Date of Award
Summer 8-17-2017
Degree Name
Master of Science (MS)
Degree Type
Thesis
Abstract
The goal of this thesis is to compare the performance of several eddy-viscosity turbulence models for computing supersonic nozzle exhaust flows. These flows are of relevance in the development of future supersonic transport airplane. Flow simulations of exhaust flows from three supersonic nozzles are computed using ANSYS Fluent. Simulation results are compared to experimental data to assess the performance of various one- and two-equation turbulence models for accurately predicting the supersonic plume flow. One particular turbulence model of interest is the Wray-Agarwal (WA) turbulence model. This is a neat model which has demonstrated promising results mimicking the strength of two equation k-ω model while being a one equation model. Compressibility corrections are implemented for CFD simulations with SST k-ω, k-ε and low Reynolds versions of k-ε models which improved the results compared to the baseline models without compressibility correction. A compressibility correction for WA model is also developed to compare the performance of a compressibility correction to WA model with the compressibility correction to other models. Results show that the standard eddy-viscosity models can capture the shock structure and shear layer of the plume accurately when the thickness of the shear layer is small compared to plume diameter. However, when thickness of the shear layer is relatively large, a compressibility correction should be implemented to predict the supersonic jet flow. However, the use of compressibility correction consistently overestimates the length of potential core on the centerline of the plume although it improves the prediction of the velocity profile in other regions of the flow field such as the mixing region. Also, it is speculated that an accurate prediction of boundary layer profile at the nozzle exit has an influence in the model’s ability to predict the length of potential core as well as the shear layer growth rate. No single model appears to capture all features of the plumes’ flow fields without or with compressibility correction. In particular, WA model shows an excellent potential for computation of supersonic nozzles’ exhaust flows; however further improvements and investigations in WA model are warranted.
Language
English (en)
Chair
Ramesh K. Agarwal
Committee Members
Qiulin Qu Mark Meacham
Comments
Permanent URL: https://doi.org/10.7936/K7Z31Z2T