Date Awarded


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Virginia Institute of Marine Science


In recent years, different approaches for modeling Reynolds stresses have been pursued by various authors. Two different closure schemes, representing two schools of thought, have been selected. The first closure scheme is the k-&\varepsilon& model. The second closure scheme is the level two model. Each group has claimed that the proposed models are an improvement and an advancement over previous models employing the mixing length hypothesis as a closure scheme. Neither group however, has made thorough comparisons between the newly proposed schemes and the most commonly used approach the mixing length hypothesis. The main objective of this paper is to test the applicability of the standard k-&\varepsilon& and the level two models in buoyant and non-buoyant flows. In addition, comparisons have been made between these new closure schemes and an existing mixing length model. Results obtained by employing the level two model for buoyant and non-buoyant flows are considered to be an improvement over those obtained by employing the mixing length model. A single set of constants was used for all level two model applications. Unlike the mixing-length closure scheme which requires tuning of the constants employed by the model for each individual application, there was no further tuning required for the level two model. Comparison between model results has shown no substantial improvement of the k-&\varepsilon& model over the level two model.



© The Author