Multiscale Issues in DNS of Multiphase Flows

Author(s):  
Bahman Aboulhasanzadeh ◽  
Siju Thomas ◽  
Jiacai Lu ◽  
Gretar Tryggvason

In direct numerical simulations (DNS) of multiphase flows it is frequently found that features much smaller than the “dominant” flow scales emerge. Those features consist of thin films, filaments, drops, and boundary layers, and usually surface tension is strong so the geometry is simple. Inertia effects are also small so the flow is simple and often there is a clear separation of scales between those features and the rest of the flow. Thus it is often possible to describe the evolution of this flow by analytical models. Here we discuss two examples of the use of analytical models to account for small-scale features in DNS of multiphase flows. For the flow in the film beneath a drop sliding down a sloping wall we capture the evolution of films that are too thin to be accurately resolved using a grid that is sufficient for the rest of the flow by a thin film model. The other example is the mass transfer from a gas bubbly rising in a liquid. Since diffusion of mass is much slower than the diffusion of momentum, the mass transfer boundary layer is very thin and can be captured by a simple boundary layer model.

2005 ◽  
Vol 51 (6-7) ◽  
pp. 69-76 ◽  
Author(s):  
S.R. Smith ◽  
T. Taha ◽  
Z.F. Cui

Tubular membrane ultrafiltration and microfiltration are important industrial separation and concentration processes. Process optimisation requires reduction of membrane build-up. Gas slug introduction has been shown to be a useful approach for flux enhancement. However, process quantification is required for design and optimisation. In this work we employ a non-porous wall CFD model to quantify hydrodynamics in the two-phase slug flow process. Mass transfer is subsequently quantified from wall shear stress, which was determined from the CFD. The mass transfer model is an improved one-dimensional boundary layer model, which empirically incorporates effects of wall suction and analytically includes edge effects for circular conduits. Predicted shear stress profiles are in agreement with experimental results and flux estimates prove more reliable than that from previous models. Previous models ignored suction effects and employed less rigorous fluid property inclusion, which ultimately led to under-predictive flux estimates. The presented model offers reliable process design and optimisation criteria for gas-sparged tubular membrane ultrafiltration.


1998 ◽  
Vol 37 (4-5) ◽  
pp. 139-147 ◽  
Author(s):  
Harald Horn ◽  
Dietmar C. Hempel

The use of microelectrodes in biofilm research allows a better understanding of intrinsic biofilm processes. Little is known about mass transfer and substrate utilization in the boundary layer of biofilm systems. One possible description of mass transfer can be obtained by mass transfer coefficients, both on the basis of the stagnant film theory or with the Sherwood number. This approach is rather formal and not quite correct when the heterogeneity of the biofilm surface structure is taken into account. It could be shown that substrate loading is a major factor in the description of the development of the density. On the other hand, the time axis is an important factor which has to be considered when concentration profiles in biofilm systems are discussed. Finally, hydrodynamic conditions become important for the development of the biofilm surface when the Reynolds number increases above the range of 3000-4000.


Sign in / Sign up

Export Citation Format

Share Document