Date of Award

January 2017

Document Type


Degree Name

Master of Science (MS)


Mechanical Engineering

First Advisor

Clement Tang


It is well recognized in continuum mechanics that the theoretical origination for single-phase flows fundamentally based on the beginning of field equations paraphrasing the conservation laws of mass, momentum, and energy. These equations are accompanied by the appropriate constitutive associations for thermodynamic state and energy transfer, which specify the thermodynamic and transport of a specific fundamental material. Nevertheless, the multi-phase flows equations derivation is, in general, considerably more convoluted than for single-phase flows. This is predominantly due to the presence of significant discontinuities of the fluid properties through the interfaces and complicated flow characteristics approximating the interfaces, separating the individual phases that co-exist within the flow, and multiple, deformable, and moving interfaces.

The purpose of this study is to use computational and experimental approach to understand the gas-liquid two-phase flow, and single-phase flow behavior in channels with sudden contraction/expansion. With the advances in computing capabilities and resources, researchers have continued to reevaluate the reliance of computational analysis and predictions to better understand the transport phenomena, and intricacies in multi-phase flows. Reliable computational approach can often be a more cost-effective tool than an experimental approach that requires accurate sensors and instrumentation. Four test sections with different sudden contraction and expansion in the cross-sectional areas were used in this study. The diameter ratios for the sudden contraction/expansion are 1.33, 1.59, 2.63, and 3.57. The range of flow rates are, for liquid, from 5 to 30 g/s (0.005 kg/s to 0.03 kg/s), and for gas, from 0.49 to 29 g/s (0.00049 kg/s to 0.028 kg/s).

using gauge pressure values form experimental setup, measured and analyzed data, and its assistance to validate computationally modelled data with detailed visualization of pressure profile, is a one of principle topic of study. Along with it, pressure drop data collected from experimental analysis, and computationally acquired values, and validation between them. This is too prove that, the current computational model can be utilized for complex multiphase flow systems in industries. After validation, another goal of study is to generate pressure profiles, and local velocity profiles and study of their visualization to propose physics behind it.

The computational study for the two-phase flow used an Eulerian-Eulerian multiphase approach and the Reynolds stress turbulence model for two-phase gas-liquid flow with input from the experimental data for boundary conditions of solver. Prior to the two-phase and single-phase flow, grid independence study, and turbulence study is carried out. Turbulence study shows Reynold’s turbulence model provides more accuracy than that of the k-ϵ model for higher flow rates. The optimized grid is implemented with the Reynolds stress dispersed multiphase flow turbulence model. Pressure drop along the channels of different area ratios was observed to be influence by the Reynolds number, along with that it is found to be directly proportional to pressure drop for channel. The Reynolds number calculated in the turbulence analysis is found to be 400 to 10000 as the pressure drop value increases and flow rate increases, for the area ratio of 0.1444. The pressure drop values are in range from 0.3 to 10 kPa, for the area ratio with 0.1444. While, the pressure drop values ranges from 0.0146 to 0.8 kPa for area ratio of 0.5625. This also proves that pressure drop is inversely proportional to hydraulic diameter. The after calculation of pressure profiles, the plotted pressure drop values show precise prediction of computational analysis and good agreement with experimental data with margin error of 1 to 11% for two-phase flows and 1 to 3% for single-phase flow for channels with smaller diameters.