Frieler, Clifford Eugene (1992) Mixing and reaction in the subsonic 2-D turbulent free shear layer. Dissertation (Ph.D.), California Institute of Technology. http://resolver.caltech.edu/CaltechETD:etd-03292005-135259
Several aspects of mixing and reaction in a turbulent two-dimensional shear layer have been studied. Experiments have been performed with reacting H2, F2, and NO in inert diluent gases. Sensing the heat release by these reactions, several aspects of the mixing process can be examined without the usual resolution limitations. For example, in contrast with direct measurements of composition, the amount of mixed fluid can be conservatively estimated with the results of the "flip" experiments. These have been performed over a range of density ratios, Reynolds numbers and heat release. The effects of initial conditions are of primary importance when comparisons to other studies are undertaken. Aspects as fundamental as growth rate of the turbulent region, or as obscure as the mixed fluid flux ratio depend strongly on the boundary conditions of this flow. These effects are examined in conjunction with those of Reynolds number and density ratio. For most cases studied here, tripping of the high speed boundary layer led to growth rate decreases. An exception was found for the case of high density ratio where the opposite effect was observed. This anomalous result occurred at conditions under which a new mode of instability has been shown to exist. Parallels exist between this unusual result and those of Batt in the uniform density case. An extensive study of the effects of density ratio on the mixing and reaction in the 2-D shear layer has been performed. Results indicate that several aspects of the mixing process are remarkably similar. Profiles of mixed fluid change little as the density ratio varies by a factor of 30. The integral amount of mixed fluid varies less than 6% for all density ratios examined. This insensitivity contrasts with that of the profiles of mixed fluid composition. While having very similar shapes the profiles are offset by an amount which depends very strongly upon the density ratio. The entrainment into the mixing layer has also been examined. Power spectral densities of the temperature time series were calculated and found to collapse upon normalization with the adiabatic flame temperature and large structure passage frequency. Least squares fits of the probability density functions were also examined. The initial work of Mungal and Frieler (1988) on the effects of chemical kinetics on the formation of product in the 2-D mixing layer have been greatly expanded. Measurements have been extended to include a wider range of NO concentrations and have been performed for two other stoichiometries. Results indicate that the simple model envisioned in Mungal and Frieler may only be suited for cases with extreme stoichiometry (very high or very low). Further investigations have turned up a serious discrepancy reflecting both on the experimental technique and on theory and modeling of this reacting flow. Experiments run under otherwise identical conditions demonstrate that more product is formed when F2 is the rich reactant than when H2 is the rich reactant. This dependence upon molecular character is counter intuitive and stems from a coupling of the effects of differing diffusivity and chemical kinetics. Numerical calculations based on simplified flow models are reported which demonstrate this coupling. These results indicate that even subtle diffusion effects can measurably effect reacting flows and imply that assumptions common among current modeling efforts must be re-examined. The effects of Reynolds number on mixing and reaction in the 2-D turbulent mixing layer have been examined. Evidence of the remnants of the initial roll up and mixing transition are seen for Reynolds numbers as large as 30,000. Indications of a resonance with the acoustic mode of the apparatus exist which affect results for Reynolds numbers up to 60,000. Natural transition of the high and low speed boundary layer on the splitter plate complicate comparisons of the high Reynolds number data with the remainder. In spite of all of these qualifications, the amount of mixed fluid is nearly constant. Over the range of Reynolds numbers 10,000 to 200,000, it varies by less than 12%. No evidence of an asymptotic decline in the amount of mixed fluid is observed.
|Item Type:||Thesis (Dissertation (Ph.D.))|
|Degree Grantor:||California Institute of Technology|
|Division:||Engineering and Applied Science|
|Thesis Availability:||Restricted to Caltech community only|
|Defense Date:||26 May 1989|
|Default Usage Policy:||No commercial reproduction, distribution, display or performance rights in this work are provided.|
|Deposited By:||Imported from ETD-db|
|Deposited On:||29 Mar 2005|
|Last Modified:||04 Mar 2014 20:07|
- Final Version
Restricted to Caltech community only
See Usage Policy.
Repository Staff Only: item control page