Implementation of Breakup and Coalescence Models in CFD

1
Implementation of Breakup and Coalescence Models
in CFD

• Peng Chen,
• Chemical Reaction Engineering Laboratory
• Washington University
• St. Louis, MO 63130
• CREL Group Meeting
• December 14, 2000

2
Overview

• Introduction
• Bubble number density approach
• Result
• Future work

3
Introduction

• The drag force term in one of the key issue of
  two-fluid model and the predictive capabilities
  of this model depend crucially on the closure.
• Most numerical simulation resorts to single
  particle drag correlation with a mean bubble
  size. This is mostly because modeling different
  sizes of bubbles as individual phase lead to
  unrealistic computational cost and has numerical
  convergence problem.
• However, in reality, bubble-bubble interaction
  results in local variation in bubble sizes that
  are substantially different from the mean
  bubble size assumption.
• In order to get reasonably good simulation
  result, the mean bubble size need to be
  adjusted so that it could be far away reality.

4
Motivation

• Bubble size distribution should be resolved
  locally by implement bubble coalescence and
  breakup into CFD framework.
• There is rare implementation of such bubble
  breakup/coalescence model in CFD simulation of
  bubble column reactors.
• Bubble breakup (Martínez-Bazán, 1999) and
  coalescence (Luo, 1993) model was implemented
  using bubble number density approach into FLUENT
  5.48 in the context of Algebraic Slip Mixture
  Model (ASMM).

5
Bubble Number Density Approach

• The population balance equation for the ith
  bubble class
• The source term may be written as

6
Closures - Breakup

• Martínez-Bazán et al. (1999)

7
Closure - Coalescence
Luo (1993)
, ?ij di/dj,
8

• Geometric Grid was used, xi1 2vi. x0 1mm.

Source term for particles of size xi is
9
(No Transcript)
10
Result
11
Overall 17.5 Experiment, 19.
12
Class 4 d 2.5 mm Class 6 d 4.0 mm
13
Class 12 d 16.0 mm Class 14 d 25.4 mm
14
Future Work

• Tune up the parameters, try to got a universal
  one or some sort of correlation.
• Test the sensitivity of boundary condition
• Test other breakup/coalescence closure
• Implement area transport equation approach
• Run 3D simulation
• Expand to Euler-Euler two fluid model
• With optical probe data, verify the closures and
  propose own closures