MODELING AND EXPERIMENTAL DATABASES ON POLY-DISPERSED BUBBLY FLOWS

Dirk Lucas, Eckhard Krepper,

Matthias Beyer, Lutz Szalinski

PRESENTATION TOPICS

• The Institute of Fluid Dynamics @ Helmholtz-Zentrum Dresden-Rossendorf;

• Motivation & background;

• Poly-dispersed bubbly flows

• Homogeneous and Inhomogeneous MUSIG models;

• Experimental databases on vertical pipe flow;

• Conclusion and next steps.

Helmholtz-Zentrum Dresden-Rossendorf

• Member of the Helmholtz Association since January 1st, 2011- funding 10% local state of Saxony, 90% German state

• Basic and applied long term research in selected fields of energy, health, and materials sciences

• Employees: ~ 750

~ 350 scientists (including PhD students, postdocs)

• 7 Institutes

* Ion-Beam Physics and Materials Research

* Resource Ecology * Radiation Physics

* Fluid Dynamics * Dresden High Magnetic Field Laboratory

* Radiopharmacy * Helmholtz Institute Freiberg for Resource Technology

• Departments of the Institute of Fluid Dynamics

Exp. Thermal Fluid Dynamics CFD Magnetohydrodynamics

Qualification of CFD codes for two-phase flows • The qualification of CFD codes comprises model development and validation

• The activities are embedded in the German CFD initiative which aims on the qualification of CFD codes for future use in Nuclear Reactor Safety.

• For large scale industrial applications Two- or Multi-Fluid model requires closure models

• During the last decade the Inhomogeneous MUSIG-model for the simulation of poly-dispersed flows was developed at HZDR suitable closure models required need of experimental data with high resolution in space and time

Motivation & background

Poly-dispersed bubbly flows – Example: Upwards vertical pipe flow

Phenomena to be considered • Two-phase turbulence • Bubble forces

• drag • virtual mass • turbulent dispersion • wall • lateral lift

• Bubble coalescence & breakup Depend e.g. on:

• gradients in the liquid velocity • turbulence parameter

Bubbles influence the liquid flow field • velocity field • turbulence

LLGLLLIFT VrotVVCF

)(

• Transition wall peak – center peak at 5.5 ... 6 mm Lateral lift force!!!

Volume fraction profiles decomposed according to the bubble size

Homogeneous MUSIG model

• Multiple bubble size group model (MUSIG) S. Lo (1996 CFX-4):

- for the gaseous phase only one velocity field - only one momentum equation for the gaseous phase - consideration of bubble break-up and coalescence only in the continuity

equation

VG

d1 dM

bubble

coalescence

bubble

break-up

Gas velocity

Size fractions

K=1..M dk

Source: Krepper et al., NED 238(2008)1690-1702

Inhomogeneous MUSIG model

• Improvement: Inhomogeneous MUSIG Model Krepper et al. (2008):

• N velocities fields – allows separation of small and large bubbles • M size fraction groups in the mass balance • consideration of bubble break-up and coalescence in the continuity equation • Improved model for bubble coalescence and breakup by Liao et al., NED 241 (2011)

1024–1033 • Improved two-phase turbulence modelling

V1 V2 VN

d1 dM1 dM1+1 dM1+M2

bubble

coalescence

bubble

break-up

Velocity groups

J=1..N

Size fractions

K=1..SMJ

...

dSMJ

Source: Krepper et al., NED 238(2008)1690-1702

TOPFLOW: Transient Two Phase Flow Test Facility

Two-phase flow in vertical

pipe configurations:

• Wire-mesh sensor

• Fast X-ray tomography

Pressure tank:

steam-water flow

experiments at

pressure

equilibrium

Wire-mesh sensors

Measuring frequency up tp 10.000 frames/s

Figs. from http://www.hzdr.de/db/Cms?pOid=10412&pNid=1014

Wire-mesh sensors for 200 mm pipe

64 * 64 wires, 2500 frames/s

10 s measuring time

3D matrix 64*64*25000 of values for

conductivity

Calibration: matrix of

64*64*25.000 values for gas volume

fraction

Two sensors behind each other

Gas velocity from cross-correlation

Wire-mesh sensor measurements for 200 mm pipe

JL = 1.017 m/s

JG = 0.534 m/s

Test section: Variable Gas Injection – length for experiments: 8 m

Investigations on evolution of two-phase

flows along a 200 mm vertical pipe

Gas injection via holes in the pipe wall

6 injction devices each 3 injection

chambers, 1 mm and 4 mm orifices

Quantitative data from time averaging

radial gas volume

fraction profiles

radial gas velocity profiles

Bubble size distributions

Also available: Gas volume fraction decomposed

according to bubble size and radial position

Validation of closure models for Inhomogeneous MUSIG in CFX

0.000 0.020 0.040 0.060 0.080 0.100r [m]

0.00

0.10

0.20

0.30

0.40

0.50

G

[-]

A: 0.221 mexp

CFX (total)

dB<6 mm

dB>6 mm

0 10 20 30 40 50 60 70dB [mm]

0.0

0.5

1.0

1.5

2.0

2.5

d

G/d

dB [

%/m

m]

L12_118R: 7.802 m

exp

CFX

0 0.05 0.1 0.15 0.2 0.25

[-]

0.0

2.0

4.0

6.0

8.0

10.0

z [

m]

L12-118exp

compressible

incompressible

Level A

Level R

0 10 20 30 40 50 60 70dB [mm]

0.0

0.1

0.2

0.3

0.4

0.5

d

G/d

dB [%

/mm

]

L12 118A: 0.221 m

exp

CFX

0.000 0.020 0.040 0.060 0.080 0.100r [m]

0.00

0.10

0.20

0.30

0.40

0.50

G [-]

R: 7.802 mexp

CFX (total)

dB<6 mm

dB>6 mm

JL = 1.017 m/s

JG = 0.219 m/s

• Using the Inhomogeneous MUSIG approach, poly-dispersed flows can be modelled over some range of flow parameters

• Further improvements of closure models is necessary to extent the range of applicability

• Close connection between model development & validation and experiment is important

• Future experimental data: Ultrafast X-ray tomography (non-intrusive)

Conclusions and next steps

Fischer et al., Meas. Sci. Technol. 19(9), 2008

Acknowledgements

Thank you for your attention!

This work was carried out within

the frame work of research

projects funded by the German

Federal Ministry of Economics and

Technology, project numbers

150 1265 and 150 1329.