing may support corresponding experimental work, as well as the design of gas micro-mixers.

The Correctness of the Simplified Bernoulli Trial (SBT) Collision Scheme of Calculations of Two-Dimensional Flows

“…The results obtained using the SBT and SBT-CDF collisional schemes were compared to the results in [ 29 ]. The number of each species at the ( I , J ) cell is denoted as , with a = 1, 2.…”

“…The number of each species at the ( I , J ) cell is denoted as , with a = 1, 2. It is compared relative to the number density difference defined in [ 29 ] as where is the average number density of species a = 1, 2 of cells in column I , given by …”

A direct simulation Monte Carlo approach on the Riemann problem for gas mixtures

“…Very common working fluids are gases and so far single gas flows in micro scale have been extensively studied [5][6][7][8][9][10][11][12]. However, only recently the focus has shifted to gas mixtures and the process of gas mixing that are frequently met in real applications [13][14][15][16][17][18][19][20]. MEMS refer to devices with characteristic lengths starting as low as 1μm up to 1mm, therefore very often the characteristic length of the problem in study is critically small, comparable to the mean free path of the gas molecules.…”

Thermally Developing Flow and Heat Transfer in Elliptical Minichannels with Constant Wall Temperature

“…Microchannels are a fundamental part of microfluidic systems. A number of researches have been conducted to understand the mechanism of fluid flow and heat transfer characteristics in microchannels with circular [1,2,3], parallel plates [4,5,6], annular [7,8,9], rectangular [10,11,12,13,14,15] and trapezoidal [16,17,18] cross-sections. Compared with a circular duct, a better thermal efficiency can be found in an elliptical passage because the perimeter of an elliptical configuration is wider than the other for the same cross-sectional area [19].…”