No abstract

Although the role of gas purging in liquid steel systems is well recognized, it has yet to be adequately analyzed. One key aspect of this process is the prediction of gas voidage in the bath, which has been studied in great detail beginning with water modeling in the early days and using advanced multiphase models more recently. Still, there are significant unresolved issues with gas purging systems. When gas is introduced through a nozzle at high flow rate, a jet may form which is undesirable. The break-up of this jet into bubbles is a separate topic of research. The more common practice in the steel industry is to use porous plugs for gas injection. Gas entry through a porous plug can be characterized by the stretched bubble regime, and the laws of coalescence and fragmentation used to analyze bubble column reactors are generally applicable. Calculation of the bubble size distribution is important for two reasons. First, the voidage distribution in the bath is significantly modified by the injection system and flow rates used, primarily due to changes in flow regime and bubble dynamics (collision, break-up, coalescence). Second, the voidage distribution directly determines the buoyancy, that influences the physical mixing process, and the specific-area-density, that influences surface reactions (for example, decarburization, desulfurization and nitrogen pick-up). In this paper, a numerical study is presented that combines a bubble dynamics model with an Eulerian multiphase model. The results of the simulation are compared with the experimental data from Anagbo and Brimacombe (1990). Relevant discussion and reviews will be presented to distinguish the differences of this detailed bubble dynamics model with the uniform bubble diameter approximations reported in various recent studies.

Fluent version 6.2 computational fluid dynamics environment has been enhanced with a population balance capability that operates in conjunction with its multiphase calculations to predict the particle size distribution within the flow field. The population balance is solved by the quadrature method of moments (QMOM). Fluent's prediction capabilities are tested by using a 2-dimensional analogy of a constantly stirred tank reactor with a fluid flow compartment that mixes the fluid quickly and efficiently using wall movement and has a feed stream and a product stream. The results of these Fluent simulations using QMOM population balance solver are compared to steady state analytical solutions for the population balance in a stirred tank where 1) growth, 2) aggregation, and 3) breakage, take place separately and 4) combined nucleation and growth and 5) combined nucleation, growth and aggregation take place. The results of these comparisons show that the moments of the population balance are accurately predicted for nucleation, growth, aggregation and breakage when the flow field is turbulent. With laminar flow the mixing is not ideal and as a result the steady state well mixed solutions are not accurately simulated.