Many chemical manufacturing processes are carried out in bubble column reactors in which gas bubbles are dispersed throughout a liquid to perform chemical reactions. These reactors are designed to contain a large number of small bubbles to increase the contact area between the gas and liquid phases. The dynamics of gas-liquid flow inside these reactors and in similar multiphase processes is often described using mathematical models called two-fluid models. This project focuses on improving two-fluid models by explicitly taking into account the effects of interactions among the bubbles on certain terms in the models. The investigators will evaluate these critical terms by using computational simulations of bubble motions that account for interactions among bubbles and the development of bubble assemblies in flow. The improved two-fluid model will then be used in computational fluid dynamic simulations to compare with experimental data and help validate the model. The results of the project will be useful to design and optimize a broad array of manufacturing processes such as fermentation, wastewater treatment, algae growth, and antibiotic production.
The accuracy of volume averaged two-fluid models for gas-liquid flow depends on terms that estimate momentum transfer between gas and liquid phases. These terms include several forces such as drag, lift and virtual mass forces. These forces are usually evaluated for the case of an isolated bubble in the flow. This project will study the effects of assemblies of interacting bubbles in the flow on the lift and virtual mass forces, and will use the results to develop improved momentum exchange terms for two-fluid models. The forces will be determined by conducting particle-resolved direct numerical simulations of spherical buoyant particles, which are a surrogate for bubbles in the limit where bubble deformation can be neglected. The simulations will be used to develop expressions for the lift and virtual mass forces for use in the two-fluid equations. The improved two-fluid models will then be used in computational fluid dynamics simulations over a wide range of operating conditions for which experimental data are available to validate the newly developed terms.
