Gas-Liquid Columns: Experimental Characterization and CFD Simulations

1
Experimental Characterization of Gas-Liquid
Column Effect of nozzle orientation and
pressure by Peter Spicka
CREL group regular meeting, November 26th
2
Objective

• Study the sparger nozzle orientation effect on
  gas hold-up, liquid velocity and turbulence in
  gas-liquid column

Motivation

• Only few data, i.e. liquid velocity and gas
  holdup, available in the literature for
  churn-turbulent flow regime
• CARPT and CT techniques allow relatively accurate
  acquisition of needed data
• Different pressures and UGS can be covered
• Additional experimental database for CFD
  simulations can be created

3
Experiment

• 6.375 stainless steel column
• Cross-sparger, two nozzle orientations facing
  upward and downward
• Air-water system
• Dynamic height maintained at 11 D
• Pressure 1 bar and 4 bars
• UGS 5 cm/s (only CT) and 20 cm/s
• CARPT setup
• Typical setup, 30 detectors
• Only photo peak acquisition
• 50 Hz sampling frequency
• CT setup
• 5 detectors, 7 projections per view
• 4 axial levels 2.5D 3.5D 5.5D and 9D
• 20 Hz sampling frequency

4
Detector alignment and calibration
CT CARPT
5
CT Results Effect of Nozzle Orientation- Global
View
Gas holdup at UGS20 cm/s and p1 bar
nozzles facing downward nozzles facing
upward

• Bubbles formed from nozzles facing upward are
  smaller Þ increased hold-up
• Similar behavior was found for all the studied
  regimes

6
CT Results Effect of Nozzle Orientation and
Pressure
Gas holdup profiles UGS5 cm/s
UGS20 cm/s

• Nozzle orientation
• particularly pronounced at high pressure and high
  UGS in the sparger zone
• diminishes with axial position
• Pressure
• Typical increase of gas holdup magnitude

7
Axial velocity profiles Radial
velocity profiles
CARPT Results Liquid velocity

• Steeper velocity profiles observed at high
  pressure Þ higher bubble momentum
• However, effect of nozzle orientation on liquid
  velocity is visible only at near-sparger region

8
Liquid velocity calculation

• CARPT processing algorithm considers uniform time
  step
• However, relatively large amount of data is
  excluded from calculation (25 and more)

• Time step is non uniform
• Calculated velocity will be biased towards higher
  values

Comparison with Boon Chengs data
9
CARPT Results Turbulent kinetic energy
Nozzles facing down Nozzles facing up p 1
bar p 4 bars p 1 bar p
4 bars

• Turbulent kinetic energy is higher for nozzles
  pointing downward and at higher pressure
• Nozzle effect is significant mainly at low
  pressure
• Significant effect of bubble-induced turbulence

10
CARPT Results Reynolds stresses

• uxux are approximately 2.5 x higher than urur
  and they are weakly coupled
• Magnitude of uxux is comparable with the
  corresponding mean velocities
• Highly anisotropic flow !

11
Concluding Remarks

• Nozzle orientation
• Significant effect on gas holdup and turbulent
  kinetic energy mainly near the column bottom
• More pronounced at high UGS and high pressure
• Effect on liquid velocity profiles is less
  significant
• Uncertainty in magnitude of turbulent parameters
  due to gas holdup fluctuations

Outlook for future

• Filtering
• Elimination of gas holdup fluctuations from CARPT
  data
• CFD
• Examination of nozzle orientation effect in
  churn-turbulent regime