I ADRIB 433 PROPELLER DESIGN OF US NAVY'S SMATH (SNALL WRTERPLANE 1/1I ARER THIN HULL) T-AGOS 19(U) DAVID Ni TAYLOR NAVAL SNIPI RESEARCH AND DEVELOPHENT CENTER BEEHESDA ND7 UNCLKSSFE KIN TET AL. AUG 97 DTNSRDC-87/6034 F/G 13/160 N

T jIK~t.i,1flhI== _13151*1 r11111 ~11111 1*8tflht~ ~ II

- sI _________ 1I 125 1114 14- 1W~

C

David W. Taylor Naval Ship Research and Development CenterBetw , MD 20084-5000.

FILE COPY

in DTNSRDC-87/034 August 198700 Ship Hydromechanics Department

Research and Development Report

< "

Propeller Design of U.S. Navy's SWATH T-AGOS 19byKi-Han KimArthur M. Reed

0 Presented at the Third International Symposium onCPractical Design of Ships and Mobile-Units (PRADS '87)

Trondheim, Norway, 20-25 June 1987

in- DTIC-,

z ItEECTESCT0 1 9 87f0 S-

ci

0

0

z Approved for public release; distribution is unlimited.

.......

SECURITY CLA551IIATION OF THIS PAGuE

REPORT DOCUMENTATION PAGEI&. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS

UNCLASSIFIEDZa. SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/AVAILABILITY OF REPORT

2b DECLASSIFICATION /DOWNGRADING SCHEDULE Approved for public release;distribution is unlimited.

4 PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)

DTNSRDC-87/034

64. NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATIONDavid W. Taylor Naval Ship (if applicable)Research & Development Center Code 1544

6C ADDRESS (City. State, and ZIP Code) 7b ADDRESS (City. State, and ZIP Code)

Bethesda, MD 20084-5000

Ba NAME OF FUNDINGISPONSORING Sb. OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION (If applicable)

Naval Sea Systems Command SEA 5046Sr ADDRESS (City. State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERS

PROGRAM PROJECT TASK WORK UNITCrystal City ELEMENT NO NO NO ACCESSION NOWashington D.C. 20362 64567N WRlO511 S1803528 506007

I I TILE (Include Security Classification)

PROPELLER DESIGN OF U.S. NAVY'S SWATH, T-AGOS 19

12 Pf RSONAL AUTHOR(S)

Kim, Ki-Han and Reed, Arthur M.13a TYPE OF REPORT 13b TIME COVERED 114 DATE OF REPORT (Year, Month, Day) I~S PAGE COUNT

Final I FROM 486 _TOAL87 1987 August I24'6 SUPPLEMENTARY NOTATION

Presented at the Third International Symposium on Practical Design of Ships and MobileUnits (PRADS '87). Trondheim, Norway, 20-25 June 1987

I? COSATi CODES 18 SUBJECT TERMS (Continue on reverse if necessary and identify by block number)

FIELD GROUP SUB-GROUP Propeller Design, SWATH, Lifting-Surface,

- I Lifting-Line, Blade Surface Cavitation

19 ABSTRACT (Continue on reverse if necessary and identify by block number)

'This paper presents the design process and model evaluation of a propeller for the U.S.Navy's research vessel, T-AGOS 19, with a SWATH (Small Waterplane Area Twin Hull)Configuration. The design procedure is discussed in detail including considerations ofcavitation and propeller-excited vibratory forces. The design objective was to minimizethe blade cavitation at various operating conditions. Model experimental resultsincluding powering and cavitation characteristics are presented which confirm the

validity of the design process.

20 DStRiuTON'AVAILABIUTY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATIONIlJN(LASS4FIOUNLIMItED 0 SAME AS APT 0OTIC USERS UNCLASSIFIED.

.'2 NAME OF RESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL

l'-u-- lip an 111100Ahtt 1 R- (202) 227-4308 1 Code 1544DO FORM 1473,84 MAR a) APR edition may be used until emausted SE(URITY CLASSIFICArION OF T-4S PAGE

All other editionS are obtolete UaAS TE

UNCLASSIFIE

Page

ADMINISTRATIVE INFORMATION ................................ v

1.• ABSTRACT ............................................................ 1

2 INTRODUCTION ................... ... .. .............. * . ............ . 1

3. PROPELLER DESIGN REQUIREMENTS .................................... 2

4. WAKE SURVEY AND PROPULSION EXPERIMENTS USING STOCK PROPELLERS ....... 3

5. PROPELLER DESIGN ...................... ............. ... . ........ 4

5.1 DESIGN CONDITION AND PROCEDURE ............................... 4

5.2 ESTIMATION OF FULL-SCALE WAKE .................................. 5

5.3 DIAMETER OPTIMIZATION ......................................... 6

5.4 CIRCULATION DISTRIBUTION ...................................... 6

5.5 THICKNESS AND CHORD LENGTH DISTRIBUTION ....................... 7

5.6 LIFTING-LINE DESIGN ........................ *.................. 7

5.7 LIFTING-SURFACE DESIGN ........................................ 8

6. PERFORMANCE PREDICTION .............................................. 8

6.1 PREDICTION OF DELIVERED POWER AND RPM ......................... 8

6.2 UNSTE.ADY BEARING FORCES AND MOMENTS ........................... 9

7. EXPERIMENTAL RESULTS AND ADJUSTMENT OF BLADE PITCH .................. 9

7.1 OPEN-wAreR AND POWERING EXPERIMENTS ........................... 9

7.2 ADJUSTMENT OF BLADE PITCH...................................... 9

7.3 CAVITATION TESTS .............................................. .11

8. DISCUSSION .......................................................... 12

9. ACKNOWLEDGEMENTS .................................................... 12

10. REFERENCES ........................................................... 1 . .-1

iii

* -

Page

1. Sketch of SWATH T-AOS 19 Configurationn I

2. Schematic Drawing of Stabilizing Fin and Propeller Guard Fins ...... 2

3. Circumferential Variation in the Axial, Tangentialand Radial Velocity Components ......................................... 3

4. Comparison of Measured Model Nominal Wake andPredicted Full-Scale Nominal Wake ..................-.............. 5

5. Propulsive Efficiency at Different OperatingConditions for Different Diameter .................................... 6

6. Circulation and tanB i Distributions ............................... 6

7. Radial Distribution of Thickness ........................ 000...... 7

8. Radial Distribution of Chord Length.... ............ 0.......... 7

9. Pitch Distribution (P/D) Predicted byLifting-Surface Code, PBD-O... ..... ............................... S

10. Final Camber Distribution (fM/c) Predicted by

Lifting-Surface Code, eBD-lO ... ....... .. .. .. .. ..... ......... ... 8

Ii. Blade Outline of Design Propeller .................................. 9

12. Comparison of the Original and Final Pitch Distribution ............ 10

13. Comparison of Experimental Open-Water Performancesbetween Original and Repitched Propellers .............................. 11

14. Powering Performance of the Repitched Propeller .................... 11

15. Comparison of Velocity in the Propeller PlaneMeasured at DTNSRDC and SSPA ..................................... 12

16. Maximum Radial Cavitation Extension on Original andRepitched Propellers at Full Power, Model Wake.................... 13

iv

TAUXI

Page

1. Principal Characteristics of SWATS T-AGOS 19....................... 2

2. Powering Predictions Based on Model Propulsion TestUsing Stock roeer............................ 4

3. Effect of Radial Circulation Distributionon Propeller Performance................... ................. ...... 7

4. Geometric Characteristics of Design Propellerof SWAT! T-AGOS 19 ......... 9

5. Predicted Full-Scale Speed and RPM at Full PowerBased on Model Tests Using Design Propeller........0............. 10

6. Predicted RPM for Original Design andRepitched Propellers at Full Power ............ 10

A£IUZSTEtATIVIt lJfoinlna

This project was carried out for the Naval Sea Systems Command under

Work Request N00024-86-VR-10511. The work was performed at the David W.

Taylor Naval Ship Research and Development Center under Work Unit number

1235-680.

V

?WE~AM mizce orU.S. NTV's VMi., T-AAO 19

Ki-Man Kim and David W. Taylor U. S. A.Arthur K. Rood Naval Ship Research

and Development Center

1. S2w

This paper presents the design process and modal evaluation of apropeller for the U.S. Navy's research vessel, T-AGOS 19, with a SWATH (SmallWaterplane Area Twin Bull) configuration. The design procedure is discussed indetail including considerations of cavitation and propel ler-extcited vibratoryforces. The design objective was to minimize the blade cavitation at variousoperating conditions. Model experimental results including powering andcavitation characteristics are presented which confirm the validity ofthe design process.

2. INIT28=WU

At the request of the U.S. Naval gas Systems Commeted (IIAYSIA), the DavidW. Taylor Naval Ship Research Center (DtUSRIDC) carried out a propeller designfor a research vessel, T-AGOS 19, having a Imall-Watorplans-Area-Twin-Hull(SWATH) configuration asshown in Figure 1. Bach lowerhullI has inboard canard andstabilizer, fore and aft, res-pectively. acting as controlsurfaces. The droop anglesfor canard# and stabilizersare 20 degrees and 18 degrees,respectively, relative to thebaselIi ne. The propellers areprotected by ring-saped pro-peller guards, each with foursupporting struts (see Figure2). The details of the shipdesign are presented inReference 1. The principalcharacteristics of the shipare listed in Table 1. Themain objective was to design a_________prope ler with minitsm cavita-tion resulting from the pro--___pelleI r operat ing in thea spa--tially nton-uaiform imflo V_veloctty field in the ship'swake at different operating, i I Sketch tWSWATH I A(i)% 14 (millwIMc~coind it ions.

This paper presents the Table 1. Principal Characteristics ofdesign procedure and ratio- SWATH T-AG0S 19" a. which led to the finalfive-bladed propeller with -----------------------a 10-foot (3.05 a) diaae- Length Overall (LOA) 71.32 a (234 ft)ter, and the results ofpowering and cavitation Length at Waterline (LWL) 57.91 a (190 ft)tests with the design pro-peller. The following **c- Be&= on Design Waterline 24.38 m (SO ft)tions describe the speci-fied design conditions, the Beam (Maximum) 28.35 a (93 ft)pertinent information re-quired as input to the Draft 7.54 a (24.75 it)design problem such as wakesurvey and propulsion cha- Displacement 3366 Metric Tonsracteristics, the designprocedure, performance pre- Wetted Surface Area 2814 a2diction and powering and -----------------------cavitation test results.

3. PWUJZZ =uI=m 35K I f3

The design requirements for the T-AGOS 19 propel ler were provided byNAYSIA: propel ler geometry should be as simple as possible, minimise thecavitation on the propeller operating at full power of 800 SHP (597 kW) persaft, sustained speed a minimum of 9 knots at 50 percent of the fullI power,propel ler rotational speed of 185 rpm at full1 power in calm water, maximumpropeller diameter of 11 feet (3.35 a) with 1.5 feet (0.46 a) clearancebetween the propeller tip and the propeller guard. The depth of immersion ofthe shaft centerline in calm water is 16 feet 9 inches (5.11 a) with even keelcondition.

The full-scale propeller material will be Nickel-Aluminum-Ironze al-loy, American Bureau of Shipping (ASS) Grade 4. The propeller strengthrequirements will follow ASS 1985 Class C Ice Strengthening rules.

NO 11** Iia f We la -I l

las AWN;

em F"001

ONG~i ucmm

NOWT/ /7 _ _

s. .10OWI ~11c a

Fg2 %Lhemratic txawing ot Siahiwiing Fin and Pr(ollcr Guard Fin%

2

4. 3*11 SUaVEY AID FIPMLICE UIPIITS USING STOCK P3Dh'ELLEE

The wake survey, resistance and self-propulsion tests were conductedin the DTNSRDC towing tank using Model 5445 representing the SWATH T-AGOS19 with scale ratio of 12.5. The nominal wake wasn measured at modelconditions corresponding to design draft, zero trim and model apeed equiva-lent to 10 knots (5.14 m/s) full-scale. The model vas fully appended withcanards, stabilizers, and ring-shaped propeller guards with supportingfins. For the wake survey, canards and stabilizers were fixed at xerodegree angle of attack.

The measured values of the circumferential variation in the axial,tangential and radial velocity components at four nondimensional radii areshown in Figure 3. From these data, similar results are obtained for otherradii by interpolation. The sign convention of the velocity componentsare shown in Figure 2.

1.2 1 1 I 1 1 1 1 1 1 I 1 1 1 1 7 1 1I1.0

v 0.3

VTv

V -0.2

VR0.4

-0.2

0.0

-0.2

0.2

0.2

seaMi 4 2 0s4 u eMa 0VOMNAW(" OIMAGEI"

-0.2 icmeeta Vraini h xalrneta n aia ~iNC np et

0.2

Table 2. Powering Predictions Based on Model Propulsion TestUsing Stock Propellers

SHIP SPEED EHP(PE) DHP(PD) 'R RPM 1-t l-wT(KNOTS)

3.0 27.1 45. .975 39.9 .83 .8104.0 60.5 100. .975 52.3 .83 .8105.0 116.6 190. .980 65.2 .83 .8156.0 202.5 330. .980 78.4 .83 .8157.0 320.0 520. .980 91.2 .83 .8158.0 480.0 785. .980 104.6 .83 .8159.0 700.0 1140. .985 118.4 .83 .81510.0 955.0 1550. .985 131.4 .83 .81511.0 1230.0 2000. .985 143.8 .83 .82512.0 1580.0 2560. .990 157.1 .83 .835

The resistance and self-propulsion experiments were performed on Model5445 using the stock propellers. The appendage suit was the same as for thewake survey except that 'he angles of the control surfaces were adjusted tothe position which required minimum delivered power at full-scale speed of 10knots. The angle of attack on the forward canards was 4 degrees trailing-edgedown, and for the aft stabilizers it was 6 degrees trailing-edge up. Thepowering predictions made by the standard DTNSRDC prediction method [21 arepresented in Table 2.

5. PROPKLLEI DESIGN

5.1 Design Condition and Procedure

The propeller was designed to absorb 800 horsepower per shaft at 185 rpmbased on the results of wake survey, and resistance and propulsion tests withstock propellers.

The propeller was designed in five phases;(a) Preliminary Design: Estimates are made for number of blades, diame-

ter, blade area ratio and blade outline at design condition such that thepropeller is compatible with the ship and the propulsion machinery from thestandpoint of efficiency and vibration.

(by Lifting-Line Design: The radial load and the corresponding hydrody-namic pitch distributions are computed using Lerbs [31 induction factormethod.

(c) Propeller Global Geometry: The initial blade chord length and thethickness distribution are refined by considering various hydrodynamic as wellas structural aspects such as cavitation, erosion and thrust breakdown, andblade strength. Lifting-line calculations are repeated for different geomet-ries as the geometry is refined.

(d) Lifting-Surface Design: The final pitch and camber distributionsare determined by lifting-surface calculations.

(e) Design Check: In this phase the unsteady forces and moments arecalculated and compared with the design requirements. If the predicted valuesdo not meet the design requirement, the skew distribution is modified. Sincethe skew distribution will affect the resulting pitch and camber distribution,the lifting-surface calculations must be repeated for a new skew distribution.The design propeller is reviewed to check the off-design performance and tosummarize the final design predictions in terms of required speed margins andother specified constraints.

These phases are specified as a guide to the major steps in the designprocedure. However, most of these steps are closely interrelated, and itera-tions between the steps are necessary.

4I

U U U U- U U U ~ ~ ~

5.2 Estimation of Full-Scale Wake 12 1 1 1 1 1

One of the essential piecesof information for propeller design > 1.0is the ship's wake in which thepropeller will be operating. Twoaspects must be considered for 0. Fship's wake; the effect of Reynolds

number for full-scale ship and theeffect of propeller action. z 06

WMODEL WAKE

The effect of different EXeRItNTReynolds numbers, or scale effects, 04

are important for ships such ascommercial ships, SWATHs and navalauxiliaries where the propeller S 0operates in the boundary layer ofthe ship's hull. There is no re-liable universal method that is 0 . ,400 02 04 00 08 18applicable to surface ship wake NONDIMENSIONALA"usI,R1scaling. Simple methods includingthe one proposed by Sasajima and Fig. 4. Comparison of Measured Model Nominal Wake andTanaka [41 are available, but none Predicted Full-Scale Nominal Wake.of them fully represents the com-plicated flow behind the ship.

In the present design, the full-scale nominal wake was approximated byadding to the measured SWATH model nominal wake the difference between thecomputed model- and full-scale wake for the equivalent axisymmetric unap-pended hull. The nominal wake of a body of revolution was calculated based onthe potential flow/boundary layer interaction theory developed by Huang, et al15). T)ie submerged lower hull of T-AGOS 19 excluding the strut was approxi-mated by an unappended body of revolution having the same length and displa-cement as the T-AGOS 19 hull excluding the strut. Figure 4 shows the compa-rison of the circumferential average of the measured model nominal wake andcomputed full-scale nominal wake,

When a propeller operates behind a ship, the inflow to the propeller ischanged due to the action of the propeller. The propeller accelerates theflow over the stern, thus resulting in a decrease in the pressure on the hulland changes the boundary layer characteristics. This inflow to the propellerwhen the propeller is operating is called effective wake.

Although one can use Huang's method to calculate the effective wake, anaverage correction method was used in the present design. This method hasproven reasonably successful in the past at DTNSRDC for obtaining the desiredship speed and shaft rpm at design power. This method uses the ratio of theeffective wake, i-WT, based on the thrust identity from the propulsion experi-ment using a stock propeller, and the volumetric mean wake, l-wv, from thewake survey, i.e.:

(I - wT)

(Vx/V)Corrected * (Vx/V)Nominal (1)(I -w v

where V./V is the nondimensional circumferential mean, axial velocity compo-nent from the wake survey.

To account for scale effects, the effective wake of the ful l-scale shipwas empirically estimated by increasing the model effective wake by 6 per-cent. The radial distribution of the full-scale effective wake was alsoestimated by using Equation (1). For final propeller design, the correctedfull-scale wake distribution was used.

5

5.3 Diameter Optimization

The propeller diameter was de-termined by considering the propul-sive efficiency at full power and lowspeeds with different propellerloadings and the sustained speed )requirement. The propulsive effi-ciency was calculated by a lifting-3line procedure for these conditions. 0For each operating condition, thepropeller diameter was varied from 8to II feet. In these calculationsthe radial load distribution wasassumed to be Lerbs optimum for allconditions.

U

As shown in Figure 5, the opti-mum diameter for the full power con-dition is about 9 feet. Although not 6. 9.5 .0 ,1o

shown in this figure, the optimum PRoUMAMM

diameter increases with increasingloading at low speed. Therefore, .5. PvoEnionys DDiffetOp er ingconsidering the low speed conditions CitiusflffrnDiuuirthe diameter was taken as maximumallowable which satisfies the sustained speed requirement. With a 5 percentspeed margin added to the minimum required sustained speed of 9.0 knots, theeffective power is 800 horsepower and the required propulsive efficiency is0.625. From Figure 5, the maximum diameter which satisfies the propulsiveefficiency of 0.625 at the sustained speed of 9.45 knots is 10 feet. In thefollowing analysis, the propeller diameter was set to 10 feet.

5.4 Circulation Distribution30

Three circulation distribu- G, (MALItions were investigated as shown G,in Figure 6. The corresponding G -/ SS OPIUMI

hydrodynamic pitch angle, tan 20 Q -

are also shown in Figure 6. The "/Lerbs optimum circulation distri-bution, designated Go, and the 10mathematically unloaded root and

tip circulations were investi-gated to serve as a guide toselect the degree of tip un- 00loading. Lerbs optimum is anestimate of the circulation dis-tribution which will produce the

07

highest efficiency for a wake- 06 G2 G06Gadapted propeller. G O

0S

Table 3 shows the speed --predicted by the lifting-lineprogram at full power for thethree circulation distributions 03for the same diameter of 10 feetand the expanded area ratio (EAR) 02

of 0.4. Load distribution G2 was I , , ,selected as final which repre- 02 04 06 08 i0sents a calculated sacrifice in % s* US,,R,ship speed at full power of 0.02knots as compared to the Lerbs Fig. 6. Circulation and tanpi Distributions.optimum propeller.

6

,

5.5 Thickness and Chord Table 3. Effect of Radial CirculationLength Distribution Distribution on Propeller Performance

The thickness and -------------------------------------------------cbordlengtb distributions TYPE OF SHIP SPEED (KNOTS) PROPULSIVEshould be carefully CIRCULATION RPM - 185 EFFICIENCYselected to provide ade- -------------------------------------------------quate blade strength and Go (OPTIMUM) 10.17 0.624sisaltaneounly minimizethe tendency towards CI 10.13 0.615cavitation erosion. Inaddition, the propeller G2 (FINAL) 10.15 0.619efficiency must not be -------------------------------------------------materially sacrificed.

In order to select a blade shape for best cavitation characteristics, usewas made of theoretically predicted cavitation inception data for two-dimen-sional sections 16). For a given blade shape, i.e., c/D and consequently t/c,the cavitation characteristics were evaluated by calculating the angle ofattack variation with cavitation number, and comparing it with the minimumpressure envelopes, which are commonly called "cavitation buckets".

The minimum thickness was determined first according to the 1985 ABSClass C, Ice Strengthening Rule (71. The initial shape of the chord lengthdistribution was taken from the existing DTNSRDC Propeller 4381 (8]. Theradial distributions of c/D and t/c have been varied until the predictionensured no leading edge and back bubble cavitations at the design condition.The final thickness and chord length distributions thus obtained are shown inFigures 7 and 8, respectively. The resulting expanded area ratio was 0.484.

5.6 Lifting-Line Design

Since we have sufficient information on geometry, ship wake and poweringcharacteristics, a more elaborate design can be made using lifting-line modelto predict the ship speed, effective power, blade lift coefficients at designcondition, and other information necessary for lifting-surface design. In thelifting-line calculations, viscous effects are included by specifying thesectional drag coefficient, CD. In the present design, full-scale section

03 I 03 1

alFINAL

• 0

I'

9 1 00 I02 04 06 08 10 02 04 06 06 110

NOO4DIMNSIONAL RADIUS 41111 NONDIMENSiONAL RAOIUS WIr

Fig. 7. Radial Distribution of Thickness. Fig. 8. Radial Distrbution of Chord Lenglh.

7

drag coefficients were used *a , . .based on the experisenta.1and empirical data in Re-ference 9. The C value at .0.7 radius was 0.1066. Thepredicted ship speed at thefullI power design condition a 0.is 10.24 knots and thethrust coefficient. K., i E - 0urTNO-URPCs ounwo ipo-W0.103. e - FAInO VALu.S OF P§30- OUTMf

--- uFTnO-UrN 1.

5.7 Lifting-Surface Design 0.4 /

The final pitch andcamber distributions cor- 0.3responding to the selectedradial load distributionand other geometry from o 0.3 0.4 0.6 0. . 0.7 0.8 0.3 1.0l ifting- 1 ine computations NONDIMENSIONM RADIUS i,/llwere determined fromlifting-surface computa - Fig. 9. Pitch Distribution (P/D) Predicted by Lifting-Surface

Code, PBD-10.tions using the computercode, PBD-10, developed byGreeley and Kerwin [10]. I I IThe section thickness shapewas selected as NACA 66 Owith NSRDC modified nose 0g- .... VALUES OP P80u10 OUTPUTand tail [61 and camber as z - L--IFTING LINENACA a-0.8 meanline. I

Figures 9 and 10 showthe final pitch and camberdistributions determinedfrom lifting-surface codetogether with the values 1 oocomputed from lifting-line •code. Table 4 presents asummary of the design pro- 02 03 04 05 06 07 0a 00 10peller geometry. Figure 11 NONDIMNSIONAL RADIUS ,011shows the blade outline ofthe design prope Iler. Fig. 10. Final Camber Distribution (fM/c) Predicted by Lifting-Surface

Code. PBD-10.

6. PKRFORMANCE PREDICTION

6.1 Prediction of Delivered Power and RPM

In order to predict the delivered power and propeller rpm, open-waterperformance is necessary. The open-water performance of the design propellerwas predicted by using lifting-surface computer code, PSF-2, developed byGreeley and Kerwin [101.

Once the design propeller open-water performance and the stock propellerpowering characteristics are known, the propeller rotational speed and thedelivered power can be estimated assuming that the propeller/hull interactioncoefficients, (l-t) and (l-wT), and the relative rotative efficiency, R I aresame for both the stock and the design propellers.

Then, the shaft power can be obtained from the effective power and thepropulsive cificiency. For a given ship speed, the advance coefficient wasobtained from the intersection of KT/J2 curve and the propeller open-watercurve; KT versus J. From this advance coefficient, the propeller rotational

speed was computed.

8

At full power, the predicted Table 4. Geometric Characteristics ofship speed was 10.45 knots and the Design Propeller ofpropel 1er rpm was 186.2. The dis- SWATH T-AGOS 19crepancy may be caused by inaccura-cy in the predicted open-waterperformance. Diameter 3.05 m (10 ft)

Number of Blades 5Rotational Direction Outward

6.2 Unsteady Bearing Forces and Expanded Area Ratio 0.4276Moments P/D at 0.7R 0.6471

Section Meanline NACA a-0.8The unsteady bearing forces Section Thickness NACA 66 (Hod.)

and moments are calculated by using Design Advance Coeff. JA - 0.484a lifting-surface code, PUF-2, de- Design Thrust Coeff. K - 0.103veloped by Kerwin and Lee 111 . Thrust Loading Coeff. CT = 0.835The predicted fundamental blade-frequency (5th harmonic) thrust andtorque values were 5.74% and 4.33%of the design thrust (KT-0.10 3 1)and torque (KQ-0.012) values, res-pectively. Since the basic designcriterion was to produce a siuplestpossible blade geometry, no attempthas been made to incorporate skewin order to reduce the calculated SWEPT OUTLINEunsteady forces and moments.

OUTLINE

7. IZPBRIMKETAL RESULTS AND

ADJUSTHEUTr OF BLADE PITCH PRJETE

7.1 Open-Water and Powering OL

Experiments

Three aluminum model propel-lers; one with a diameter of 12inches for cavitation tests, andtwo with a diameter of 9.6 inches Fig. II. Blade Outline of esign Propeller.for powering tests, were manufac-tured by Maritime Research & Con-sulting (SSPA) in Sweden. Theopen-water and powering experiments were carried out at DTNSRDC. The predic-tions of full-scale powering performance based on the model tests with thedesign propellers are summarized in Table 5.

The predicted full-scale rpm was about 5 rpm lower than the design rpm of185. The major cause of this low rpm is presumably due to the large dif-ference in Taylor wake fraction (l-wT) between stock propeller test (0.864for full-scale wake), which was used for the design, and the wake fractionobtaianed from the test with the design propeller (0.755 for full-scalewake). The effective wake, 1-wT, from stock propeller test is about 152higher than the value from design propeller test. As a result the designpropeller is overpitched and turns at a slower speed than the design rpm.Since 5 rpm was out of acceptable range, it was decided to modify the propel-Ier geometry.

7.2 Adjustment of Blade Pitch

A simple way of correcting the low rpm is to reduce the pitch by rotatingthe blades about the blade reference line so that the new rpm will match thedesign rpm. In order to determine a new pitch distribution which will produce185 rpm at full power, three rotational angles; 1.0, 0.5 and 0.6 degrees in

9

order, were tested. The Table 5. Predicted Full-Scale Speed and RPMnew pitch was obtained by at Full Power Based on Model Testsreducing pitch by the same Using Design Propellerpitch angle at all radii.For each new pitch, the------------------------------------------------open-water performance was Full Power With Still Without Stillcalculated by PSF-2. 1600 HP Air Drag Air Drag

The rpm's for the re- V (Knots) 10.49 10.56pitched propellers were RPM 180.0 179.9calculated based on the l-wT 0.755 0.755predicted open-water per- 1-t 0.820 0.820formance of the repitchedpropellers and the poweringtest results for the origi-nal propeller, with the Table 6. Predicted RPM for Original Design andassumption that the propel- Repitched Propellers at Full Powerler/hull interaction coef-ficients were same for both

the original and the re- PROPELLER RPMpitched propellers. Table6 shows the predicted rpm Original Propeller 181.3of the repitched propel-lers. For the same predic- Repitched Propeller (1.0 degree) 190.5tion conditions, the dif-ference in predicted rpm Repitched Propeller (0.5 degree) 185.2between the original pro-peller and the propeller Repitched Propeller (0.6 degree) 186.3repitched by 0.6 degrees is5.0.

Based on this paramet-ric study, the pitch of 0the original propeller wasreduced at SSPA by rotatingthe blades about the blade 07 ORIGINALPIT Hreference line by 0.6 deg-rees. Figure 12 and 13show the comparison of the a 0. 0pitch distributions and theexperimental open-waterperformance of the original / ,oa. ROTATIONand the repitched propel-lers, respectively.

A new power and rpm

prediction was made for therepitched propeller utili- 03zing the open-water testresults, and assuming thepropeller-hull interaction 02 1 1 1 1coefficients and the pro- 02 04 00 a$ to

pulsive coefficients were NONOIMENSONAL RAOIUS rRI

the same as for the origi-nal propeller. Fig. 12. Comparison ot the Original and Final Pitch Dirihution

Figure 14 shows the comparison of the predicted full-scale power and rpmfor the original and the repitched propellers. At full power, the full-scalerpm and the ship speed for the repitched propeller were predicted to be 184.8and 10.4 knots, respectively. Based on this prediction, NAVSEA decided toaccept the repitched propeller as the final propeller for SWATH T-AGOS 19. Nopropulsion tests were carried out using the repitched propeller.

10

7.3 Cavitation Tests@I---

Wake survey and psA.rm

Cavitation tests were r"

carried out in theSSPA's large cavitationtunnel (cross section

demihul I mounted in thetest section on thedesign waterline [121.The demihull model wasequipped w it h thecanard, stabilizer, andthe prope lL er protec-tion ring with its four 5support guard.

war* performed using 5- A.OwACICOSP111ci00 Whole pitot tubes with Fig I ICornpan..on of Exprimntal Open Water PertitntAn..e htwt.een Originalthe optimum angle of and Repitched PriOilcvkattack for the controlsurfaces; i.e., 4 deg-rees trailing-edge downfor the canard and 6 degreestrailing-edge up for the stabi- I T-TI Iliter. The axial velocity corn .111'cG -epLASP0-aM aponent at 0.98 radius measured - 011111100"ti, - MM"Pa - 011at DTNSRDC and SSPA are compared C) FUL 5w5m -O1 @UWin Figure 15. The wake peak due We~ 1109to the strut near 0 degrees,which controls the blade cayvita- M Ition, shows good agreement.Large discrepancies are shown at g-the blade position angle of 240 to'

degrees. The major source of Idifference in control surfaceangle of attack.

Cavitation inception and mD Iii 'sobservation tests were performedwith the optimum angle of attack 6-4of the control surfaces at theo r ig inalI p it ch o f(P ID)0 7R'0.6722 and at the . ..final pitch of (P/D) 0 7 1t0.6471.IuFor the original pitch, testsam "M.

were performed for both modeland fu I1 -ac sca e wakes. For the I ig 14 P-crnp F'eiiiniii, VI ihc Rfpol lI, I1,.'%t,,repitched propelIler, tests wereperformed only for model wake.

Figure 16 shows the polar diagram of the maximum radial cavitat ion exten-sion at full power for model wake, and photographs reproduced from the videotapes recorded during the tests. The white spots near the leading edgebetween the 0.85 and 0.9 radius indicate the leading-edge sheet Cavitation onthe suction sides. The repitched propeller showed slightly improved cavita-tion performance. The cavitation erosion test results showed no erosiontendency.

Li 0 6l

Propeller design requires complex trade-off. to satisfy various designrequirements. As demonstrated in the present design, even it the design hasbeen carefully developed, utilizing *tate-of-the-art design tools, there is noguarantee that the propeller will perform as designed.

The results of the propulsion experiments with the design propel Loeshowed that the propel lore lid not meet the required rpm at fu' I power. Anexamination of the propel ler-hullI iteration coefficients showed that theTaylor wake fraction, "~T, had changed substantially between the stock &addesign propelleI rs. If it had not changed, the rpm should have been close tothe design rpm.

As indicated by the very low rpm from the powering tests (see Table 2)using stock propel lers, the pitch Of the stock propel ler was substantiallyhigher than the design propel br, as was the wake fraction. [dealI Ly, thegeometry of the stock vropeller should be as ,list as possible to the finalpropel I tr geometry.

In general, the m"*e wake has been used in surface-ship propel ler de-sign, assuming that scale effects are negligible. In the present design, theestimated fullI-sCale wake vas used, which was ibtained by increasing themoel value by 6 percent. The performance of the full-sc:ale propeller at seais very such dependent upon how wel I the ful 1-sale wake was predicted.

The authors are grateful to the Propul sor Design Review Committee of C0oe1544, DTNS&DL, F. Peterson, E. Caster and S. J~ssup, for their discussions andcriti&que of the design, and tj) Dt. . Rutget s)n it SSPA for prov iding themwith the photos shown in Figure ib.

-7 - - - T- 7- T~. T-u T.,o. T.

is4 ~T

110 .. on*,** *w 0* 0 1

5

IU

*

U -I0

U-

U

L4)

C

S I'

LW

WEIp C S -~

-'WI,oE

* hU C

U-

3

i 1/

13

A. ~

i. Covich, P.M.. "IWATI T-AGOi. A Producible Doeig," AIAA Paper No. 8-2314, Presented to the AIAA itb Advanced Marine Syoteme Conference, ian Diego.Califeraia (Sep. 1966).

2. Grant, J. and C. Wilson, "Dosign Practices for Powering Predictions,"DIUSUIC Report SPD-693-01 (1976).

3. Lerbe, LW., "Moderately Loaded Propellers with a Finite Number of Bladesand and Arbitrary Distribution of Circulation," Transactions iNANE, Vol. 60(1952).

4. $asojisa, a. and I. Tanaka. "On the Estimation of Wake of Ships," Procee-ding., 11th ITTC, Tokyo (1966).

5. Nuang, T.T., Wang, N.T., Santelli N. and Groves, N., "Propel ler/Stern/-Boumdary-Layer interaction on Axisynmetric bodies: Theory and Eaperiment,"DOTU1DC Report 76-0113 (1976).

6. Brockett, T.K., "Minimum Pressure Envelopes for Modified MACA-66 Sectionswith NACA a*0.S Camber and lushipe Type I and Type 11 Sections," DTNSRDCReport 1789 (Feb. 1966).

7. "Rules for Classification and Construction of Steel Vessels," Americanbureau of Shipping (19S5).

S. Boesell, UJ., "Design, Cavitation Performance and Open-Water Performanceof a Series of Research Skewed Propellers," DTOSRDC Report 3339 (Mar. 1971).

9. Abbot, I.N. and A.E. Von Doenhoff, Theory of Wing Sections Including aSumary of Airfoil Data, Dover Publication Inc., New York (195I.

10. Greeley, D.S. and J.9. Kerwi., "Numerical Methods for Propeller Designand Analysis in Steady Flow," Transactions SNAKE, Vol. 90 (1962).

11. Kerwin J.E. and C.S. Lee, "Prediction of Steady and Unsteady MarinePropeller Performance by Numerical Lifting Surface Theory," TransactionsSNANE, Vol. 86 (1978).

12. Rutgersson, 0., "SWATH T-AGOS Model Tests in CavitStion Tunnel ," SSPAReport No. 4397-1 (Jul. 198b).

15

= 1. . , , , . i .,,.t : r;: .%&..pull: ::,; -

INITIAL DISTRIBUTION

Copies Cop is

14 NAYS9A 1 NSWC, White Qak/Lib1 SEA 5046 (Bauman)3 SEA 55W3 (Costock, I NUSC/Lib

Davey,Hlough) 12 OTIC

1 SEA 56X1 (LaFons)3 SEA 56X7 (Crockett, 2 "IAMA

Platzer, 1 Div o' Ship R&DMajuadar) I Lib

I SEA 5011 (H~owell)I SEA 5015 (Cannon) 2 MITI SEA 50151 (Kennell) I Baker LibI SEA 55N3 (Lee) I J.E. KerwinI SEA SSN2 (Paladino)I SEA 99612 (Library) 3 U. of Michigan (MAKE Dept)

I Lib2 PINS 383 (Nunnery, Perin) 1 T. Brockett

I W. Vorus1 14SC 9492 (Jacobson)

I Penn State IJ/ARL LibI SSNO 100 (Jenkins)

I SNAME/LibI USCG G-MSC (Reid)

I Tracor, Hydronautics1 ASS AAJI/fd (Stromer) (S. Kobayashi)

I NAVSHIPYD PrSMH/Lib I Bird Johnson (J. Norton)

1 NAVSHIPYD PHILAJLib 1 IA Assoc. (B. Cox)

I NAVSHIPYD NORVA/Lib 1 Atlantic Research fD. Greeley)

I NAVSHIPYD CIIASN/Lib

I HAVSHIPY) LBEACH/Lib CNE 1iRBFO

I AVSH1PYD KAR.E/Lib Copies code Name

I 011.3 D.D. MoranI NAVSHIFYL Po(;E r 1!L b

121S R. LambI NAVSHIPYD PEAJL/Lib

I SNA/Lib W .B. Morgan

I NAVPGSCL/Lib l,4 VJ oaei

I NOSC/Lib 1506 S. Hawkins

I CS/121 15018 R. Boswell

1 NCSC/712

Corie" Code man*

1 152 W.C. Lin

3 1521 W. DayG. lordaH. Liu

3 1522 N. WilsonJ. LawlerK. Remmers

1 154 R. Folb

2 1542 T. HuangB. Yin

25 1544 F. PetersonK.-M. Kin (10)A. Reed (10)S. JessupC. SchottJ. PressN. Hubbard

1 1905.1 W. Blake

10 5211.1 Reports Control

1 522.1 TIC (C)

1 522.2 TIC (A)

I ., vo -S 5. . .7Izi% . - - *.- * *:** *. *U p4