- Author
-
Frankel, S. H.
- Title
- Probabilistic and Deterministic Description of Turbulent Flows With Nonpremixed Reactants.
- Coporate
- State University of New York, Buffalo
- Sponsor
- National Aeronautics and Space Administration, Langley Research Center, Hampton, VA
Office of Naval Research, Washington, DC
National Science Foundation, Washington, DC
- Report
-
Thesis,
June 1993,
295 p.
- Contract
- NASA-CONTRACT-NAG-1-1122
N00014-90-J-4013
NSF-GRANT-CTS-9012832
NSF-GRANT-CTS-9253488
- Keywords
-
turbulent flow
|
homogeneous turbulence
|
isotropic turbulence
- Identifiers
- description of the DNS computational tools; reactant conversion in homogeneous turbulence; modeling of isotropic reacting turbulence by a hybrid AMC-EDQNM closure; Johnson-Edgeworth Translation (JET) of binary mixing in isotropic turbulence; comparative assessment of closures for turbulent reacting flows; modeling of reactant conversion in a turbulent shear flow; LEM of binary scalar mixing and reaction in homogeneous turbulent flow; LEM of reactant conversion and selectivity; large eddy simulation of turbulent reacting flow
- Abstract
- Presented herein is a mathematical and computational study of turbulent reacting flows. This work is concerned with four main approaches to dealing with mixing and reaction in turbulent flows. Namely, Direct Numerical Simulation (DNS), probability modeling, Linear Eddy Modeling (LEM) and Large Eddy Simulation (LES). The DNS results discussed are predominantly used for validation studies throughout this report. These involve pseudo-spectral, spectral/finite-difference and spectral/finite-element numerical techniques and are applied to both homogeneous and non-homogeneous flows. The models discussed and employed in the probability modeling include the Amplitude Mapping Closure (AMC) of Kraichnan, a hybrid Mapping-EDQNM closure, assumed PDF methods and the method of Johnson-Edgeworth Translation (JET). Comparisons for the problem of homogeneous mixing of two initially segregated scalars undergoing fast chemistry are made with both DNS and available experimental data. The applicability of some of the PDF models is also assessed for a non-homogeneous flow situation, again employing DNS data for validation purposes. LEM is employed to study mixing and reaction in a homogeneous turbulent flow. Finally, LES are conducted using a proposed hybrid one-equation Smagorinsky/PDF subgrid scale closure model as applied to a two dimensional, incompressible, spatially developing mixing layer.