Effects Leading Edge Sweep on Cavitating Characteristics ...downloads.hindawi.com/journals/ijrm/2001/613563.pdf · 398 A.J.ACOSTAETAL. ion Inducer0 Inducer F30 "InducerB50[._[__ Inducer0--.-InducerB50.149.8

International Journal of Rotating Machinery, 7(6): 397-404, 2001Copyright (C) 2001 Taylor & Francis1023-621X/01 $12.00 + .00

Effects of Leading Edge Sweep on the CavitatingCharacteristics of Inducer Pumps

Allan J. Acosta, Yoshinobu Tsujimoto,2 Yoshiki Yoshida,2 Seiji Azuma,2

and Paul Cooper3

California Institute of Technology, Mechanical Engineering, Pasadena, California, USA;2Osaka University, Engineering Science, Osaka, Japan; Fluid Machinery Research, Inc.,Titusville, New Jersey, USA

It is well known that leading edge sweep has a favor-able effect on the cavitation of turbomachines. However, themechanisms of the improvement have not been made clear.It has been shown that the lift and the drag on a cavitatingswept single hydrofoil can be correlated fairly well basedon the velocity component normal to the leading edge. Inthe present paper, such correlations for swept cascades arederived and the results are examined, neglecting the full ge-ometrical effects of the inducer rotor. It is shown that thecorrelations can simulate the developments of various typesof cavitation, including alternate blade cavitation, rotatingcavitation, and cavitation surge. This result is based on theobservation that the steady cavity length, as well as the devel-opments ofvarious types ofcavitation, is fairly well predictedby the correlation.

Keywords Inducer, Cavitation, Leading edge sweep, Cross flowplane

INTRODUCTIONThe effect of inclining the leading edge of a wing to the on-

coming flow, called sweep, is familiar to us all from seeing mod-ern high-speed aircraft. Sweep is usually backwards (the anglebetween the leading edge and the flow is less than 90), but it canbe forward as well. The primary effect of sweep in aerodynamics

Received in final form on 10 July 2001.This study was triggered by a discussion at US-Japan Seminar spon-

sored by NSF and JSPS. Experimental results were obtained under thesupports of SNECMA, division SEP and the Ministry of Education,Japan. These supports are gratefully acknowledged.

Address correspondence to Allan J. Acosta, California Institute ofTechnology, Mechanical Engineering, Pasadena, CA 91125. E-mail:[email protected]

is to forestall the influence of compressibility. Similar problemsoccur in gas turbine blading described in detail by Wadia et al.(1998). But in normal industrial practice, the leading edges ofconventional pumps often tend to be normal to the oncomingstreamlines as seen in a meridional section, i.e., the view of theimpeller blades projected onto a radial plane. In that case, wewould say there is no sweep; this is a common feature of manypumps. The occurrence of cavitation in liquids causes specialproblems for pumps and, in many cases, the inlet of a pump isfitted with a special rotor to operate satisfactorily with extensivecavitation and it is usually called an inducer pump This rotor

may be an integral part of the main pump to follow or a sepa-rate device. Typically, this is an unshrouded rotor with small tipclearance with, as a result, cantilever-mounted blades. Cavita-tion is a notoriously unsteady phenomenon resulting in highly-fluctuating blade loadings and, as a consequence, most designershave intuitively inclined the blade leading edge backwards (inthe meridional plane) to minimize structural and vibration loadson the unsupported leading edge. But there also can be con-

sequences for cavitation performance, as many designers havenoticed that such swept-back leading edges are able to operate anlower inlet pressures than radial ones, thereby achieving a lowercavitation number or a higher suction specific speed. In fact, oneof us (Cooper, 1973) has patented an impeller design in whichthe leading edge in the meridional view (see Fig. b) is highlyinclined backwards--or forwards with the blade elements simi-larly inclined. Most inducer pumps, however, traditionally haveblade elements (or blade generators) that are radial or nearlyso. Thus, these blades are somewhat like that of a conventionalwing. Interestingly enough, experimental measurements on a

single foil in cavitating flow have been made (Ihara et al., 1989)in which a simple scaling rule, the same as for compressibleflow, collapses the cavitating performance data as a function of

See, e.g., the article "Inducer Pumps--An Apercu," ISROMAC IV, section

B, page for a recent review of these pumps.

397

398 A.J. ACOSTA ET AL.

ion Inducer 0Inducer F30

"Inducer B50[ ._[__

Inducer 0

--. -Inducer B50

.149.8

(a) The end view (b) The meridional view

FIGURE 1Geometry of leading edges of test inducers. (The leading edgeprofiles of F30 and B50 are shown in (a) only for one blade.)

sweep angle surprisingly well. They find that (from their con-clusions) the lift drag ratio for highly swept foils is superior tothe unswept case at the same and even lower cavitation number(all at constant angle of attack) and that the flow velocity normalto the foil should be selected to calculate a cavitation number.

With this background, we want to propose similar rules forinducer pumps based on the idea of a cascade of straight bladessimulating the flow in the tip region of an inducer of radial bladeelements with an axial inflow. The leading edges of the bladesin the meridional view can be inclined backwards or forwardsas shown in Fig. l(b). The end view of the inducer shown inFig. (a) shows the leading edge swept backwards or forwardsbut subtending a relatively large angle.

GEOMETRY OF THE SWEPT CASCADEThe cascade is shown in Fig. 2; several views are shown

which are needed for clarity. The upper or plan view of the cas-cade shows straight uncambered blades spaced s apart along thecascade axis (the plane normal to the inducer axis in Fig. (b)).The blades are inclined at blade angle fl with respect to this axis.The leading edges of these blades are shown in the meridionalview inclined at angle 3 from what would be a radial line inFig. (b). Let us select two points on one blade, O and A. Cor-responding points in the meridional plane lie along 1-1 spanninga vertical height r/s (in the radial direction). These points and atthe next blade O’, A’ are observed in the true view of the bladeleading edge defined by cut B-B. The bold line OA is the trueview of the leading edge. This line makes an angle ) to the planeofthe inducer axis, namely, the projection of O-O’. The adjacentblade in this true view is O’-A’. The plane normal to OA,and to the plane of the paper is the cross flow plane. Note thatthe line O’A’ is hidden from view by the first blades. Now in thetrue view plane, BB, project a normal to OA from O’ ending atP, a point on the leading edge of blade B1. Point P also appearsin the plan view on the leading edge of blade B1. Imagine nowwe progress from P normal to the leading edge along blade

B2

A

leadin, edge

lyl

i/1 Inducer axis

Meridional view AA

FIGURE 2Geometry of swept cascade.

B2

in the cross flow plane, CC, until we are underneath the normalto the next blade, B2, at point Q. From Q we move to point O’on blade B2 seen in the plan view. The cross flow plane in theplan view may be identified by the points P, Q, O (a portionof it is shown, cross-hatched for clarity).We will be concerned with the flow velocities and the effective

cascade geometry in this cross-flow plane. Before completingthe definitions of the cross-flow cascade geometry, let us de-scribe the velocity components. The velocity approaching theinducer seen in Fig. (b) is presumed to be purely axial; relativeto the rotating inducer the flow speed is V1 and is inclined tothe blades with incidence angle o shown in the plan view ofFig. 2. This velocity vector lies in the plane of the inducer inFig. 2. There is a component of V1 that is normal to the cascadeblade, Vn, and a component tangential to the blade surface Vtalso shown in Fig. 2. The tangential component is resolved intocomponent V. normal to the true view of the leading edge andVp parallel to it. We see in the cross-flow plane (Fig. 2, sectionCC) component Vc approaching a cascade but one characterizedby a new spacing, Se, and a new blade angle, fie. The normal dis-tance between the blades B1, B2 shown in the cross-flow andplan view is of course the same, i.e.,

ssinfl SeSinfle, or tic sin-((S/Se)sinfi). [1]

LEADING EDGE SWEEP AND CAVITATION CHARACTERISTICS OF INDUCER PUMPS 399

The sweep angle X is constructed in the true view plane from

or Z tan-1 [2](Ts tan 6/sin/3) \ tan

The effective spacing is determined from its projection P-O’ inthe true view plane and the normal distance above to get

Se v/(S COS/3 sin/)2 -Jr- (S sin 13)2. [3]

Note that s < S. The true thickness of the blades is t; the ratiot/s is an important geometric parameter governing cavitation.Clearly then the effective thickness-spacing ratio (t/S)eff is

>_ [4]S eft S

Thus the cross-flow geometry is blunter than the normal flow.For a leading edge in the shape of a wedge of included angle 0,it follows that

0e= tan-1 ( tan0 ) [51

If, as is the case with many small commercial inducers withleading edges machined on a lathe, 0 ft. Finally, there is theflow incidence angle in the cross flow plane for which

ot. tan-l(v/Vc), or Otc tan-l(tanot/sin;k), [6]

so that cc > or. This completes all the geometrical features inthe cross-flow plane.

CAVITATION SCALINGLet the pressure in the cavity be Pv and the velocity there be

Vk. Then in the usual way from the Bernoulli Equation, i.e.,

and, based on this velocity, we can define a cross-flow cavitationnumber as

(pl- p l(pv /2)total

And on the cavity we will have

[11]

Vkc- VctotalV/1 -1I- kc. [12]

These definitions must make the true velocity on the cavity pre-cisely the same requiring that

Vff V2c + Vp2 or V12(1 + k) (Vc + Vff)(1 + kc)+ Vp2. [13]

After substituting these definitions we have the simple result

kc k/(cos20 sin2 * + sin2 [14]

Note that if Z -+ 7r/2 (no sweep), k -+ k.. For typically smallincidence angles ot << 1,

kc (k/sin2 Z)(1 0(oe2)) , k sin2 Z, [15]

which is the relation we will use. The equivalent formula for a

swept-isolated wing quoted by Ihara et al. is (in our notation)

kc k sin2 Z, 16]

the slight difference arises because our cross-flow velocity isparallel to the chord.

The effect of sweep appears through two mechanisms. One isthrough the change of cascade geometry in cross flow as shownby Eqs. (3)-(6) obtained in the preceding section. The otheris through the cavitation scaling as shown by Eq. (16). It hasbeen shown by two-dimensional linear analysis such as Acosta(1955) that the cavity length, and hence the cavity development,is a function of k/2ot. If we combine Eq. (6) and Eq. (16), weobtain

Pv 1P P___}_l + V12 + V? [7]

we define the cavitation number

k Pl PpV2/2

[8]

and so we can say

k/2oekc/2ot [17]

sinZ

If we increase the leading edge sweep, ,k is decreased. Then

kc/2otc is increased and hence the cavity length 1/s (lc/se)(cos/3/cos fie) is decreased. This can be the major reason whythe cavitation performance is increased by simply sweeping theleading edge.

Vk-" VI/’I + k [9]

in the physical plane. The total velocity approaching the cascadein the cross-flow plane is

Wctotal //V(.? + gr/2 [101

COMPARISONS WITH EXPERIMENTS ANDTHEORETICAL CALCULATIONS

In order to study the effect of leading edge sweep, system-atic experiments were carried out at Osaka University under thesupport of SNECMA, division SEE Three inducers tested herehave helical blades with the same camber line with straight partnear the leading edge. Forward and backward swept inducers


TABLE IGeometry of test inducers

Inducer 0 B50 F30

WithoutSweep sweep

Backward forwardsweep sweep

47.3 deg. 25.0 deg.

Sweep angle . at tip 90 deg. 35.4 deg. 49.2 deg.Number of blades 4 4 4Tip diameter, D1 149.8 mm 149.8 mm 149.8 mmInlet tip blade angle, fit] 7.5 deg. 7.5 deg. 7.5 deg.Outlet tip blade angle,/3t2 9.0 deg. 9.0 deg. 9.0 deg.Hub/tip ratio at inlet 0.25 0.25 0.25Hub/tip ratio at outlet 0.51 0.51 0.51Solidity at tip 2.97 2.44 2.86Tip clearance 0.5 mm 0.5 mm 0.5 mmDesign flow coefficient, 4d 0.078 0.078 0.078

were produced by cutting back the straight part of an unsweptinducer so that the inlet blade angle is not changed. Thus, all ofthe inducers have the same inlet and outlet blade angle. Dimen-sions of the inducers are shown in Table I. The basic design is thesame as for the LE-7 LOX turbopump inducer except that threeblades are employed for LE-7. Inducer 0 is without sweep andhas a straight radial leading edge. Inducer B50 is produced bycutting back the leading edge by about 47.3 as shown in Fig. 1.Inducer F30 is produced by offsetting the leading edge by 35 at

0.4

0.35

0.3

0.25

015

0.05

0.02 0.04 0.06 0.08 0.1Flow coefficient (I)

0.12

7,5 6 5 4 3 2cz (degree)

FIGURE 3Non-cavitating performance of test inducers.

1.5

0.5

o=2.9 (=0.080)],I: o=4.0 (4=0.060)

0 0.05 O. O. 15 0.2 0.25Cavitation number, k

(a) Inducer 0

----0.65s

--- 1.5

0.5

Unsteady Z,,: a =2.9 ( =0.080)IT],==: a=3.5 (=0.070)0,0: z =4.2 ( =0.058)

0.05

<1--0.65s

0.1 0.15 0.2Cavitation number, k

O..5

(b) Inducer B50

A,,: =3.0" (=0.079)Unsteady<:}" C],l: =3.3 (=0.074)cavitation (,}: =4.0" =0.060)

0 0.05 0.1 0.15 0.2 0.25Cavitation number, k

(c) Inducer F30

0.65s

FII3URE 4Plot of cavity length against k. Open symbols show shortercavity length for the case with alternate blade cavitation.

the root and then giving forward sweep by 25. The tip/leadingedge corner is rounded with radius 4 mm. The leading edgecurve is obtained by shifting the circumferential location of aninvolute curve with a base radius 26 mm (this produces a sweepwith 85.3) proportionally to the amount of sweep. The bladethickness is 2 mm and the suction surface near the leading edgeis filed to wedge angle 2.75 with the leading edge radius of0.2 mm.

Figure 3 shows the non-cavitating characteristics of the in-ducers. Nominal incidence angle at the tip (or / tan-] 4)is also shown. As expected, three inducers have nearly the samenon-cavitating performance for q > 0.06 (oe < 4 deg).


Figure 4 shows the plot of cavity length at the tip 1/s againstthe cavitation number k for the three inducers. For all inducersand all incidence angles shown, alternate blade cavitation (inwhich cavity length differs alternately) starts to develop when thecavity length exceeds about 65% of the spacing s. The cavitationbecomes unsteady for the condition with k smaller than thatwith the data point. These observations agree fairly well withthe theoretical findings by Horiguchi et al. (2000).

In most cases, unsteady cavitation starts to occur when thelength of the shorter cavity exceeds 65% of the spacing. Asexpected, the cavity develops faster for the cases with a largerincidence angle oe.

Figure 5 shows the plot of cavity length against k/2oe whereoe is the nominal incidence angle at the tip. As expected from lin-earized analysis, the development of cavity is nearly the same forall the incidence angles. The comparisons among three inducersclearly show that the development of a steady cavity is signifi-cantly delayed by giving both forward and backward sweep.

Neglecting all the difference of the cascade geometry inthe cross flow plane, the cavity length l/s is replotted againstkc/2cc in Fig. 6. Nominal values at the tip have been usedfor replotting. We find that the alternate blade cavitation startsto occur kc/2o&.0.9 and it shifts to unsteady cavitation at

k./2o&. 0.4.

2

.o.5

0 0.5 1.5 2k/2 o =kc/2 o c

(a) Inducer 0

2.5

<1---0.65s

1.5

0.5

Unsteady /X,&: e =2.9" q} =0.080)cavitation [’-J, I: 1 =8.5 4} =0.070)

0 058)

0 0.5 1.5 2 2.5k/2 z

(b) Inducer B50

<F--O.65s

2A,&: e =8.0 (4} =0.079)Unste.a,_ I--I,.’. {I =8.8 (4}=0.074)

1.5 cav’tal 0,0:!=4.0 =0.060)

0.65s

0.500 0.5 1.5 2 2.5

k/2

(C) Inducer F30

FIGURE 5Plot of cavity length against k/2ot.

1.5

().5

A I, =2 9 4} =o 080)

0 0.5 1.5k/2 ot =kc/2 ot c

(a) Inducer 0

<I--0.658

1.5

0.5

Unsteady ZX,,I: ot=2.9" (4}=0.080)n,n. =3.5 4} =0.070)O,O: o=4.2" (4}=0.058)

0.5 1.5

kc/2 c

(b) Inducer B50

<1"--S

<F--O.65s

a,,: =a.o" {,=o.o79} ]1Unstead Fl,n: ot=8.,:3 q}=0.074) IIcavitati

t--.

<F--O.65s

0.5 1.5 2kc/2 o c

(c) Inducer F30

FIGURE 6Plot of cavity length against kc/2otc.

2.5


TABLE IICascade parameters in physical and cross flow plane

Inducer 0 B50

Sweep angle ,k at tipInlet tip blade angle,Solidity at tip, C/sLeading edge wedge angle, 0Effective blade angle, fieEffective solidity, Ce/SeEffective leading edge wedge

angle, 0e

F30

90 deg. 35.4 deg. 49.2 deg.7.5 deg. 7.5 deg. 7.5 deg.2.97 2.44 2.862.75 deg. 2.75 deg. 2.75 deg.7.5 deg. 12.8 deg. 9.9 deg.2.97 2.40 2.842.75 deg. 4.74 deg. 3.63 deg.

The present result shows that the delay of cavity developmentcan be explained by the cross-flow effect. The secondary flowcaused by the centrifuging of blade boundary layer should bequite different for forward and backward sweep. However, thedelay of cavity development is quite the same for forward andbackward sweep as shown in Fig. 5 and it can tbe explained bythe cross-flow effect as shown in Fig. 6. This fact shows thatthe cross-flow effect is more important than the secondary-floweffect caused by fluid viscosity.

It has been shown (Tsujimoto et al., 1998) that various kindofunsteady cavitation depends only on the steady cavity lengthor equivalently on k./2oe.. In this respect, the present correctionwith k./2oe, explains not only the steady cavity development,but also the onset of unsteady cavitation for k./2c. < 0.4.

Table II shows the cascade parameters in the cross flow plane.Comparisons in Fig. 6 have been made by neglecting the dif-ference in the cascade geometry in the cross flow plane. Toexamine the effect of the geometrical difference, calculationsare made by using a singularity method based on a linear closedcavity model (Horiguchi et al., 2000) on thin flat plate cascades.Figure 7 shows the cavity length in the cross flow plane. It is

1.5

0.5

0

"" Inducer 0:) (C) Inducer BSOb /k Inducer F30

0 2 4 5 63kc/2 a c

FIGURE 7Calculated cavity length in cross flow plane. The results areshown for cross flow cascade geometries of Inducer 0, B50,F30. Two cavity lengths are shown where alternate blade

cavitation was found.

shown that the geometric effects also suppress the developmentof the cavity when the cavity is shorter than the spacing. Figure 8compares the exact cavity length in the physical plane estimatedfrom the exact cascade geometry in the cross flow plane with theapproximate cavity length estimated from the original cascadegeometry in the physical plane. Although the difference in thecascade geometry cannot be ignored, the major effect of sweepcomes from the k./2oe, effect. Unfortunately the agreement with

Inducer 0

1.5BSO (original geometry)

0 BSO (cross flow geometry]

0.5 8 "OOOOoooooooooeOooo

00 2 4 5

<1---8

<}-- 0.65s

3 6kl2a

(a) Inducer 0 and Inducer B50

2Inducer 0

1.5 L F30 (original geometry),., F30 (cross flow geometry)

r- ::........_ o.6s

0 2 3 4 5 6k12 a

(b) Inducer 0 and Inducer F30

FIGURE 8Calculated cavity length in physical plane, exact value fromcross flow cascade geometry, and approximate value from

original cascade geometry in physical plane.


FIGURE 9Geometry and the flow with wedged leading edge cascade.

the experiment in Fig. 5 is not good, perhaps caused by 3-Deffects.

The results in Fig. 8 show that the sweep does not affect thecavity development largely for cavities longer than the spacing.This is caused by the canceling of the favorable effects of kcby the deteriorating effects of the cross-flow cascade geometryas shown in Fig. 7 for lc/sc > 1. Analyses based on the cor-relation kchoke 0(/3 Or) for the choke cavitation number ofa thin-bladed cascade show that the sweep does not affect thechoke cavitation number and is caused by the cancellation asmentioned.

To examine the effect on the choke cavitation number forthe cascades with wedged leading edge cascade, calculationsare made based on a linear theory by Acosta for wedged lead-ing edge cascade as shown in Fig. 9. The cavity is assumed tostart from the end of the leading edge wedge as shown in thefigure. Figure 10 shows the choke cavitation number obtained.The results are not shown for a larger angle of attack for whichthe cavities will start from the leading edge tip. The choke cav-itation number is clearly decreased by the sweep and the effect

0.06

0.05

0.04

0.03

0.02

0.01

,,A_

O X

0 0.5

0.13

0 InducerB50

A Inducer F30

o

1.5 2 2.5a [deg.]

0.12 0.11 0.1 0.09Flow coefficient

FIGURE !0Choking cavitation number evaluated by an unpublished

theory by Acosta for wedged leading edge cascade.

is more significant for the cases with smaller angle of attack.This shows that the leading edge geometry significantly affectsthe cavity development. Unfortunately, the breakdown cavita-tion number could not be determined in the series of experi-ment due to the limitations of the experimental apparatus. Care-ful experiments are needed to determine the effect of sweepon the choke cavitation number, paying attention to the lead-ing edge geometry and the location of the cavity detachmentpoint.

FURTHER REMARKSApplying this type of development to the kind of "sweep"

one obtains by leaning back the blade elements from the typ-ically radial orientation (thereby not changing the view fromthat of Inducer 0 in Fig. (a)mi.e., the projection of the leadingedge of the leaned-back blade onto the end view of Fig. (a)is still a radial line) should reap the benefits on the breakdownvalue of cavitation number computed and plotted in Fig. 10.Adding a cut back to this geometrymin view of the devel-opment provided in this papermay show further im-provement.

CONCLUSIONS1. A cross flow model is proposed to estimate the effect of

leading-edge sweep on the cavity developments in inducerpumps with predominantly radial blade elements. Correla-tions are derived for geometrical parameters of the cascadein the cross-flow plane and for cavitation scaling.

2. The correlations are applied to the cases of forward and back-ward swept leading edges of inducers neglecting the detailedgeometrical effects of the full impeller. Beneficial effects ofboth forward and backward sweep for cavitation are shown.The correlation using kc/2o, explains the cavity develop-ments observed (e.g., cavity length, alternate blade cavita-tion, and unsteady cavitation) fairly well. This shows that theeffect of sweep is mainly caused by the cross-flow effectsproposed in the present paper with minor contributions ofsecondary (viscous) flow effects.

3. These effects ofsweep are further studied with a linear closed-cavity model in cascades without blade thickness. A favor-able effect of sweep is found (causing shorter cavities) forcavities shorter than the blade spacing, but this effect dis-appears for longer cavities and in the choked cavitycase.

4. The effect of a blunt leading edge blade shape, specificallya wedge, on the choked cavitation number is also examinedwith linear free streamline theory. Unlike the case of zerothickness blades, there is a favorable effect of sweep on thechoked cavitation number which lowers the choked cavita-tion number and increases the resultant suction specific speedof the inducer pump.


5. The analysis and results of this paper apply to blade "sweep"as generated largely by cutting back the leading edge ofa typical radial-element-blade inducer. Similar and perhapseven greater improvements may be possible for leaned-backblade elements that are also cut back (at the leading edge)beyond this amount of lean.

NOMENCLATUREk inlet cavitation number (p pv)/(pV2/2)k. cross flow cavitation number

(p pv)/(pVctotal2/2)Pl static pressure Upstream of impellerp pressure in the cavity (presumed vapor)Pt inlet total pressures spacing of blades along the cascade axis

true blade thickness(t/S)eff effective thickness/spacing ratio

V1 inlet relative velocitygctotal (go2. + V2)1/2, total cross flow velocity in cross flow

planeoe angle between inlet relative velocity V1 and the blade/3 angle of blades from cascade axis6 angle from radial direction of leading edge in merid-

ional plane leading edge in the meridional view

true angle of blade leading edge-if ) rr/2, leadingedge is radial (In aerodynamic use the complement iscalled r.)liquid density

REFERENCESAcosta, A. J., 1955, A Note on Partial Cavitation of Flat Plate

Hydrofoils, Caltech Hydro Lab. Report No. E-19.9, Oct.Cooper, E, 1973, High Flow Pump Impeller for Low Net Positive

Suction Head and Method of Designing Same, United States Patent3,737,249.

Horiguchi, H., Watanabe, S., Tsujimoto, Y., and Aoki, M., 2000, A The-oretical Analysis of Alternate Blade Cavitation in Inducers, ASME,J. Fluids Engineering, vol. 122, pp. 156-163.

Ihara, A., Watanabe, H., and Shizukuishi, S., 1989, ExperimentalResearch on the Effects of Sweep on Unsteady Hydrofoil Loadingsin Cavitation, ASME, J. Fluids Engineering, vol. 111, pp. 263-270.

Sasaki, T., and Breugelmans, E, 1998, Comparison of Sweep andDihedral Effects on Compressor Cascade Performance, ASME,J. Turbomachinery, vol. 120, pp. 454-464.

Tsujimoto, Y., Watanabe, S., and Horiguchi, H., 1998, Linear Analysesof Cavitation Instabilities of Hydrofoils and Cascade, US-JapanSeminar Abnormal Flows in Turbomachinery, Osaka, Japan.

Wadia, A. R., Szues, E N., and Crall, D. W., 1998, Inner Workingsof Aerodynamic Sweep, ASME, J. Turbomachinery, vol. 120,pp. 671-683.

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttp://www.hindawi.com Volume 2010

RoboticsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation http://www.hindawi.com

Journal ofEngineeringVolume 2014

Submit your manuscripts athttp://www.hindawi.com

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

SensorsJournal of


Modelling & Simulation in EngineeringHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks


Documents

Effects Leading Edge Sweep on Cavitating Characteristics ...downloads.hindawi.com/journals/ijrm/2001/613563.pdf · 398 A.J.ACOSTAETAL. ion Inducer0 Inducer F30 "InducerB50[._[__ Inducer0--.-InducerB50.149.8