Date Author Address
October, 2003
Prof.dr.ir. T.J.C. van Terwisga Delft University of Technology Ship Hydromechanics Laboratory Mekelweg 2, 26282 CD Delft
Phone: +31 15 2786873
TUDelft
Delft University of Technology
Performance and cavitation analysis of a
Waterjet system on a cavitation tunnel
byT.J.C. van Terwisga
Report No. 1401-P
October 2003
FAST2003, The 7th International Conference on Fast Sea Transportation, Ischia, Italy
Università di Napoli "Federico II"
FAST 2003
Ischia (Italy)
7'
-
10' October 2003
The 7th International Conference
on
FAST SEA TRANSPORTATION
PAPERS (VoL 1)
Pasquale Cassella
Editor
Dipartimento Ingegneria Navale - Università di Napoli "Federico II"
Istituto di Navigazione "G. Simeon" - Università di Napoli "Parthenope"
"The Organizing Committee FAST 2003" shall not be responsible for statements
or opinions advanced in papers or printed in these volumes.
In order to make these proceedings as economically and rapidly as possible, the
authors's papers have been prepared for final reproduction and printing without any
reduction, correction, etc.
Therefore the authors are fully responsible for all the infon-nation contained in their
papers. The printing process has been performed in a standard way for all the papers
submitted.
ISBN No: 88-901174-0-0(Set)
ISBN No:88-901174-9-4 (Vol. 1)
Copyright © 2003
Comitato Organizzatore FAST 2003
All Right Reserved
Printed in Italy (September 2003) by
"ARTI GR.AFICHEZACCARIA SRL" Via Loggia dei Pisani, 15/19
Tel. 081 5512628 - Fax 081 5516924 E-mail: zaccariasrl@libero.it
These Proceeding,s consist of papers presented at the 7th International Conference on
FAST Sea Transportation. The Conference was held at the Hotel Continental Terme,
Ischia Porto (Italy), on 7-10 October 2003.
The Conference was Organised by:
Department of Naval Architecture and Marine Engineering
University of Naples "Federico II"
Institute of Navigation "G. Simeon"
University of Naples Parthenope
Under the patronage of:
Senato della Repubblica Italiana
Marina Militare Italiana
Universitä degli Studi di Napoli "Federico II"
Università degli Studi di Napoli "Parthenope"
Ordine degli Ingegneri della Provincia di Napoli
The Organizing Committee organized the Conference under supervision of the
International Scientific Committee. The Conference benefited from the generous
sup-port of a number of Sponsors.
We are grateful to:
Università. degli Studi di Napoli "Federico II"
Università degli Studi di Napoli "Parthenope"
FINCANTIERI S.p.A.
Registro Navale Italiano
Lloyd's Register of Shipping
American Bureau of Shipping
CETENA S.p.A.
Ordine degli Ingegneri della Provincia di
Napoli
SNAME (Society of Naval Architects and Marine Engineers)
VOLAVIAMARE
INTERNATIONAL COMMITTEE
Mr. K. Holden, MARINTEK A/S, Norway (Chairman)
Mr M. Bewick, The Harbor Consultancy Int. USA
Prof. L. Doctors, UNSW, Australia
Prof. O. Faltinsen, NTNU, Norway
Prof. G. Ferrara, University of Naples "Parthenope"
Mr. N. Gee, N. Gee and A. Ltd. UK
Prof. T. Moan, NTNU, Norway
Mr. J.A. Moret, IZAR Construcciones Navales, Spain.
NATIONAL COMMITTEE
Dr. Ing. C. Antonini
,FINCANTIERI S.p.A
Ing. G. Bernardi, CETENA S.p.A.
Dr. U. Bulgarelli, INSEAN (Rome Towing Tank)
Prof. A. Campanile, University of Naples, "Federico II"
Prof. A. Cardo, University of Trieste
Ing. G. De Domenico, Tugboats Company of Naples
On. S. Lauro, Shipping-Line "Alilauro"
Prof. L. Nicolais, University "Federico II" of Naples
Adm. Gen. Isp. (G.N.) E. Piantini, Italian Navy
Prof. C. Podenzana Bonvino,University of Genoa
Adm. Gen.Isp. (C.P.) E. Sicurezza, Italian Coast Guard
Ing. N. Squassafichi Nicola, RINA (Italian Register of Shipping)
ORGANISING COMMITTEE
Prof. P. Cassella (Chairman), University of Naples, "Federico II"
Prof. R. Balestrieri, University of Naples "Parthenope"
Dr. C. Bertorello University of Naples, "Federico II"
Prof. G. Russo Krauss, University of Naples, "Federico II"
Prof. A. Scamardella, University of Naples "Parthenope"
FAST 2003 Secretariat
Dr. Carlo Bertorello
Mr. Pasquale Cioffi
Department of Naval Architecture and Marine Engineering
University of Naples "Federico II"
Via Claudio 21 80125, Naples (ITALY)
Tel. +39-081-7683310 /3714
Fax +39-081-2390380
fast2003@unina.it
Table of Contents
Vol. 1
Preface
Opening Lectures
The Fast Ship Fever
From Trondheim To Ischia,The Gulf Of Naples.
1 Kjell O. HoldenHigh Speed Craft - Regulations And Efforts To Minimize Accidents
5Eugenio Sicurezza
Keynote Lectures
The Optimisation Of Trimaran Sidehull Position For Minimum Resistance
1Lawrence J Doctors, Robert J Scrace
Hydrodynamic Aspects Of High-Speed Vessels
13Faltinsen 0.M, Claudio Lugni, Maurizio Landrini
The Effects Of Length On The Powering Of Large Slender Hull Forms
23Nigel Gee, James Roy
Towards Structural Design Of High Speed Craft Based On Direct Calculations
31 Torgeir MoanSimulation Based Resistance And Seakeeping Performamce Of High-Speed Monohull And Multihull
Vessels Equipped With Motion Control Lifting Appendages
51Paul D. Sclavounos, Shiran Pun;in, Ta/ha Ulusoy, Sungetin Kim
Developments On A Probabilistic Risk/Cost Model For Large-Scale Flooding Consequence Analysis
Of High Speed Monohulls
67Dimitris Konovessis, Daria Cabaj, Dracos Vassalos
A New Generation Of Large Fast Ferry - From Concept To Contract Reality
75Tony Armstrong, Kjell Holden
Session Al Hydrodynamic Performance
From Model Scale To Full Size. Investigation On Turbulence Stimulation In Resistance Model Tests
Of High Speed Craft
1
Carlo Bertorello (invited speaker,) Dario Bruzzone, Sebastiano Caldarella ,Pasquale Cassella, Igor Zotti
Hydrodynamic Improvements Of Catamaran Hulls When Using Streamlined Bodies Of Revolution
9Igor Zotti (invited speaker)
Transom Hollow Prediction For High-Speed Displacement
Vessels 19Simon W Roba rds, Lawrence J Doctors
Series 64 Parent Hull Displacement And Static Trim
Variation 27Gabor Karafiath, Todd Carrico
Hullform Analysis And Optimisation For A Fast Ropax Ferry
31II
A Critical Survey On Performance Assessment For Classical And Advanced Ships
39 Luigi IannoneThe Effect Of Demihull Separation On The Frictional Resistance Of Catamarans
47Tony Armstrong
Performance And Cavitation Analysis Of A Waterjet System On A Cavitation Tunnel
57Klaas Kooiker Tom Van Terwisga, Rob Verbeek, Peter Van Terwisga
The Accuracy Of Wave Pattern Resistance Determination From Wave Measurements In Longitudinal
Cuts
63Nastia Degiuli, Andreja Werner, Zdravko Doliner
Influence Of Hull Shape On The Resistance Of A Fast Trimaran Vessel
71Antonio Cardo, Marco Ferrando, Carlo Podenzana-Bonvino
Hydrodynamic Performance And Exciting Force Of Surface Piercing Propeller
79 Kazuo NozawaDrag Reduction On A High Speed Trimaran
87 Robert Latorre, Aaron Miller; Richard PhilipsHydrodynamic Developments Of 154m Class Ro-Ro Passenger Ferry
93 Soon Ho Choi, Jung Joong Kim, Gun Ho Lee, Se Eun Kim, Sung Mok AhnOn The Hydrodynamic Performance Of High Speed Craft
101C. Bertorello, S.Brizzolara. D. Bruzzone, P Cassella, I. Zotti
A Comparative Analysis Of The Resistance Qualities Of A Series Of Semi-Displacement Hi-Speed
Mono-Hull Forms
109Jacques. B. Hadler And Richard Vanhooff
Session A2 Propulsion
Applicability Of A Rim Drive Pod For High Speed Ship Propulsion
1John Richards,Jon Eaton, Juergen Friesch, Michelle Lea, Donald Thompson, Bill VanBlarcom
Design Of A Waterjet Propulsion System For An Amphibious Tracked Vehicle
7M. C. Kim, H. H. 'hun W. G. Park
Enhanced Performance Scaling Methodology
17Jon E. Eaton, Eckhard Praefke, Friedrich Mewis
Erosion Problems On Fast High Powered Ships
27 Jiirgen FrieschDesign of optimal inlet duct geometry based on vessel operational profile
35 NWH. Bulten, R. VerbeekInvestigations About The Use Of Podded Drives For Fast Ships
41Cornelia Heinke, Hans-JUrgen Heinke
Dynamic Performance Analysis of a Gas Turbine/Waterjet Propulsion System for a Fast Trimaran
Ferry
49Giovanni Benvenuto, Ugo Campora
Minimising The Effects of Transom Geometry on Waterjet Propelled Craft Operating In The
Displacement and Pre-Planing Regime
59James Roy, John Bonafoux
Latest Developments in Fast Ship Gear Technology
67Franz Hoppe
Session A3 CFD
Fully Nonlinear Wave-Wave Interactions Between A High-Speed Vessel And
Incident Waves 1Ray-Oing Lin, Arthur M. Reed. William Belk-nap
A Numerical Method For Performance Prediction Of Hydrofoil-Assisted
Catamarans
9PERFORMANCE AND CAVITATION ANALYSIS OF A WATERJET SYSTEM ON A
CAVITATION TUNNEL
Klaas Kooiker, Tom Van Terwisga, Rob Verbeek, Peter Van Terwisga,
SUMMARY
To study the effect of intake working point (defined by Intake Velocity Ratio IVR) on pump and waterjetperformance, experiments with a Waterjet System mounted on a Cavitation Tunnel have been conducted. The effect of the intake on efficiency and head and torque coefficient for both the pump and the complete jet system were studied, as well as the
inception of cavitation on the pump. The tests have been performed on a mixed flow waterjet system that is
representative for current industrial standards. The effect of IVR on pump head coefficient and pump efficiency was tested for two working points. Inception of cavitation was determined for the design IVR and for an off-design IVR. At these IVR's, cavitation inception was found for several flow rate coefficients by lowering the ambient pressure in the tunnel. The paper clearly shows that pump-intake interaction effects are present and need to be accounted for. The test set-up described appears suitable to determine the overall jet system characteristics and to assess theeffect of the intake on pump performance.
MARIN, The Netherlands
MARIN, Delft University of Technology, The Netherlands Wärtsilä Propulsion Netherlands, The Netherlands
Royal Netherlands Navy, The Netherlands
1. INTRODUCTION
Little is known about cavitation inception of waterjet
pumps when mounted in the complete wateijet system. Most often, when cavitation inception is discussed in the
literature, the pump working point is meant where the
pump head starts to decrease for lower ambient pressure
(or increasing specific suction speed nss). Head
breakdown criteria of 1 or 3% are often used. In this situation however, the impeller usually shows already a lot of cavitation and consequently produces a lot of noise. Furthermore, these "inception points" are usually determined in a pump loop set-up without an intake that absorbs boundary layer flow, that was built up along the
hull of the ship.
Figure 1: Test set-up of waterjet system in MARIN
Large Cavitation Tunnel
Session Al
57For naval applications, cavitation is to be avoided as long
as possible. The question remains however at what
specific suction speed the impeller, or any other part of
the jet system, shows the first signs of cavitation and
starts to produce excessive noise. Because cavitation
inception is very sensitive to local flow distortions, it is important to model the local flow conditions at the pump
inlet as good as possible. This was already argued by
Kruppa [1993], who cites results on the effect of several upstream bend configurations on pump efficiency and
torque and head coefficient. And these results dealwith
integral quantities! The effect of an intake distorted flow on cavitation inception is likely to be more pronounced than it is on the integral quantities.
The objective of the study presented here is twofold; The first objective is to obtain more knowledge on cavitation inception limits of a representative waterjet pump. The second objective is to obtain insight in the effect of the
distorted flow, provided by the intake to the pump, on
pump performance.
This paper presents the results of an experimental study of the effect of intake working point on both the integral quantities such as pump efficiency and torque and head
coefficient, as well as on cavitation inception number
based on
visual identification of cavitation in arepresentative waterjet system. A detailed description of
the test set-up is included.
Furthermore,
results of the current test
set-up arecompared with those from a traditional pump loop set-up without the presence of a waterjet intake and an ingested
2. TEST SET-UP
The test set-up consisted of a waterjet system with retour
conduit mounted on top of a cavitation tunnel. A
photograph of the test set-up is shown in Figure 1. A more detailed drawing of the waterjet mounted in the tunnel cover is presented in Figure 2. The aeometry of
the
waterjet was provided by Wärtsilä Propulsion
Netherlands.
+ 4 + 1- +
1-+
PORSHLOL - A
Figure 2: Waterjet system mounted on tunnel cover The experimental set-up enabled independent variation of the intake operating point (IVR) and the pump working point (KQ), allowing for a study of the effect of IVR on pump performance Visualisation of cavitation inception
on the impeller and the stator was made possible by using a Perspex pump housing, containing both the
impeller and the stator. The working points of the pump and the intake could be adjusted independently.
2.1 BOUNDARY LAYER AND ARTIFICIAL THICKENING
LDV measurements on the boundary layer velocity
distribution have been conducted in the past for a smooth tunnel wall and a tunnel wall fitted with a serrated strip
as depicted in. w
Strip 1: w=30 mm, h=19 mm, 11=40 mm, 12=30 mm,
b=510 mm, a =approx. 33 deg
Figure 3 Serrated strips for boundary layer thickening The serrated strip is fitted to the tunnel top, at the point
where the contraction in the tunnel connects to the
parallel part of the tunnel. This
is some 630 mm
upstream of the intake ramp tangency point.
Session Al
58The boundary layer velocity profile has been measured with LDV equipment by Van Terwisga [1993] for two transverse locations and two streamwise locations. The sen-ated strip resulted in a boundary layer thickness of approx. 28 mm at a downstream position of the strip of 480 mm at a tunnel velocity of 7 m/s.
In line with MARIN's best experience, carborundum
grains were alued in the leading edge area of the impeller blades with an average size of 40 Jim over a chord length of some 4 mm f-rom the leading edge to induce transition to turbulence in the boundary layer.
2.2 TRANSDUCERS AND DATA REDUCTION
The objective of the tests is to determine the waterjet system hydraulic performance characteristics and to
determine its cavitation characteristics. To this end, the
following transducers were mounted:
Flowrate transducer
An Electromagnetic Flow Meter (type IFM 4080 K from Krohne-Persenaire) was used during the tests to measure
the flow rate. This meter was, following the
specifications, mounted in the return duct, well away
from disturbances in the conduit that might have affected
the measurements.
Head measurement
To determine the hydraulic performance of the system,
total head over the pump and the waterjet system are
important parameters. The total head is defined by the
following equation:
H = -1 p(t7,25, (P, 7153) p g
-2 where
= average energy velocity in longitudinal direction at position i;
1 f f
Q jAj,
= flow rate
= velocity along longitudinal co-ordinate
= unit normal vector rectangular through stream
tube cross sectional area
= Area
= mean static pressure at position i (see figure 4)
= mean z
co-ordinate (z co-ordinate positive pointing vertically downward)It is seen from eq. (2) that the velocity distributionacross
the relevant cross section A, is needed to determine the
average energy velocity Fie4.
In the absence of the
detailed flow field, the average energy velocity ii-eg is approximated by the average volumetric velocity Wig
Figure 4: Definition of station numbers and normalized energy fluxes according to ITTC [1996]
The mean volumetric velocity is however only equal to the average energy velocity in a uniform flow field. As
the actual flow field at the pump inlet (pos. 3) and the pump outlet (pos. 5) is non-uniform, an error is
introduced here. Estimates of the error introduced by this
simplification can be calculated for distinct velocity
distributions. Scherer et al. [2001] present the following table of deviations:
Table 1 Typical energy factors for various velocity distributions (mean velocity based on unit free stream
velocity)
Session Al
59the measuring system. The total absolute pressure at any pressure tap i can thus be obtained from:
= pa,, + Po+ pgzi (5)
where
P,,
= atmospheric ambient pressure measured witha mercury barometer mounted near the cavitation tunnel
Po pressure measured at the centreline cavitation
tunnel, relative to the atmospheric pressure
= pressure measured at station i, relative to pressure at centreline tunnel Po
zi vertical co-ordinate of station i, relative to centreline cavitation tunnel. Positive z direction pointing downward.
Torque and rotation rate
A torque transducer was mounted between the sealing of the watetjet pump and the electric motor drive (after the
1:1 transmission). The torque transducer was designed to measure a maximum average torque of approx. 125 Nm. During most of tests, the torque transducer measured in the range of 10-20% of its maximum. The calibration of the torque transducer in its measuring range was checked with a static load in "built in" condition, so as to make sure that no residual torque due to the set-up would affect the calibration. The results from the static load test were
all within 2% of the calibrated torque. As the original
calibration was considered more accurate, it was decided to use this original calibration. A rotation rate transducer was mounted on the outgoing shaft of the electric motor (in front of the 1:1 transmission to the pump shaft). This rotation rate transducer was checked with an independent transducer prior to testing. The difference in results of these two meters appeared to be well within 0.5%.
Tunnel velocity
The tunnel velocity or free stream velocity was measured
from the pressure difference in the contraction of the
tunnel, upstream of the measuring section. This pressure
difference signal has been calibrated
by LDV
measurements to arrive at the most accurate estimate of
the tunnel flow velocity (outside the tunnel boundary
layer).
3. REVIEW OF TESTS
Three types of tests have been conducted:
1. Jet system performance
'"). Cavitation Inception
3. Cavitation observations
The Jet System Performance tests can be subdivided into three distinct types:
Exploratory tests where pump rpm and tunnel
velocity were fixed (at n = 1250 rpm and Uo = 6 m/s)
and where the pump working point (in terms of KO
Type of velocity distribution Energy
factor l3E
Mean
velocity Uniform vel. distribution 1.0 1.0
Linear Distribution
O - Free stream 2.0 0.50
1/3 Free stream - Free stream 1.25 0.67
1/2 Free stream - Free stream 1.11 0.75
Exponential distribution
Exp = 1.0 (linear) 2.0 0.50
Exp= 0.5 (parabolic) 1.35 0.67
Exp = 1/7 (boundary layer) 1.05 0.88
The lcinematic contribution to the energy flux canE/cm be expressed in terms of the energy factor by:
43Ek, = Q. pV 2 0,1
2
(4)
The average static pressure T? is approximated from 5
pressure taps fitted circumferentially at the required station at equidistant positions. The z co-ordinate is measured relative to the centreline of the cavitation tunnel, where the origin of the co-ordinate system is positioned, the z axis being positive when pointing
vertically downward. All pressure transducers measured a pressure difference between the actual pressure tap and the ambient tunnel pressure at the centreline tunnel. This ambient tunnel pressure is referred to as Po. Prior to every
has been varied by gradually shutting of the valve in the return duct
Jet system performance tests at two distinct operating points of the intake (IVRri values of 1.6
and 1.2)
Jet system performance test to get an appreciation of the effect of IVR, on jet system performance. To this
end, two distinct pump operating points were
adjusted (1(Q=0.40 and 0.37)
Cavitation inception tests were conducted to determine at
which specific suction speed number nss, cavitation
would first start to occur visually.
To this
end,visualisation
of the impeller and stator was made
possible by the Perspex pump housing, containing both the impeller and the stator (see Figure 5).
Figure 5 Photograph of Perspex pumphousing
For each cavitation inception test, a certain flow rate coefficient K.Q at a prescribed IVRi. was adjusted by adjusting the valve for a pre-set tunnel and impeller
velocity. Cavitation inception was subsequently found by lowering the ambient pressure in the tunnel. The limits in the flow rate coefficients over which inception could be found were set by the limits of the water tunnel and test set-up. The non-dimensional pressure at which cavitation inception would occur is defined by the specific suction
speed nss, defined by:
nfd
nss= (gHs)Y4 where 1 , Hs=-Q9t+
Po+ P3meos + pgT3 PgPv = vapour pressure of water
The suction height Hs is the pressure at which incipient cavitation was observed first. One can understand that for
an increasing non dimensional rotation rate, the
resistance against cavitation decreases, that is for
increasing suction specific speed.
Session Al
60Finally, cavitation observation tests were conducted with the aim to detect possible cavitation forms that could be
erosive in nature.
4. DISCUSSION OF RESULTS 4.1 JET SYSTEM CHARACTERISTICS
Figures 6, 7 and 8 show the jet system characteristics in
non-dimensional form: torque coefficient, pump head
coefficient and pump efficiency for three distinct tests. The scatter in results of these tests are essentially caused
by differences in intake working point IVR. as will be
discussed in the following. For reference purposes, the pump characteristics as derived from tests in a different
set-up on a slightly larger pump diameter (factor 1.6
times bigger) have been plotted in the same graphs. It is
shown that significant differences with these earlier results occur. These differences may be caused by the different set-up of the tests. It has been demonstrated
here that the intake worldng point does have a significant
effect. This is caused by the intake velocity profile,
offered to the pump. As the tests in the pump loop circuit did not have the possibility to easily and truthfully model the intake working point, it is considered to be the major cause for the discrepancies measured. Another possibility for the deviation could be the presence of roughness on
the leading edge of the impeller blade. Based on
MARIN' s experience with propellers, this effect is
considered of secondary importance.
Figure 6: Jet system characteristics: Torque coefficient
1.4 1.3 1.4 .3 O 1.2 1.1 t 1.0 20 0.8 a 0.7 Ic 0.5 Torque coefficient 1, 1.1 .11P1111... -.11PP%111.
' , '41IM
I
1.0 o 8Mmi...
"
.16M it 0.7 -o- 254 mm intake 0.6 -0-160 mm intake 1 0.5 . -6- ivrr=1.6=
0.4 II -8-ivrr=1.2 05 0.6 0.7Pump head nondimensional
-- 254 mm intake
0160 mm intake
-ivrr=1.6
0.4 -,--ivrr=12
05 0.6 0.7
Ko/Kc best performance
(6)
2P1732 Pv) (7)
0.8 09 1.0 11 1.2
08 0.9 1.0 1.1 1.2
Ko/K. best performance
Figure 7 Jet system characteristics: Pump head
; 1.0 0.9 0.8 0.7 of' -è 0.8 1.20 0.92 0.90 1.1 0.5 254 mm intake -6- 160 /TIM intake s- ivrr.1.6 Pump efficiency K./Kb beat performance
Figure 8: Jet system characteristics: Pump efficiency 4.2 REPEATABILITY OF RESULTS
In order to get an appreciation of the repeatability of the results, results that were obtained at the same pump and intake worlcing point (KQ and IVR respectively) were compared with each other. IVR identity was selected as an additional constraint, as it was clearly demonstrated that this intake working point did significantly affect the
results.
An indication of the repeatability of the torque
coefficient could be obtained from the jet performance test series at constant IVR, the cavitation inception tests at constant IVR and the corresponding IVR values in the
IVR variation tests at constant KQ. This gave a number of
21 pairs of results that could directly be compared with each other with the same independent variables KQ and IVR. From this comparison, a sample standard deviation in torque coefficient was found of 0.83%.
4.3 EFFECT OF IVR ON PERFORMANCE
The effect of the Intake Velocity Ratio IVR, on head
coefficient and pump efficiency is presented in Figure 9. Pump characteristics as RIVR)
Figure 9: Effect of IVRR on head coefficient and pump
efficiency
The coefficients are normalized relative to the
coefficient's value
at an IVRR=1.2, which can be
regarded as the design point of the intake, where the most uniform flow is delivered by the intake. It is seen that the head coefficient increases with 3% for an IVRR=1.6 and
even up to 8% for an IVRR=2.0. The effect of IVR is
greatest for the lightest pump loading (KQ=0.40).
Session Al 61
A similar trend is seen in the pump efficiency 17e, although the effects are somewhat increased due to
similar trends in torque coefficient.
The effects of IVR on torque coefficient are presented in Figure 10. The torque coefficient decreases for increasing
IVRR. A decrease of 2.5% at an IVRR=1.6 occurs,
whereas the torque coefficient decreases even with some 7% for the IVRR=2.0. Azain, the effect is strongest for
the lightest pump loading.
Figure 10: Effect of IVRR on torque coefficient 4.4 EFFECT OF IVR ON CAVITATION INCEPTION
Figure 11 shows the results of the cavitation inception
tests. It is seen that tip vortex cavitation occurs first,
followed suit by sheet cavitation at the outer radii of the impeller. In addition, cavitation was observed first in the blade top position, where the axial velocity is relatively low due to the intake flow and impeller shaft wake, and where the pressure is also at a low due to its high vertical
position.
Cavitation could only be noticed on the stator at very high specific suction speed numbers nss, which were
adjusted during the cavitation observation tests.
2 1.8
Pump characteristics as t( IVR)
Cavitation inception
09 0.92 0.94 0.96 0,98 1 1.02 1.04 1.06 1.08 11 K01K0 best pertorrnance
Figure 11: Results of cavitation inception tests
Although the results in Figure 11 show a corresponding tendency, the reliability of two points can be questioned; Sheet cavitation inception at the highest loading (lowest
KQ) and tip vortex cavitation (tvc) inception at the lowest
loading (highest K,Q). Both questionable points were
measured for an off-design rvRR of 1.6.
...4-Kh,Kg.0.40 -s--etap. 6,0.40 _
-..
-KR 69=0.37 --/, -etap, Kg-0.37 em ...." ...., ...-- ...--onti=1.6 tvc inc--4-ivrr=1.6, sheet inc --ii-ivrr=1.2, tvc inc
--11-ivrr=1.2, sheet inc
...10 :idge."'"AgelEl . -1.10 1.30 1.50 1.70 1.90 2.10 VRh 05 0.6 07 0.8 09 1.0 1.1 1.2 13 1.10 R 1 00
/-..
Kcm0.40 -ea -61g, 6,0.37 0.80 0.90 1.10 1.30 1.50 1.70 1.90 2./0 IVRFor decreasina, pump loadinas (increasing flow rate
coefficient KO, cavitation inception occurs at higher nss values. For the flow rate coefficient KQ at the maximum pump efficiency point, the specific suction speed number
at which incipient tip vortex cavitation is found is the
approximately the same for both IVRR values of 1.2 and 1.6. Incipient sheet cavitation is found for slightly lower ambient pressures. The sheet cavitation seems to be more
sensitive to the pump input velocity field than the tip
vortex cavitation, as the difference between nss values for IVRR = 1.6 and 1.2 at maximum efficiency is about
12%.
For a 8% lighter pump loading, cavitation inception can
be delayed until approx. 1.7 times the nss value at the
pump loadina at maximum efficiency.
During the cavitation inception and observation tests.
only suction side cavitation has been observed.
Figure 12: Photograph of developed cavitation
5. CONCLUSIONS
The current study has demonstrated the applicability of waterjet system tests on a waterjet that was mounted on
top of a cavitation tunnel. The complete range of flow
rate coefficients of interest could be investigated, without the additional use of another pump in the circuit to adjust intake and pump working point independent from each
other. The conduit resistance was relatively small and
further reduced by some 50% by the suction effect of the outlet scoop. By applying a transparent pump housing,
cavitation could be observed on both the impeller and
stator . In addition, the intake lip could be observed
through Perspex windows in the intake and throual
windows in the cavitation tunnel.
The pump characteristics derived from the present tests
are in good qualitative agreement with the pump
characteristics measured earlier on a sliahtly bigger
pump (scale factor 1.6) in a conventional pump loop set-up. The absolute values differed however (some 2 % in efficiency at maximum efficiency point).
This is most likely ascribed to differences in the test set-up. The repeatability of the new test set-up was found to
Session Al
62be adequate. The torque coefficient was checked and showed a sample standard deviation of s=0.83% on a
sample of 21 pairs of measurements that could be
compared directly.
Clear effects of the intake working point, expressed in
IVRR, on pump performance and cavitation inception were measured. Effects in head and torque coefficient
were measured up to some 3% between a design and an off design condition for the intake at the lightest pump loading condition tested. Due to accumulating effects, the effect in efficiency appeared to be some 5% for this case. Cavitation inception started with tip vortex cavitation at the suction side of the impeller blade, followed suit by
sheet cavitation at
the outer radii of the
impeller.Inception of sheet cavitation appeared to be more
sensitive to intake working point than inception of tip
vortex cavitation. For a 8% lower loading of the pump,
the cavitation inception number found for sheet
cavitation appeared to be some 70 % higher. Based on
the cavitation extent, the character and position of the cavitation at inception, it is expected that cavitation
inception can be sigmfficantly improved by modifying the
impeller design. It is, however, likely
that such a
modification will be at the cost of pump efficiency.
Based on the present study, it
is concluded that forreliable measurements of pump performance, including cavitation, it is important to properly model the complete waterjet system, including a hull boundary layer.
REFERENCES
1. British Standard BSI Guide 7405:1991; 'Guide to
selection and application of flow meters for the
measurement of fluid flow in closed conduits', Aug. 30 1991, ISBN 0 580 19335 7
ITTC Quality Manual; 'HMI Speed Marine Vehicles
- Waterjets', Section 4.9-03-05.2, 21st ITTC 1996
Kruppa, C.; 'Contribution to the 20th ITTC
workshop on waterjet propulsion Pump Installation
efficiency', 1993
Scherer, O., Mutnick, I and Lanni, F.; 'Procedure for conducting a towing tank test of a waterjet propelled craft using laser doppler velocitimery to determine the momentum and energy flux', 26th ATTC, Webb
Institute, July 23-24, 2001
Van Terwisga, T.J.C., Radstaat, G and Van der
Weerd, G; 'LDV measurements on waterjet intake
MARIN Data Report', MARIN Report