Cavitating Herschel Venturi nozzle test rig

Heiko Warnecke, Dep. 1.53 Liquid Flow

� Flow rate control device� Proportioning of liquids� Flow meter

introduction | characterisation | different liquids | load profiles | conclusion 2 Heiko Warnecke

introduction

� Reproducibility� Flow rate stability� Fast reaction time� Small � Cheap� …

introduction

introduction | characterisation | different liquids | load profiles | conclusion 3 Heiko Warnecke

� Influence of numerous factors on thecavitation

• Flow cross-section

• Liquid

• Reynolds-number

• Upstream and downstreampressure cylindric?

diameter

surface roughness

axial?

Nozzle throat

d

1,9 x d

edge3,3 x d

10 x d

set-up

introduction | characterisation | different liquids | load profiles | conclusion 4 Heiko Warnecke

� 6 nozzles with a nominal diameter of0.74 mm, 1.4 mm and 2.6 mm arecharacterised

� flow rates between 37 l/h and 588 l/h

characterisation

introduction | characterisation | different liquids | load profiles | conclusion 5 Heiko Warnecke

Nozzle nr. i

DiameterD

mm± 1 µm

Flow Q (4 bar)

l/h± 0.1 %

Slopeb

l/h/bar

1 0.740 42.69 5.4 ± 0.2

2 0.738 42.65 5.4 ± 0.2

3 1.396 151.80 19.2 ± 0.4

4 1.397 152.40 19.5 ± 0.5

5 1.393 154.50 19.5 ± 0.5

6 2.601 527.80 71.0 ± 1.4

characterisation

introduction | characterisation | different liquids | load profiles | conclusion 6 Heiko Warnecke

3 4 5

40

60

80

100

120

140

160

180

nozzle 1 nozzle 2 nozzle 3 nozzle 4 nozzle 5

flow

/ l/h

upstream pressure / bar

2 3 4 5300

400

500

600

flow

/ l/h

upstream pressure / bar

nozzle 6 linear regression

equation y = a + b*x

a 238.45

b 71.02

r² 0.99776� Assumption of a linear

correlation between flow rate and upstream pressure

Nozzle nr. i

DiameterD

mm± 1 µm

Flow Q (4 bar)

l/h± 0.1 %

Slopeb

l/h/bar

1 0.740 42.69 5.4 ± 0.2

2 0.738 42.65 5.4 ± 0.2

3 1.396 151.80 19.2 ± 0.4

4 1.397 152.40 19.5 ± 0.5

5 1.393 154.50 19.5 ± 0.5

6 2.601 527.80 71.0 ± 1.4

characterisation

introduction | characterisation | different liquids | load profiles | conclusion 7 Heiko Warnecke

� Different nozzles with same nominal diameter

� Significant difference of theflow rates

additivity

introduction | characterisation | different liquids | load profiles | conclusion 8 Heiko Warnecke

0 200 400 600 800 1000 1200

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

rel.

dev.

/ %

flow / l/h

���� =�����

+ � − � ∗ ��) Qi: single nozzle flow rate

bi: calculated slope of each nozzle

pi: upstream pressure of single nozzle flow

pj: upstream pressure of multiple nozzle flow���. ���. = ���� −��������

different liquids

introduction | characterisation | different liquids | load profiles | conclusion 9 Heiko Warnecke

Medium Principalconstituents

Density

g/cm3

Vapourpressurembar

Viscosity

mm2/sWhite spirit Hydrocarbon,

C10 – C13,n-Alkane,iso-Alkane,cycliccompound

0.785 0.5 1.2

Haku 1025-310

Hydrocarbon,C11 – C14,iso-Alkane,cycliccompound

0.761 0.6 1.3

Water Tap water 0.998 23.4 1.0

physical properties density, vapour pressure and kinematic viscosity at 20°C

different liquids

introduction | characterisation | different liquids | load profiles | conclusion 10 Heiko Warnecke

⇒Similar flow rates of white spirit and cleaner solvent⇒The flow rate of water is higher at the same

upstream pressure⇒Comparable standard deviation for different liquids

different liquids

introduction | characterisation | different liquids | load profiles | conclusion 11 Heiko Warnecke

3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

30

33

36

39

42

45

48

51

mas

s flo

w /

kg/h

upstream pressure / bar

water white spirit 180/210 Haku 1025-310

d = 0.74 mmd = 1.4 mm

Medium Standard deviation

White spirit 0.097 %

Haku 0.018 %Water 0.023 %

load profiles

introduction | characterisation | different liquids | load profiles | conclusion 12 Heiko Warnecke

Load profiles of a typical german household are generated

⇒How can a load profile be realised in a test rig?

load profile

introduction | characterisation | different liquids | load profiles | conclusion 13 Heiko Warnecke

Apartment house

Example

[Schumann et al. 2017]

New characterisation for water meters close to real world conditions

load profile

introduction | characterisation | different liquids | load profiles | conclusion 14 Heiko Warnecke

� Flow rates at constant upstream pressure� pressure ratio smaller than 0.6

load profile section

introduction | characterisation | different liquids | load profiles | conclusion 15 Heiko Warnecke

⇒pressure excursions of 7% for about 1 s⇒No effect seen on the balance signal

� Flow rates of the individual nozzles sum up to theexpected flow rates

� A constant mass flow can be generated withdifferent liquids, including liquid mixtures

� With the nozzles a load profiles can be generatedwith fast flow rate changes

conclusion

introduction | characterisation | different liquids | load profiles | conclusion 16 Heiko Warnecke

� Characterisation and verification of liquid properties (vapour pressure, viscosity, etc)

� Characterisation of nozzle properties(surface roughness, edges, etc)

� Variation of measurement conditions(pressure, temperature, etc)

outlook

introduction | characterisation | different liquids | load profiles | conclusion 17 Heiko Warnecke

Physikalisch-Technische BundesanstaltBraunschweig and BerlinBundesallee 10038116 Braunschweig

Heiko WarneckeTelefon:0531 592-1389E-Mail: heiko.warnecke@ptb.dewww.ptb.de

Stand: 06/2019

⇒Different flow rates of water and the two other liquids at the same pressure

introduction | characterisation | different liquids | load profiles | conclusion 20 Heiko Warnecke

• Nominal diameter 1.4 mm (left) for water and cleaner solvent and 2.6 mm (right) for water and white spirit

3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0440

480

520

560

600

640

mas

s flo

w /

kg/h

upstream pressure / bar

water white spirit

Different liquids