ARCHIE'
PUBL. NO. 3"
OF THE N.S.M.B. Authorized Reprint fromSpecial Technical Publ atihn 474 a'
Copyright 41.0,,a)cheepsbouwkunde
American Society for Testing
1916 Race Street, Philadelphia, Pa. 19103
1970
Technische ;Hogeschooi
J. H. J. van der M euleni
Delft
Cavitation Erosion of a Ship Model
Propeller
162
REFERENCE: van der Meulen, J. H. J., "Cavitation Erosion of a Ship Model Propeller," Characterization and Determination ofErosion Resistance, ASTM ST P 474, American Society for Testing and Materials, 1970, pp. 162-181. ABSTRACT: A four-bladed propeller model has been tested in a water tunnel
with a flow regulator. Cavitation erosion of the aluminum alloy blades was caused by simulating the wake pattern behind a single screw ship in the test
section of the tunnel. The relationship between the rate of weight loss and the
time of exposure to cavitation was measured for each blade individually.
Comparative tests, using the same material, have been carried out in a rotating
disk apparatus and a magnetostriction oscillator. The erosion rates of the propeller model and of the specimens tested in the rotating disk apparatus show a similar dependence on time during the incubation, acceleration, and
deceleration period. A steady-state period does not occur. The erosion rates
of the specimens tested in the magnetostriction oscillator show a considerably
different dependence on time. The mean erosion intensity of the test devices
is 2.5 to 3 times larger than with the propeller model, based on the actual time of exposure to cavitation.
KEY WORDS: ship, propeller, cavitation, erosion, rotating disk,
mag-netostriction, aluminum alloys, evaluation, tests
The cavitation damage resistance of materials is tested mostly in devices
where cavitation is produced artificially. Very often magnetostriction
oscillators or rotating disk apparatuses are used for this purpose. They are sometimes called accelerated devices, but Thiruvengadam [1J2 has shown that the intensity of cavitation damage, experienced in field installations, can be much higher than in the above mentioned devices.
A comparison between the erosion resistance of materials is based usually
on the weight loss measured in a test device or, more specifically, on the
rate of weight loss as a function of the time of exposure to cavitation. According to Thiruvengadam and Preiser [2], the rate of weight loss as
a function of time can be divided in four typical periods: (1) an incubation 1 Scientific officer, Netherlands Ship Model Basin, Wageningen, Holland.
VAN DER MEULEN ON A SHIP MODEL PROPELLER 163
period, (2) an acceleration Period, (3) a deceleration period, and (4) a
state period. In their opinion the rate of weight loss in the
steady-state Peiiod -should be used for comparative purposes. This suggestion
was disputed by Plesset and Devine [3]. They photographed the cavitation cloud over a specimen tested in a magnetostriction oscillator and noticed
a much reduced intensity when the surface roughness of the specimen
increased due to erosion. Thus, they attributed the change in the damage rate to hydrodynamic effects. Moreover, they found a different dependence
of the rate of weight loss on time, namely, the maximum erosion rate
showed a much more nearly constant rate of weight loss than indicated by Thiruvengadam and Preiser [2].
Similar results are reported by other investigators. For example, the results found by Hobbs [4], using a magnetostriction oscillator, closely agree with the results found by Plesset and Devine [3], whereas, on the
other hand, the erosion results of steam turbine blade shield materials by
Smith et al [5] agree with the results found by Thiruvengadam and
Preiser [2]. It should be noted, however, that erosion in this case was due
to the impact of water droplets. Heymann [6] mentioned quite another type of rate of weight loss versus time pattern. Here, the erosion rate
begins at a maximum value and then decreases with time until a steady-state value is reached. These results were obtained in a spray impingement erosion test facility.
The different opinions about the significance of the rate of weight loss versus time measurements are brought together, in a way, by Heymann [6]. He has developed a statistical model of the erosion process on the assump-tion that fatigue will be the predominant mechanism for material removal. The analysis predicted rate of weight loss versus time patterns which were almost identical to those measured. According to the analysis, the erosion rate will tend to reach a steady-state period, but, as pointed out by
Hey-mann [6], this state is probably unattainable in practice because the
increasing roughness of the surface itself will affect the erosion rate. A more physical approach of the problem is suggested by Engel [7]. Her
model of the erosion process gives a nice interpretation of the existence of four typical periods in the rate of weight loss versus time diagram.
At the Netherlands Ship Model Basin (NSMB), much attention is
paid to the prevention of cavitation erosion of ship propellers by eliminating
or limiting the incidence of cavitation. But little is known about the weight loss or the rate of weight loss as a function of the exposure time to cavi-tation of actual ship propellers or model propellers. The flow conditions on a propeller which cause cavitation are much different from the "flow conditions" in a magnetostriction oscillator, where cavitation is produced
by the vibration of the specimen itself. Even the flow conditions in a
rotating disk apparatus, where cavitation is produced by a hole in a sub-merged rotating disk, are different from those with a propeller. Hence, it
may be expected that the rate of weight loss versus time diagram of a cavitating propeller will give a different picture than a corresponding
diagram obtained from tests in an erosion device. In order to investigate
this matter, a model propeller was eroded in the cavitation tunnel with
flow regulator at the NSMB, and the rate of weight loss results were
com-pared with those in a rotating disk apparatus and a magnetostriction
oscillator, using specimens of the same material.
Erosion tests with ship model propellers have been reported before by
Georgievskaya and Stumpf [8]. They used model propellers that were
coated with lead and measured the weight loss due to cavitation erosion.
Cavitation Phenomena on Single Screw Ship Propellers
Cavitation on single screw ship propellers is produced mainly by the
unequal velocity field of the wake behind a ship. Wake simulation in a cavitation tunnel is a prerequisite in testing model propellers of single
screw ships with regard to a reliable prediction of the expected degree of
safety against erosion. At the NSMB, the wake flow is simulated by means of a flow regulator consisting of a multiple valve system. This
way of testing propeller models has been analyzed by van Lammeren [9]
in 1955. With single screw ships, it often has been observed that the
trailing edges of the propeller blades were eroded and bent towards the pressure side. An explanation of these phenomena has been given by van Manen [10, 11].
When observing a propeller blade section passing the high wake region,
the following is noticed. The blade enters the high wake region, and a
sheet cavity is initiated at the suction side of the blade, near the leading
edge. During the subsequent passing of the high wake region, first the
length of the cavity increases until it almost covers the suction side of the blade section, and next the leading edge of the cavity moves towards the trailing edge of the blade and bubbles remain behind. These bubbles are dragged by the flow, but their velocity will be much smaller than the
free-stream velocity because of the positive pressure gradient which exists
towards the trailing edge of the blade. Such effects have been discussed,
for example, by van der Walle [12]. Finally, the pressure will exceed a critical value, and the bubbles will collapse near the trailing edge and cause damage to the blade. Van Wijngaarden [13] has calculated the
average pressure generated by a collective collapse of a large number of gas bubbles near the trailing edge of a propeller blade and has found it to be sufficiently high to explain the bending of the trailing edge.
The above mentioned phenomena mainly occur on single screw ship propellers. For that reason a wake distribution of such a ship has been
111111111r7111111111
e
AN DER MEULEN A'-SHIPMODEL-PROPELLER. I 65
Test Equipment and Propeller Model
The test equipment, used in the propeller erosion experiments, is the
cavitation tunnel with flow regulator of the NSMB. The test section
con-sists of a "slotted-wall" construction with inner and outer diameters of 40 and 80 cm, respectively. Some details of the flow regulator and test
section are shown in Fig. 1.
The propeller actually used in the erosion experiments was a four-bladed propeller with a diameter of 22.4 cm and a constant pitch ratio of 0.929. The propeller drawing is shown in Fig. 2. In order to increase the thickness of the blades at the trailing edge, a part of this edge was cut off. All blades were made dismountable. So, weight losses could be measured for each blade individually. A fifth blade, identical to those of the propeller, was
attached to a bar in the slot between the walls of the test section. When operating the tunnel, the water in this slot is whirling round in eddies.
This blade was used in order to measure a possible weight variation due
to corrosion. The axial wake distribution which was simulated in the
cavitation tunnel is shown in Fig. 3. Here, the wake fraction w is given by:
V-
V4V
where:
V = ship speed, and VA= intake velocity.
.1 ;i:::--,VAWDER.:-'MEULEN.iONT'A5,5111P. /410DEL 'OROPELLER ' 167 7 0.50 " -0.65,V 0:1 i A , i . iFit..4:a4 I ' .," " - D. 224 Min:
-tFIG. 3Axial wake distribution"simulateei in caritatimPtunizel (w :=wake fractiara)
-,
i
The wake.-distribution,correapOnds' to an actual'7,one.'which'Wakineasured
-behind Of1Weig1itef
tunnel presatirewaksetron a lowei: value than would c-orrespond, With Oat ShiP-7cJiiditiO:thio,-*Wdone to intenOfy_Tavitation-and'thui--accelerate
the *skin process. The propellers -revolutions per minute were kept
"- constant at 2400.
- The material which, was used manufa,ctliring the prOP'eller-hlideir win-i an aluirdinini'-alloy:: The -Chemical coinp-O`sition of this rnateriaLl in.
weight percent: is: aluininum94.16;'copper 3.60(ntninganeSe 0.02; ifoii 0:42, .
silicon 0.57, titanium 0.11, .-magnesium 0:56, and chromium 0.02. The
material does not have either an equivalent American Society for Metals'
'7:-designation or any other- '7:-designation. It was cast into- . . a-. sand moUld:,_
14 12 2 0 o Q002 0004 Q006 (1008 001 Strain
FIG. 4Stress-strain diagram of aluminum alloy.
aluminum suffered from porosity which resulted in a low-strength brittle material. The cast propeller blades were machined by the tape-controlled
milling machine of the NSMB, and finished by hand. The mechanical
properties of the aluminum alloy are given below:
Ultimate tensile strength 13.0 kg/mm2
Ultimate elongation 0.92 percent
Modulus of elasticity 6 2 X 103 kg/nare
Brinell hardness (HB10/1000/30) 72.5 kg/aim'
Specific gravity 2 72 g/cm3
These properties were measured by using specimens and a test bar which
were cut from the riser of a casting of the same alloy. The stressstrain
diagram, Fig. 4, shows that the strain energy is only 0.0915 kg/mm2. The properties may be considered to be representative of the propeller blades, disk, and magnetostriction specimens.
Experimental Results
The total test duration of the propeller model in the cavitation tunnel
was 40 h. The propeller passes in and out of the cavitation region twice
per revolution. Using stroboscopic lighting, it was observed that the passage of this region in the upper position of the blades took about
'500
VAN DER:MEDLENg0N; LAHIP*ODEC PROPELLER
TABLE.1Wiiiei.'ionditiores during erosion,t4t in-ccuntatiah tunnel.
I .
Exploa'tire'Tirae;31'^: '-Water 'Temperature;, . Air Content at .
" = . deg C STPa p6r.liter of water.
-,.3i-- ,,,- - -,.. ' 4f : ''I - :23.: .-, 16 ,3,.5', ..61- '.. --7-.:- - 1--. '-' -e-8i ... ,- 25 . .' '9..17 .... ' ' 101,, '. - icilgrr--7 - , ,
a STP =st tireatt,id presaure. ---, 10.9 12.8, 10.6 14.6 10.1 :Y. 25i ." '12.2
31-
, ,26- 37k'
' - 2-4t 0 26 : ,3
111J. 1.0 s."11 1 ,,,,77ps . .4U Time hours,. d'-.;;r; 1 . ' 1 ; ' ; ;; ' FIG,..1-5 Effect of time onrWeight losk and rate of weight loss of propeller : .., -ill c , , ' ,-9F744' i,fi.. F,-.,,f7, ti --;'1 f_... .., 't.'-:. .) l -ci 1!-)1),-,--is ' '..: -.;! ,t/.--,'il; ... : Nt)..",-.. , L t!il,',,,,?:r:, - ! `--"'" i1 z7.# -,,i -' ...1...z , ,,f-1,0,.,
,
---.'"7.:- ---, - 1,)' .,. iti-,;1" ''''';',-..,..kt !' :14.;TI, .1' , ,i/
. li, :-.P1 / .i. 0 ,, '' t:',::FIZ::T.: ::. , t ''' 11. tb.L. ., , '-s -' 71,/
. I, . _. t lf. _ 1: ... ''i LL !,
' V.1.-. -. 1 1 I , i ..z.:,,- `1:7 ---i 41' =1.c:_r:::: P i -I , tl .10" ./. , ') 7; c ;1:.:s ,f,' , . rt.g.. .!1 i ':"..1-N..., I . '" il, ,'f,. ' 'cn
0
-25
0 -10 15 20 25
Time. hours.
FIG. 6Effect o time cm weight variation due to corrosion of Blade 5.
the blades 4 percent. This corresponds to an actual exposure time to
cavitation of 5 h.
-During the initial stage of the experiments, weight losses were measured
after short time intervals, but a large scattering of data occurred. This was due to the porosity of the material which caused water absorption.
This difficulty was oversorne by heating the blades in an electric furnace
at a temperature of 170 C for 1IA h, before weighing took place. The water temperature and air content in the cavitation tunnel are given in
Table 1. .
The weight loss data and the rate of weight loss versus time diagram
of the propeller model axe shown in Fig. 5. The rate of weight loss curve is determined from slope tangents to the weight loss curve. .
Weight loss data of the individual blades differed less than 10 percent from the average at the end of the test. The weight variation of the fifth
blade as a function of time, Fig 6, shows that the weight of the fifth
blade increased during exposure to the water in the cavitation tunnel..
A photograph of the propeller model in the cavitation tunnel after an exposure to cavitation of 40 h is shown in Fig. 7. Because the material
used was very brittle, only a slight bending towards the pressure side of the blades was noticed. In Fig. 8, six photographs are given of Blade 4, showing the erosion patterns after different exposure times to cavitation. They are quite representative of the erosion patterns of the other blades. It can be seen that the trailing edge is crumbling off in the final stage.
Comparative Tests Using Rotating Disk Apparatus
To compare the erosion rate--time pattern of the propeller model With
the one measured in standard erosion fest devices, it waS neceSSary io
carry out tests with these devices using the same Mit:66117'6f which the propeller Was made. As the NSMB does not have any standard erosion-test -device, on our -request we received :full cooperation of Lips Propeller Works, The Netherlands, and Hydronantics Inc :USA.; Who were willing -to execute compaYative tests with their eqUipmerit..
-.VAN, DER MEULEN,:ON .-A SHIF! MODEL ,PROP'ELLER 171
-,
FIG. 7=-Eroded Propeller model after an exposure to caintationof-.40 h.
At Lips Propeller Works a . rotating disk apparatus is Routine
tests usually are made with salt water, at a temperature of 28 c: The-SulD merged disk is rotating horizontally, with revolutions per minute of 3000:,_.
,
,Cayitating holes cause erosion on the specimens, The locations Of hole and
. . J
Specimen .ate shown.inFigg. Two specimens areinserted, on each side the disk The holes, are nOt.: perforating the disk;-;tIeirAepth.. being ,,Only 7 mm. The circumferential speed of the specimens is.40 m/s: Stilling.vanes
RESISTANCE
-r7,
3
231/2 h
400 ch E 300 1 -Test-.S .imen so 100 0 _
VAN DER MEULEN,ON3 A .5111P MODEL,PROPELLER'. 17;37,
FIG. 9=31-D 's,k.:Of, "rotating disk apparatus used at I.-iPsf"ropeller Works (dimensions
inmm). f;' :- - ,',-.':,';', , - --1;'-',::: - '
. . .,,.-i '
-. ,. ,
are used -on eithet.siaiy iil:the-:disk'-'todamp the induced bireUlation of the
water. ...,-..17.' S.A,:1-'
More &tail.s.r.oni,, the rotating disk -apparatuS ire ',given bj;...:-HanS6 and Rasmusse-n-[14 ]:.,
Four specitijetis,Weie=fab-riedied-`,,-andtestedjii'the rotating..diskapparatus i':'. '
using tap water a:S-:iiiiiid:TWo=i4keiri-iens (a.i'Ajcile A) were,cutiifOm the
--riser of the propeller'eaSting.-They were embedded iii Araldite",(an refioxy. : -cr,qt.'". ' '.. ' . . ,:: ..t...,--"-. .?"7' r'.,f;g-70. ,
FIG..1OEffect of time On weight loss and rate of weight loss of Sample A, measured with rotating disk apparatus.
.---2,.Z2,*:3 :..2i.q- . ' .... - ..: 1 .,.'"-..,1.kr-: ,i, ?.... ...1 .... ,..:. , .-.., / , .zi:-..--,1. ;1.,:. ,.. , 7 6 8 0 4 5 Time, hours.
174 aimiActiRdii1ON `AND -DETERMNATION OF EROSION
RESISTANCE.-0
1; 't)
'
el 5 ,8:Time, hours.'
i
-17,1FIG':11'=Effeet:of tirne",OnWeight'lass andriati'af weight lose. of .Sample Bçmeasured with
roiating,disk.cipparatas, ;";,e; ; `1..7
?..V..; ' PJ ;
re-sin). The other two specimens (Sample p) were made by .fedasting the
remainingo parts of the propeller, baStingl-_: No mechanical _property tests were run on this .ipecinieii.,The porosity of the material-,;:wairedUced only
shghtl3 due to `r'eCaSting: ?, 0,Lt. -±gr-_ 2-4 _ r 7 c4,z(3.14?, 00 o '
80f.
(left) A stieciinerir(right) Sample B specipen.
- ,
FIG. 12 Eroded ItOeiniens after difeiiilemricVito'6.4ii,itcition Of kh in rotating disk apparatus (botiont neralest to center of disk).
.1- .!;;-:--4:4---, !!,i,-!:--\:: j-. _ ..._ ,,-,4 ::: f: ., f c.!, !!.; -,i, !I i ; ,., A.'s---,"-..-,,..: 1 --!! ...,:.11.; ?i/, r , -:, -,1 . I - 'It. . 14 ,:l ' ., t- ; !... 1"7... 4 444 ,!.. . : I . I , , . : - -. .,.,.... ''.c.:-.C.'_:> ;--- :-, b .. 1 nt ' ,i -.f.)-1-- ".°1 I I r-r,-..-...,... .34 7r! r '` P'. - 4,-,-; . to ! . . ,-, ...,--:, ; 7 . I . i T) I --,.._ . 4 1 4.4. - 1 ,,,, L ' .,', Ft, 600 .500 400 E. ,n;;;3150 0'200 1. -100
VAN DER. MEULEN ON A SHIP MODEL ,PROPELLER. 175
The average weight loss and rate of weight loss versus time diagram of Sample A ia shown in Fig. 10, and the corresponding diagram of Sample B is shown in Fig:-11:.:The data have been corrected for water absorption. During the erosion test the temperature of the tap water *as 30 C and the static :pressure 75 cm of mereury. The tests were stopped whenthe.eroded
-area was noticed to run off the specimen -area: Then; the depth of erosion had become so large that Cavitation was induced by the eroded specimens
themselves. A photograph of each specimen is shown in,Fig: 12, With
these specimens (and with other specimens too) a remarkable observation was made, namely,:the depth of erosion is increasing gradually towards the center of the disk. The depth of erosion reaches its maximum value at
-the nearest point to -the center. Here, a sharp edge separates -the eroded
and not eroded area.
Comparative Tests Using Magnetostriction Oscillator
_
-The erosion characteristics of the aluminum alloy in a naagnetostriction oscillator have been measured at Hdronautics, Inc. The principle of this
device is based on .a high-frequency oscillation ofa.specimen which.is
sub:-merged in a liquid. Dile to the oscillation; the pressure On the specimen
face alternates which causes the formation and subsequent collapse of
'
FIG. 13Effect 0/ time on weight loss and rate of weight loss of Specimen 1, measured with magnelOstriction oscillator.
;11 7.° ,TI, :i;f ,:.r.". :. , ATI= -' . ;1. '.4 0v31 60 '-';--=.''''' 6011
'
100 -'. '4";-. 1201:1_ .. :=----. ' ,,,,.
,, Tirrie.,min.i,. -, .,F1G. 14Effect,of,time on 7ieight loss andrate of weight,loss of Specimen 2,-measured
:, ., ..' 1 Le ,":''': L: ..., ill...7 r:"' ,:'..; ` Ft ,
'with inagnetostriction oscillcitor.
150 1 "E. so a' 30 a. fr%i:
cairities: onsequent y, the specimen face,..becomest eroded. More details
- on the device and on the influence of some test Parameters are given, for_
example, by` "Thi'iuvengadam :and Preiser
_Two specimens;--with.:a'dianieter of Veinl; have been tested, theypwere
cut from the riser of the propeller casting. The test conditions are amplitude, 0.05 min; frequency, 14.2:kHz; solution, 'distilled water; and tern-per-attire,
-2016 .21 C. The weight ,10s and 'rate of-',Weight,.loss,yersus..fime diagrarns
of the specimen areshown inFig 13 andi14.
J. ' ;, :
Discussion Of Test Results'
-The 'eatelot,Weight loss diagram of the propeller model, (Fig.' 5), shows the existence of 4oni characteristic periods': During thd,fitstl-period:
i(incu-
bation period) the rate Of Weight loss is 'almost zero. Then; the fate of,weight loss increllsOs and reaches a Maximum 'value,.:(acceleriition-period)..' This is followed by a decrease (luring the ,deceleration.fieri6C1._ The rate of
weight loss reaches a minimum valtief_Aiikincreases again in the fourth period A coii-tant rate of weight loss is observed nowhere. 1 he first three
VAN DER MEULEN ON A SHIP MODEL PROPELLER 177
periods agree with those described by Thiruvengadam and Preiser [2] and
physically examined by Engel [7]. Her erosion model was applied on a low-strength brittle material, which corresponds to the aluminum alloy properties of the propeller model. During the fourth period, the erosion depth at the trailing edge of the blades approached the blade thickness,
and, consequently, the blade edges crumbled off, which explains the
in-crease of the erosion rate. This is demonstrated by the picturesgiven in
Fig. 8. The test results of the individual blades show some deviations between the rates of weight loss. In particular, the results of Blade 1
are quite different from those of the other blades since a deceleration
period is not found.
The weight measurements of the fifth blade resulted in a weight gain
(Fig. 6). Probably, this is caused by oxidation of the aluminum. Godard et al [15] have reported a rapid increase of the oxide film thickness when
aluminum is immersed in water. According to them, the rate of growth decreases with time and reaches a limiting value which, among others,
depends upon the temperature and the oxygen content of the water. Since the blade was immersed statically, its corrosion is not necessarily repre-sentative of corrosion at high velocities of Blades 1 through 4. The result
merely indicates that corrosion might have influenced the weight loss
results of the propeller blades.
The erosion results of Sample A tested in the rotating disk apparatus
(Fig. 10) show a similar relationship between the rate of weight loss and -time as found with the propeller model in the first three periods. A
steady-state period is not found. The rate of weight loss results of Sample B
(Fig. 11) show the absence of an incubation period. This might indicate that cavitation was more intense because the surface was less porous.
It would be most interesting to find the reason of the unexpected erosion shape on the specimens: the largest erosion depths appear inwardly.
The rate of weight loss of both specimens tested in the magnetostriction
oscillator (Figs. 13 and 14) begins with a maximum value. Hence, no incubation or acceleration period is found. The rate of weight loss first
reaches a steady-state period and then decreases again and approaches a
final steady-state value. A similarity between the rate of weight loss diagram of the magnetostriction oscillator specimens and the propeller
model can not be found.
Finally, we want to evaluate the mean intensity of erosion of the
pro-peller model, as tested in a cavitation tunnel, in comparison with the
intensity of erosion in the test devices. According to Thiruvengadam [16], the intensity of erosion I is defined as:
AV Se
I =
178
where: AV
rate of volume loss,
A area of erosion and
--Se= erosion .strength:
.
'denoting...the-j.4tetisity of erosionofAhe Propeller model, the rotating disk apparatus and, the foagetotijoiotir oscillator by. ./.4, re-.
Speetifely; ;k.'
7)
t t 2. A t.:;
The 'comparison is bi.iseci on the Mean rate of volume 10§§, *17-1i6h is
ob-tained siinply- bYT.d.iiiding the totalvolume loss by the total test duration. FOr the propeller model the total actual time Of exposure tO"caVitation is used; which is Only 5 h. By using the weight 'logs result', given in Figs. 5, '10,-.and 14; and by -Measuring the eroded iiieas, we' obtain:.
,.Ipro-peller model: :. c:-;,12.-,;-: . ,.1.. ',..,
(n,.=0.04,3
(AV) =0:00243 cm/h A;., =17.0cin2; t 1\
1 ,fs'ilrt. ''''1. rotating disk: ,1 (AV)
-.0 00663 crn/li Ae2\ t
2 Tniagnet*rictiOn oscillator: _ =1-.99 cm2; AV).= 0.013 Cm3/h; t Ae.=2.71 cm2; = 01_8_ ern3 /2; 13.) . . _This implies that the mean erosion intensity of the -.test devices ts.2.5 to 3 times larger thin with the --propeller model, based on the-abtiialAime of
exposure :to cavitation.-It nniSt b noticed,'...hoWevier,that the erosio
intensitv,at:the,--trailing edge of the bladeS;i;,inuch,higherrthan tfip:ercisio -intensitY-ayeragedrOver the eroded blade area:
According to Thirtivengadam [16], the erosion intensity of theinagneto-strictiOii os6i1lator is 0.8 W/m2. l'hii,value;ishase*on,rate of weight loss data in the steady-state period. If, in otirr,ca.:se; /the, erdSionNitiength is
equalized to the strain energy of the aluininum alloy, a -mean erosion
intensity of only,0:016, W/m2,is found. Probably this is true with_ all
ma-AeI
*.-0..00652.crii/hr" VAN DER MEUL61' ON A' SHIP lkopiEC-PROPRIER
terials possessing a low value of the strain energy because these materials are often very hard.
:t. Conclusions
The rate of weight loss due to cavitation erosioii'V a propeller model,_
tested in a wat,er tunnel, shows a similar dependence on time during the incubation acceleration and deceleration period as found with specimens of the _same material' jn a rotating; dig( apparatus,:--A- steady-state period
is neither found with the propeller model nor with the rotating disk.appa-ratus. The rate of Weight -loss of the magnetostriction oscillator specimens shows a considerably , different dependence on time as found- with the propeller model. Here, asteady7State period does occur. Based on these results,. one may conclude that the erosion process of the propeller model, for this specific matrial, is better reproduced bY the rotating disk
appa-ratus, Which is a fib,* facility, than by the Magnetostriction oseillator,
which is a nonflbw faciht.
The mean i'ide'tiiity of erosion of the test devices is " .5-'t,e; 3 tithes Wier
than the mean inten.sity-of erosion of the'propeller model, based on the actual exposure time to cavitation. However, if the intensity of erosion is based simply on the total test duration, the de ices may be called rightly: accelerated test devices - _
-To confirm the above iconclusions, further testing will. be necessary-, using materials which do not 'possess disturbing influences due to porosity and oxidation or corrosion in water.
Acknowledgmnts ,
The author gratefully acknowledges the cooperation, of Lipis Propeller Vv.Orks and Hydronatitic, Ihe.; who have been most helpful performing comparative tests in their erosion facilities.
References _ _
'
[11 Thiruvengadam,, A.; "Intensity of Cavitation Damage Encountered in Fie1.fiInstal-lations, " Teclinical Report :233L7, liydronauties; mc, ,Feb.. 1965. r
.[2] Tliiiiivengadam; A. :inelPreiser; H.'S., `!On Testing _Materials for Cavitation
Dam-' 'age Resistance, '! TecliniCal Rep-oft: 233-3, Hydrohanties, ina.,
Dec.;1963.-[8] _Plesset; M. S.:and Devine, R. "E'.; _"Effect of -Exposure Time on Cavitation Damage,"
Journal of Basic Engineering, 'Transactions, American Society of Mechanical.Engi-, _ . . , . .
neers, VOL 88D, No 4; 1966 pp 691705:
[4I Hobbs,L J M, "Experience Vvitli a 20-kc Cavitatioiii14.rosiOnItest;","E'rosioh by Cavitation or Impingement, :ASTM :,STP 408,-Anierican,Society,.for, Testing and
Materials, pp ,159-;1 , '
,
: [5] smith, Allen 'Kent; It. p:;,and Armstrong, R. L., `!Erosion'nf Steain-Tdibind Blade
Shield Materials, Erosion' by CaVitatioriarlrOingement, AST M STP-498, American -Society for.Testing:and Materials-,-1967; pp 125-158 ')1
161 Reymann, E J:, "On the Time Dependence of the Rate of;Er6aion,Due toImpinge-ment or Cavitation, " Erosion.`-by Cavitation Imzngemeiit, ASTM ,S7'P 408;
Engel, Olive, "A Model for Multiple-Drop-Inipact Erosion of 'Brittle Solid,"
JPL Technical Memorandum 33354, California.Institute of Technology; Jet ,ProL ptitSi6n-LabOratory;,15 June 1967; p0. 129-144:
Georgievskaka, E P and Stumpf, V. M.; "Experimental Investigations into
'Erosion Characteristics of Propellers for Large Tankers," Proceedings of Eleventh , Intertiational:Towing Tank Conference, Tokyo, 'Japan, 1.966, pp,-,:242t244.).,
v-an_f4ainnieren W..P. A "Testing Screw PrOPellerS, in kCavitat,i8n Tunnel with Coiittollahle:VelOoity Distribution over the SeieW-Dil'(,"+Trania0ioni,7-SoOiety of NiVaLArchitects and Marine Engineers, Vol. 63.;-1955,. Pp. -767499.;
[16] yaiiJVIanen,J. D "Bent Trailing Edges of Propeller:Bladesof High Powered Single
Ogre* Ships," Cavitation and -Hydraulic Machinery, Prikeedirig§ _
-
Seli'dai;laii;'1962, PP. 461474.'
' [11] A.r.p.n Manen, J: p., "On the Usefulness of a Test Willie, Propeller Model in a
Cavita-tion Tunnel With a Simulated Non-Uniform Flow, .,Sginpoiiiiin on Testing
Tech-- Cal:datum' Research, Trondheim, 'NeriiraY;1967,'.-Proccidings, Vol 2,
. ;..;; 3'
'pp; - "
[1.]F:Vran der: NVilleTlf.; "On the GroWth ,cif- Nuclei. and the Relatei:L:Scaling.Factbri in. . _ r, Cavitation Inception;:';' Fourth Syni,Poszuni on Is, ratdl.HYdrOdynarizies Washington
. ,
,D
C,F1962, .[le] Van Wijiigaarddn, L, fOn-the Collective C011apiecifittalie NuMbei-of Gas Bubbles in Wateri".Eleventhlnternational Congress of Applied iledicinics;IMunich, Germany,
,
1964, Proceedings, : :1966, pp. 854-861.
:-.[/4] Hansen, B. W:,and Rasmussen, R. E. H., 'Cairitatioii DarnEige Eiperinients in a_
Rotating Disk Appliratiii Espedialiy with Regard: to-the ' Gas' Content of Water," Journal of ',Ship Research, Vol. 12, No. 2, June. 1968, pp. 83-88:.
[15] Godard, 'H: p.-et al,. the Corrosion of Light Metals, New liork, 1967, p. 9. [15] Thiruirengadini; AS.; f.`The Concept of Erosion Strengthl". Erosion by Capitation or
--, I rripingement, AST M- P.408, American Societi-for TeSting anti Materials, 1967,
pp. 22-41-- - .
.
J. Z Lidttmaizi '(thr'itteit'disctiskin) The physical featUre§ Of the erosion
of the rotating disk specimens f obserVed and described by the author are
identical to those observed in other high velocity erosion systems, both
in the laboratory and in service . 2-7' These systems included rotating disk and axial flow systems, pump impellers, propellers, and water'brakes. It
was ' suggestad2.4 - that the features of the eroded region are related to
re-entrant._,flow§.assOciated with the cyclic cavitation cloufl, and -scouring
of the microimpact-clarnagedsinface by the high velocity. re-entrant flow..
The :directional aspects of the damage in the rotating Systems natiy be
related to the 'centrifugal .flows and resultant deviation of the cavitation - cloud flow regime from the tangential direction. The radial flow component
then would cause rd-entiant ilOW in a. direction which would result in the/.1
observed crevices and scoured Margin's. Such directional deviations are not observed iitthe damage. associated with an -axial flow system.P. This
hypothetical analysis suggests the need for further study of the highly
complex flow cOnditionS, associated with high speed and -vibrator ir cavi-tation,. Such investigations would Contribute tO, a, more complete
under-standing of the erosion process, which currently is associated primarily With the collapse of the cavitation bubble and resultant microimpact
damage.
1 Materials engineer, Naval Applied Science Laboratory, Brooklyn, N. Y. 11251.
Lichtman, J. Z., 'Possible Contributions of Reentrant Flow te Cavitation Erodion,"
'ASME Paper 62HYD73, American Society of Mechanical Engineers.
-- 3Lichtman, J. Z. and Weingram, E. R., "The Use of a Rotating Disk Apparatus
Determining Cavitation Erosion Resistance of Materials," ASMi' Symposium on
Cavita-tion Research Facilities and Techniques, May 1964, American Society of Mechanical
Engineers, 1964, pp. 18.5-196.
Lichtman, J. Z., 'Discussion, " Transactions, Society of Naval Architects and Marine-Engineers, Vol. 73, 1965, pp. 279-281.
t'Lichtman, J. Z., "Discussion," ASME 1967 Cavitation Forum, American Society of Mechanical Engineers, pp. 40, 41.
°Kailas, D. H. and Lichtman, J. Z, "Cavitation Erosion," Environmental Effects on
Polymeric Materials, Vol 1, Rbsato, D. V. and Schwartz, R. T., eds., Interseience, New York, 1968, Chapter,2, pp. 223-280.
- Morgan, W. B. and Lichtman, J. Z., "Cavitation Effects on Marine Devices/' Cavitation State of knowledge, Robertson, J. M. and Wislicerms, sG. F., eds , American Society of Mechanical Engineers, New York, 1969, Fie 10, p. 206.
