• Author / Uploaded
• Daniel Riveros

Citation preview

SNAME Transactions, Vol. 91, 1983, pp. 195-224

Effects of Propeller Design-Point Definition on the Performance of a Propeller/Diesel Engine System with Regard to In-Service Roughness and Weather Conditions Miro Kresic, 1 Member, and Bruce Haskell, 2 Associate Member The objective of this paper is to investigate the performance of a fixed-pitch propeller/diesel engine system on trials and in service as a function of various propeller design-point definitions and time. The influence of hull roughness, propeller smoothness, and environmental factors is taken into consideration in analyzing propeller absorption. The increase of resistance in service over a period of years as a function of hull roughness and fouling is analyzed and calculated. Estimations for the loss of open-water efficiency of the propeller in service, as a result of blade surface deterioration and fouling, are reviewed and used in the calculations. Also, changes in propulsion components in service as a function of roughness and environmental factors are discussed and effective wake changes are taken into consideration in the calculations. Finally, several propeller design-point definitions are compared with regard to in-service performance of the propeller/diesel engine system.

Introduction ONE OF THE many factors responsible for the good performance of a ship is the correct selection of the design point of the propeller. Depending upon the type of vessel, service route and hull maintenance procedures, there is only one fixedrpitch propeller which will provide the proper performance for a particular service. Failure to find a correct propeller design point will usually result in overloading of the diesel engine under various operational conditions. The consequences of this overloading will be complaints, such as high maintenance cost from the ship's owner, excessive engine wear and tear from the main engine manufacturer, and low engine revolutions and ship's speed loss from the crew during the service period of the vessel. Why do some diesel engine installations become overloaded during years of service? As is well known, the effects of service, wind and waves all have an impact upon a ship's performance. The ship's hull becomes rougher as a result of corrosion, painting, fouling, etc., which lead to increases in the resistance. The propeller also suffers from these effects, which decreases its efficiency. These changes all lead to increased loads for the diesel engine. If the engine does not have enough power to overcome the increased load, the engine becomes overloaded. Through the years, many definitions for the selection of the propeller design point have arisen. The intent of these definitions is to provide the proper margins to account for the increased loads of service without overloading the engine and still permit the full capability of the engine to be used. But, are 1 Director, Basic Design and Naval Architecture, John J. McMullen Associates, Inc., New York, N. Y. 2 Naval architect, John J. McMullen Associates, Inc., New York, N.Y. Presented at the Annual Meeting, New York, N.Y., November 9-12, 1983, of THE SOCIETYOF NAVALARCHITECTSAND MARINEENGINEERS.

these design points doing the job? Often ships must undergo, at some point in their life, some correction to the propeller, such as diameter reduction, repitching, or retrofit of a completely new propeller. Therefore, it is obvious that these design points do not always provide the necessary margins. In recent years, however, there has been much research done to quantify the effects of roughness and fouling on resistance and propeller performance, so that a more accurate prediction of the loads on an engine may be made. Unfortunately, most of this work has focused only upon one aspect of the roughness effect, without investigating the effects on the ship as a whole. It is the intent of this paper, using those theories already available in the literature, to see the cumulative effects of service on the diesel' engine/propeller system and how the design point affects performance later in service. This is done through the analysis of two ships and comparing each ship's performance at four different events (on trials, six months before its fourth dry-docking, just before its fourth dry-docking, and six months before its seventh dry-docking), using seven different propeller design-point definitions. The resistance, propulsive factors and propeller characteristics are calculated for each of these events. The performance and speed-keeping qualities of seven common design-point definitions are then analyzed at each event, providing an insight into the qualities of each definition. Assumptions and examples are given to illustrate the analytical procedures developed in the paper. The assumptions are believed to represent average operating conditions. For hull and propeller maintenance, other than shown in the paper, the roughening assumptions should be reconsidered for the particular case being investigated. E s t i m a t i o n o f a n i n c r e a s e o f r e s i s t a n c e in s e r v i c e An estimate of the increase in hull resistance during a ship's service is necessary so that the propeller absorption can be analyzed for various stages in service. Of the components making

195

up a ship's resistance, the frictional resistance for the majority of commercial ships represents the largest part of the ship's total resistance. The frictional resistance for a slow-speed ship can be about 80 percent of the total resistance and as much as 50 percent for high-speed ships [1], a Therefore, predicting frictional resistance for various service events is of great importance. There is a long history of research devoted to the prediction of the largest single resistance components based on experimental and theoretical work. Recently, this work has branched out to include the effects of hull roughness on frictional resistance and the change in hull roughness in service. There are two basic components which, in service, cause an increase of ship's resistance after trials have been completed: hull roughness and environmental factors. The hull roughness contributes to increased resistance by increasing friction between the hull and sea, while environmental factors contribute in the form of wind and waves. Hull roughness The hull roughness is constantly changing during a ship's service period. The causes of hull roughness can be divided into several groups [2]: plate roughness, paint/coating type, corrosion and fouling. The plate roughness of a newly completed ship is dependent upon the production quality of the steel plates, as well as on the sophistication of the methods used by the shipyard in building the ship. The roughness from paints and coatings is influenced by the type of material used and the methods of application, and can vary significantly. Corrosion on steel plates, which results in permanent damage of the surface, varies as a function of many environmental factors, as well as its location on the hull. Fouling roughness is influenced by the type of antifouling paint, method of application, location on the hull, environmental factors and, most of all, time in port. For the analysis herein, it has been assumed that the steel plating used to construct the ship was properly pretreated with shop primer and that good workmanship was emFloyed in construction. The hull was painted with two coats of anticorrosive paint and before trials with two coats of conventional antifouling paint. The estimate for the total surface roughness of a new ship when beginning service, expressed in terms of a 3 Numbers in brackets designate References at end of paper.

mean apparent amplitude (MAA) measured over a 50-mm sampling length, is about 150 microns. This represents an average new hull roughness at the beginning of service, which was adopted by ITTC-1978 [3] as the standard roughness for a new ship for use in full-scale predictions. The change in MAA in service as a result of plating deterioration from corrosion, mechanical damages caused by berthing, cable chafing, grounding, operation in ice, etc. was assumed to increase an average of 2.8/am per month [4]. Also affecting the hull roughness is the hull treatment a ship will receive in dry-clocking and the interval of dry-docking or the ship's "hull maintenance program." Depending on the procedures used in dry dock and the workmanship employed, the roughness of the hull may either become less or, as is usually the case, may actually increase. Townsin et al [4] found that 68 percent of the ships they measured increased in roughness in dry-docking and that ships with relatively smooth hulls had the largest increases during docking, while relatively rough hulls showed small declines in roughness. For this study the ships are assumed to be dry-docked at intervals of two years, as required by the classification societies for recertification. Possible extensions of six months, as well as in s i t u cleanings, were not taken into consideration for this analysis. When in dry dock the ship is assumed to get the treatment as shown in Table 1. The increase in the average hull roughness as a result of these dry-dock cleaning procedures is anticipated to be 14 g m per docking. The amount of change in hull roughness per docking is based on published data [4] and interpretation of the simulation of in-service conditions. The final component of hull roughness to be considered is fouling. The average roughness from the accumulation of marine fouling on a ship's hull is estimated and expressed in an equivalent MAA-value as proposed by Malone et al [2]: MAAfouling (sides) -~ (HRF)(PT)(CEFF) MAAfouling (bottom) = 0.75(HRF)(PT)(CEFF)

where H R F = hull roughness factor, # m per port days (see Table 2) PT = port time, days 0.75 = factor applied to bottom fouling rate because fouling does not grow as fast on bottom as it does on sides

Nomenclature AL = lateral projected area AT = transverse projected area

B = beam BMEP -- brake mean effective pressure c = chord length of propeller blade C = distance from bow of centroid of lateral projected area CO = blade drag coefficient CEFF = antifouling coating effectiveness factor CF = frictional coefficient ACI¢ = roughness allowance coefficient CL = blade lift coefficient Cs = service roughness coefficient D = diameter hMAA= equivalent mean apparent amplitude based on 50-mm apparent wave length for propeller HRF = hull roughness fouling factor J = propeller advanced coefficient k = form factor

196

KT = propeller thrust coefficient KO = propeller torque coefficient zXKro = change in thrust coefficient due to a change in drag coefficient 2xKTt, = change in thrust coefficient due to a change in lift coefficient AKoo = change in torque coefficient due to a change in drag coefficient 2XKQL= change in torque coefficient due to a change in lift coefficient Lse = length between perpendiculars Lo^ = length overall LWL = length on waterline M = number of distinct groups of masts or kingposts seen in lateral projection MAA = mean apparent amplitude based on 50-mm apparent wave length PE = effective horsepower P/D = pitch-diameter ratio PT = port time

Effects of Propeller Design-Point Definition

SS = length of perimeter of lateral projec-

tion of vessel excluding waterline and slender bodies such as masts and ventilators T = draft t = thrust deduction and propeller blade thickness t/c = thickness-chord ratio V = speed of ship Vw = wind speed " Wh = wave height wm= mean nominal wake fraction WS = wetted surface wr = Taylor effective wake fraction Z = ratio of the accumulated time, since application of antifouling paint was made, to effective life of antifouling paint; number of propeller blades r/R = relative rotative efficiency r/0 = open-water efficiency p = density of fluid

CEFF = antifouling coating effectiveness factor = 1.0 - [2.72/e z - 0.240(Z - 1.0) °.263] Z --=ratio of accumulated time, since application of antifouling paint was made, to effective life of antifouling paint In the calculations of MAA it was assumed that the hull roughness factor corresponds to moderate to severe fouling, HRF = 0.5755 (Table 2). The number of days in port (PT) is an important variable, since the majority of marine growth occurs when the speed of the vessel is less than about 3 knots. This variable is a function of trade route and type of operation and should be selected for each vessel for which analyses are conducted. For the sake of calculations made in the present paper, the time in port (PT) has been assumed to be 80 days per year. The antifouling coating effectiveness factor (CEFF) will be equal to zero during the effective life of the coating, usually one year for conventional painting systems. After the effective life of the antifouling paint has been exceeded, the CEFF factor increases exponentially toward a maximum of 1.0, which implies that the coating no longer inhibits fouling growth. Having expressed each component of hull roughness by an equivalent MAA, the total hull roughness may be calculated for any particular time in the ship's life by accumulating the effect of each component. Based on the .preceding assumptions, the mean apparent amplitude of hull roughness was calculated for all four events that will later be used in the propeller analysis. The first event is taken to be the trial test; the second is 6 months before the fourth dry-docking or 7 years and 6 months after trials; the third event is one day before dry-docking or 8 years after trials; and the fourth event is 6 months before the seventh dry-docking or 13 years and 6 months after trials. These events have been arbitrarily selected just for the purpose of the propeller analyses; however, they provide practical check points for the evaluation of the propeller absorption as it relates to the selection of a design-point definition for a particular ship. The hull roughness for each event is given in Table 3 and in a time-history diagram of dry-docking intervals as shown in Fig. 1. To estimate the effects of hull roughness on the ship's resistance, the following equation may be used:

where ACF = roughness allowance coefficient MAA = total hull roughness, m (meters) LwL = length on waterline, m This formula is the ITTC-1978 recommendation [4] to account for hull surface roughness between a model and ship for full-scale trial predictions. To predict the increased resistance after trials a service roughness resistance coefficient (Cs) may be estimated as

Cs = ACF

service --

ACF

Procedures 1 2 3 4

5 6 7 8 9

ACF service = roughness allowance coefficient in service ACE trial = roughness allowance coefficient on trials (MAA = 150 ~zm)

It is now possible to calculate the increase in effective horsepower due to roughness by

~(ws)(c~)(0.5144 v,) 3 =

150

Treatment/Location

Pressure water underwater hull and propeller shall be wash washed with high-pressure seawater Spot after survey of the hull, spot sandblasting sandblasting of damaged areas only shall be performed Hand scraping hand scraping of marine growth not removed by high-pressure water wash shall be performed Sand wash sand wash or sand sweep to remove final film of marine growth and to rough surface to allow for better adhesion of . new paint shall be made , Freshwater freshwater wash to remove all salt water wash and sand residue shall be performed Local primer primer coat on spot sand-blasted area shall coat be made Anticorrosive one full coat of anticorrosive paint shall be paint applied by airless spray Antifouling two coats of conventional antifouling paint paint shall be applied Propeller wash propeller to be cleaned by high-pressure wash, hand scraped, hand polished by wire brush machine. One coat of grease, for protection of ship's hull painting procedure, shall be applied

p = density of water, kg sec2/m 4 WS = wetted surface of hull, m 2 Vs = ship speed, knots Environmental factors Environmental factors such as wind, waves and currents should be taken into consideration for a propeller absorption analysis. The literature contains many methods for the calculation or estimation of an increase in resistance of the ship as a result of these environmental factors. Herein, the wind resistance is estimated by a method developed by Isherwood [5]. This method predicts forces and moments acting on a ship as a result of winds acting at particular angles. For the purpose of these studies an average wind speed of 16 knots (Beaufort 4) was assumed. Also, it was assumed that the vector of wind speed could act from 0 to 360 deg and then an average increase in effective horsepower was calculated, taking into consideration the speed of the ship. In other words, it has been assumed that the ship will not always sail into the wind, but that it will encounter the winds from all angles. Because there is an equal chance that the wind can come from any direction for each vessel's speed, the wind resistance was calculated for all angles of wind in 10-deg increments and the average of these resistances was used to account for the increases in PE due to wind, For resistance due to wave action, the following equation has Table 2

PE r o u g h n e s s

Dry-dock treatment

trial

where

where

Table 1

Qualitative Fouling Severity Scale 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

Effects of Propeller Design-Point Definition

Hull roughness fouling factor [2]

Fouling Severity clean trace trace to light light light to moderate moderate moderate to severe severe

HRF, #m/Day 0.0 5.334 X 10-4 7.849 X 1 0 - 3 3.828 x 10-2 0.1178 0.2822 0.5755 1.052

197

Table 3'

Hull

roughness

thrust deduction fraction (t), and relative rotative efficiency

(~R). MAA,/~m

Event 1 Event 2 Event 3 Event 4

Initial roughness Hull roughness from service Hull roughness from dockings Fouling equivalent roughness Total roughness

150 0 0 0

150 252 42 72

150 268 42 142

150 454 84 72

150

516

602

760

been applied [2] with a wave height of 1.4 m, which corresponds to the Beaufort 4 wind speed: P~ waves = 6.25(B)(A)(2.0 X Wh + O.152)2/(T)(LBp 2) where B = ship beam, m A = ship displacement, long tons Wh = wave height, m T = ship draft, m LBe = length between perpendiculars, m No provision was made for the influence of current, but, for specific cases, this should not be ignored. E f f e c t of hull r o u g h n e s s on p r o p u l s i o n factors The next step in modeling in-service conditions is to estimate changes in the propulsion factors: effective wake fraction (WT),

Changes in the effective wake fraction are a function of the changes in flow conditions on the propeller blades behind the ship. The deterioration of the hull surface and fouling are constantly altering the flow conditions around the propeller. Environmental factors such as wind and waves, as well as steering, are also changing the resistance characteristics of the vessel, which ultimately results in reducing ship speed. All these factors have a significant influence on the wake and wake distribution pattern from trials to in-service conditions. Unfortunately, to the best of our knowledge, no full-scale tests have been performed to study the difference in effective wake fraction between two dry-docking intervals. But full-scale tests performed with a 20-m-long vessel in Japan [6] give a very clear indication in the behavior of nominal wake as a function of time between two dry-dockings. The results of these tests show that the mean nominal wake fraction (Win) grew almost linearly from the first day after repainting the hull, a value of 0.8, up to the value of 0.56 after 450 days of fouling. The almost linear increment of the nominal wake fraction as a function of number of days of fouling is an indication that a similar trend can be expected with the effective wake fraction. In order to estimate the changes in effective wake fraction between two dry-docking periods we have applied a modified version of the ITTC-1978 formula for full-scale wake prediction [3]. The original formula is W T S = (t + 0 . 0 4 ) -I- ( W T M -- t --

0.04)(1 + k)CFs + ACF (l -'1- k ) C F M

A ii 800

/I

750

I

700

4

650

II . I

/I 600 I

,50q

II

I

II

450 ..-.400 tO

.g~.550 ,,~ 3 0 0

/ ,

/i

/J

~" 2 5 0 200 150 I00 50

2

3

4

5

6

?

8

9

I0

•0 = DRY-DOE KING/YEARS

Fig. 1

198

Hull roughness, time history

Effects of Propeller Design-Point Definition

II

12

13

14

E f f e c t of propeller r o u g h n e s s on o p e n - w a t e r efficiency WTS = full scale wake fraction The loss of propeller efficiency in service can occur very WTM = model wake fraction rapidly and is a function of sailing time, time in port, propeller thrust deduction for model and ship, assumed to t = loading and environmental conditions such as seawater tembe independent of scale perature and salinity. Maintaining the original efficiency of 0.04 = coefficient to take account of rudder _effect the new propeller by means of a regular servicing and maink = form factor tenance program is possible in the majority cases, if a consistent CFS = frictional coefficient of ship, ITTC-1957 line effort for propeller maintenance is made. There are three CFM = frictional coefficient of model, ITTC-1957 line main factors affecting the efficiency of an existing propeller ACF = roughness allowance coefficient in operation: blade roughness, blade fouling, and blade Since we are no longer dealing with the scale effect, but with damage. change due to roughness in service, the full-scale wake fraction The effects of propeller roughness have been investigated (Wrs) is now taken to represent the wake fraction in service and periodically on an experimental or a theoretical basis [8-10], the model wake fraction (WTM) the wake fraction on trial. It but very few data for full-scale correlation are available. is also assumed that the ship's frictional coefficient does not However, the subject of propeller blade roughness and its effect change in service and, therefore, is always equal to the frictional on location in reference to diameter and method of measurecoefficient on trials. The roughness allowance (ACF) now ments has been widely addressed [6,8-15].' represents the difference in roughness between trial and inThe effects of propeller fouling are even more difficult to service conditions. The coefficient of 0.04, which in the quantify since there is a lack of theoretical and experimental original formula took into consideration the rudder effect on work on the subject. Based on experiments by Kan et al [6], it trials when compared with the model, can now be eliminated can be seen that these effects in terms of power requirement since it is assumed that the rudder effect for both trials and are considerable and much greater than those of surface service is identical. The form factor (k) is also assumed to be roughness. Since propellers have no antifouling paint, they the same for trial and service conditions. suffer readily from fouling. Even though marine growths are removed in larger amounts from the tips of the blades, because After these modifications, the formula will read of rotation, the influence of the smaller amounts left cannot b e ignored. This is because the portion near the tips has more wr .... ice = t + (wr trial -- t) 1 + (1 + k)CF trials influence on the torque, by virtue of its longer lever arm, than that near the hub. where The effects of propeller damage will greatly differ depending WT service = wake fraction in service on location and type of damage. Trailing-edge damage will wr trial = wake fraction on trials affect the power absorption characteristics if there is any sigt = thrust deduction nificant bending which will change the local pitch. LeadC S = ACF service - - ACF triad ing-edge damage will usually affect the cavitation performance while erosion and corrosion will affect the propeller performance since the increased roughness of the blades created by A C F t r i a l = [ l O 5 ( 1 5 0 X-L~'lw-OL- 6 1 1"]/ : 3 - O ' 6 4 ] their actions is generally concentrated at the propeller tips. The magnitude of the loss of open-water efficiency as a result of the foregoing three factors is of great importance for the rv'ce-complete performance assessment of the propeller/diesel engine system. It is our intention to establish a procedure of taking into account propeller roughness and fouling by calcuC F trials = frictional coefficient on trials from ITTC-1957 lating the loss of open-water propeller efficiency for the selected line propeller and under selected service conditions. The loss of k = form factor efficiency as a result of damage will not be taken in considerMAAs = mean apparent amplitude in service, m ation for propeller design since it is assumed that any damage For convenience, the form factor (k) is calculated for our to the propeller blades will be rectified at the first possible opstudies from the work of Holtrop and Mennen, rather than from portunity. the Prohaske method, as recommended by the ITTC-1978 Performance Committee, and we refer to reference [7] for Loss in propeller efficiency details of the calculation of the form factor. Surface roughness influences propeller efficiency in two The thrust deduction fraction for trial and in-service con- ways. When compared with the smooth propeller, the drag ditions is assumed to be equal. There is a relation between of the blades of the rough propeller increases torque coefficient wake fraction and thrust deduction fraction, but tests conducted while decreasing the circulation, which affects the lift coeffion model and full-scale ships [6] have shown that the thrust cient and thus the thrust coefficient for the same angle-of-flow deduction fraction remains nearly constant for every condition incidence. The change in torque and thrust coefficients of the of roughness, in spite of the increase of the wake fraction due propeller from a trial to an in-service condition, as a function of average propeller blade roughness, may be estimated by to roughness. The relative rotative efficiency (~TR),as estimated for trial KT2 = KT1 -- AKTD -- AKTL conditions, is also assumed to remain unchanged in service. K02 = K01 - A K Q D - - AKoL Therefore, the ratio of the propeller efficiency in the behind and open-water conditions is considered to be equal for the w h e r e smooth and rough propeller. The validity of such an assumption could be checked by measuring torque in the behind KT2, K02, KTb K01 = thrust and torque constants in service and and open-water conditions for both smooth and rough propelon trials, respectively lers, which has not been done as far as we are aware.

where

Effects of Propeller Design-Point Definition

199

.04

"---

1300

d3

.8

Iz

30

./,

7 O

C~

0

o

. . . . .

C A LCU LATED

-

I~ODEL

-

/•//f

TEST

.5

\

/

\ tl. tt~

X 1 .3

,¢ I.I ..J ..J W O.

0 og o.

I.z

z 0

t~ "r 0

o

I

2

5 4 5 6 ADVANCE CONSTANT J

Fig. 2

Comparison of calculated and test [6] K T, K O and r/0

AKTo,AKgD = change in thrust and torque constants, as a result of increased drag AKTL,AKgt+ = change in thrust and torque constants, as a result of reduced circulation; that is, reduced value of lift coefficient ,SKTo, '~

tz

.o ~

f.-/,

60

80

85

90

95

100 103

108

%RPM

Effects of Propeller Design-Point Definition

215

RO/RO

VESSEL

PROPELLER

_V

AVG. SPEED

J • ROUGHENED PROPELLER $1dOOTH PROPELLER

. . . .

(K'r 19

L- . . . .

-I I-- . . . .

-L.----

-

--b_

....

18

17

i

,i

v

4

Fig. 22

6

~

8

iv

I0

12

14

Average service speed between dry~ocking--RO/RO

of 12.4 knots would be attained at 100 percent rpm at 95 per- experienced this when they compared sister ships where the cent MCR Speed could be increased to a maximum of 12.6 propeller of one ship was kept clean and the second was allowed knots by permitting the propeller to run at 102 percent rpm, to foul [19]. Considerable speed difference was experienced 100 percent MCR and 98 percent BMEP. In the case of Event between the two ships, which prompted them to begin having 4, the maximum speed would be 12.2 knots at 108 percent rpm propellers cleaned by divers before each voyage. This again and 100 percent MCR. The difference between contractual illustrates the importance of keeping the propeller as clean as and actual service speed in Event 2 of 0.8 knots would be ex- possible. perienced and in Event 4 the drop in service speed would be Conclusions 1.2 knots. All other propellers would experience similar speed loss. Therefore, the contract service speed overpredicts the The performance comparison analysis shown for the two actual speed in service. example ships demonstrates the facts which must be considered Figures 20 and 21 demonstrate the performance of the for the proper layout of propeller/diesel engine systems in order RO/RO and OBO vessels, outfitted with Propeller V, in all four to avoid abnormal maintenance and operating problems. events. In case of the RO/RO vessel out-fitted with Propeller When creating a design-point definition which is supposed to V, Fig. 18, it is evident that the contractual service speed, de- be included in the contract specifications, the amount of margin ' fined as a speed on trials at 85 percent MCR, cannot be main- and the form in which this margin shall be defined should be tained for long. Following the constant speed line of 19.0 knots based on a propeller absorption analysis prepared for the spethe propeller revolution must be increased up to about 102.5 cific service conditions the vessel will encounter. percent RPM when 100 percent MCR is reached. After that By applying the methods described in the paper, the proa drop in propeller revolutions and ship speed will follow until peller absorption curves for various propellers should be suService Event 2 propeller absorption curve is reached. At that perimposed on the proposed diesel engine load diagram to point the speed of the vessel will be 18.2 knots and the diesel evaluate which propeller will insure that there will be no engine will operate at 100 percent rpm and 100 percent MCR. overloading before the design service event. Taking into If the speed on trials is to represent the actual service speed in consideration the limitations of a proposed diesel engine, fuel Event 2, the engine would need to develop only 70 percent consumption rate at various power levels and revolutions, ship MCR. speed, and additional power margin required for operational The loss of an average service speed between two dry- reasons, such as maintaining speed in bad weather, the pitch dockings, as a functions of years in service, was calculated for can be selected. With the propeller chosen, the trial propeller the RO/RO vessel with all seven propellers. It is of interest to absorption curve can be calculated and, by again superimposing note that for this case all propellers will perform so that an av- it on the selected diesel engine load diagram, performance data erage ship speed will occur approximately 14 months after each required for the propeller design-point definitions may be sedocking. The actual service speed between dry-dockings for lected. the RO/RO vessel with Propeller V is shown in Fig. 22. The Often in contract specifications the service speed is specified loss in actual average service speed, from docking to docking, as: "Service speed or sustained sea speed shall be attained on is indicated with a solid line which assumes hull and propeller trials at 'H' percent MCR." Further, the propeller design-point maintenance procedures as outlined in Table 1. The dotted definition is very often not included at all. How valuable then line represents an average speed between dockings with the is such a statement, and what does this mean to a shipowner or assumption that the propeller blade roughness is maintained shipbuilder? Referring to Fig. 8, on trials at 85 percent MCR at 80 # m - - a n ideal situation. The New Zealand Shipping Co. all seven propellers will provide a "service speed" of 19.0 knots. 216

Effects of Propeller Design-Point Definition

The decision as to which propeller finally will be installed on the ship, from the contractual point of view, is left to the shipyard. Certainly, the diesel engine manufacturers will suggest the selection of Propeller V, VI or VII, if possible, while some shipyards will select a propeller which will provide the highest speed on trials, as has been seen in practice. By comparing Figs. 8 and 9, it is evident that the actual service speed, on the day in the life of the vessel represented by Event 2, with Propellers I, V and VII, will be 17.4 knots, 18.1 knots and 18.1 knots, respectively--lower than predicted. It is also clear from Fig. 22 that between two dry-dockings the average service speed of the ship, outfitted with Propeller V, will never be 19.0 knots as specified on trials at 85 percent MCR. The actual service speed is constantly changing between dry-dockings, as was previously discussed, and, even though the ship's speed for Event 2 is 18.1 knots, the average speed between the third and fourth dry-dockings is 18.3 knots. Therefore, it is evident that the wording of the contract specification is not adequate in specifying the propeller. If this is not satisfactory, what would be? First, the idea that the speed on trials predicts the service speed should be abandoned and, secondly, the design-point definition should be spelled out in the specification. The design-point definition should consist of the power to be absorbed by the propeller at a specific rpm and the ship resistance, for most cases on'trials. After having selected a propeller and calculating its absorption on trials, the design point data are readily available. Along with the design-point definition, the speed on trials at the design point must be specified. The demonstration of speed on trials is important, so that the predictions of ship performance in service may be verified. If a ship cannot make the predicted trial speed, then the prediction of service performance would be in question. The owner should remember, however, that the speed of the ship on trials is not a prediction of service speed, but only a criterion for acceptance. For the owner's economic analysis of the ship's future operation, the speed the ship will make in service may be predicted only by the method advocated here or by some similar method. It is also appropriate to note at this point that, with such large variations in torque and power between trials and various service conditions, the use of a controllable-pitch propeller (CPP) would be a perfect solution to problems such as overloading of the diesel engine and would permit complete utilization of the engine throughout service by maintaining nominal or any selected power level and revolution. The advantage of a CPP installation lies in the automatic load control that can be incorporated in the propeller/diesel engine system which will insure the maximum power usage at all times, along with good maneuverability, especially when direct-coupled engines are in question. The applications of CP propellers are not widespread on larger installations and commercial vessels operating on worldwide trading routes. This is because of the large capital investment necessary to purchase a CP propeller and, according to some shipowners, their doubtful reliability. Utilizing the method presented in this paper and taking into account different hull and propeller maintenance procedures and even longer periods between dry-dockings, an evaluation could be conducted in order to investigate the possible economic advantages or disadvantages of a CP propeller. This would be especially interesting for ships where hull maintenance is neglected or in situ hull cleaning must be performed. The evaluation of propeller absorption and average service speed could provide reliable data for the justification of capital investment in a CP propeller versus a fixed-pitch propeller. To conclude, the definition for the sustained sea speed or service speed on trials does not provide insurance that the ship

will actually perform in service with such a speed. Also, since the time when the "service speed" will be attained is not specified, it has limited use for economic studies. A more accurate average service speed between two dry-dockings can be estimated for a selected propeller as a function of time in service and maintenance program procedures. Such results could then be used for operational studies or analysis of required hull and/or propeller maintenance procedures. Specifying "service speed" on trials at a particular power level without also indicating propeller revolutions will not adequately define propeller characteristics. This will contribute to the confusion as to how to design the propeller and mislead the shipowner in his operational predictions. It should also be mentioned that selecting a service design event is a complex task and could be hard for a shipowner to deal with if the trade route is not known and a hull and propeller maintenance program is not routinely established. Monitoring the performance of other ships in the fleet by measuring power absorption and revolutions in service on a regular basis, and measuring MAA and hMaa at in-docking and out-docking, plays an important role in providing valuable technical data, so necessary for correct propeller design. Equipped with reliable data, propeller design-point definitions could be estimated for a similar ship on a similar trade route with great reliability. Another advantage of having propeller absorption curves calculated for various service events would be in the selection of the diesel engine. By superimposing propeller absorption curves of the design and other service events on a proposed diesel engine load diagram, the engine will demonstrate whether it has enough flexibility in its operating range to accommodate operational requirements without detrimental effects on engine maintenance. Also, this will provide an insight into specific fuel consumption for various operating points in service, and it will give a good indication when a hull and propeller maintenance program must be altered in order to protect the diesel .engine from an overloading condition. All this would permit the selection of a propeller/diesel engine to provide trouble-free operation in service.

Acknowledgments The authors would like to thank John J. McMullen Associates, Inc. (JJMA) for their sponsorship of this paper and for making available computer time for the extensive calculations necessary. They would also like to thank Mr. John Marra of International Paint Company, Inc., who provided references and other technical information. Special thanks are also due to Mr. Marko Vukasovic for his help in the preparation of the paper and to all the secretaries at JJMA for their typing of the manuscript. Metric Conversion Table

I mm = 0.04 in. 1 m = 3.28 ft 1 m2 = 10.76 ft2 1 long ton (LT) = 1.016 047 metric tons

References 1 Todd,F. H., "Resistanceand Propulsion"in Principles of Naval Architecture, J. P. Comstock, Ed., SNAME, 1967. 2 Malone,J. A., Little, D. E., and Allman, M., "Effects of Hull Foulants and Cleaning/Coating Practices on Ship Performance and Economics," TRANS.SNAME, Vol. 88, 1980. 3 Proceedings, 15th International Towing Tank Conference, M. W. C. Oosterveld, Ed., Vol. I, The Hague, Sept. 1978. 4 Townsin, R. L., Byrne, D., Milne, A., and Svensen, T., "Speed, Power and Roughness: The Economics of Outer Bottom Maintenance," Trans. RINA, Vol. 122, 1980.

Effects of Propeller Design-Point Definition

217

5 Isherwood, R. M., "Wind Resistance of Merchant Ships," Trans. RINA, Vol. 115, 1973. 6 Kan, S., Shiba, H., Tsuchida, K., and Yokoo, K., "Effect of Fouling of a Ship's Hull and Propeller Upon Propulsive Performance," International Shipbuilding Progress, Vol. 5, No. 41, Jan. 1958. • 7 Holtrop, J., and Mennen, G. G. J., "An Approximate Power Prediction Method," International Shipbuilding Progress, Vol. 29, No. 335, July 1982. 8 Bussler M., "Die Berechnung des Reibungsbeiwertes und Reibungsmassstabeinflusses von glatten und rauchen Propellern,' Schiffstechink, Band 3, 1955/56. 9 Ferguson, J. M., "The Effect of Surface Roughness on the Performance of a Model Propeller," Trans. RINA, Vol. 100, 1958. 10 Byrne, D., Fitzsimmons, P. A., and Brook, A. K., "Maintaining Propeller Smoothness: A Cost Effective Means of Energy Saving," Symposium on Ship Costs and Energy, SNAME, Sept. 1982. 11 Townsin, R. L., Byrne, D., Svensen, T. E., andMilne, A., "Estimating the Technical and Economical Penalties of Hull and Propeller Roughness," TRANS.SNAME, Vol. 89, 1981. 12 Broersma, G. and Tasseron, K., "Propeller Maintenance-Propeller Efficiency and Blade Roughness," International Shipbuilding Progress., Vol. 14, 1967.

18 Naess, E., "'Reduction of Drag Resistance Caused by Surface Roughness and Marine Fouling," Norwegian Maritime Research, No. 4, 1980. 14 Grigson, C. W. B. "Propeller Roughness, Its Nature, and Its Effect upon the Drag Coefficients of Blades and Ship Power, RINA, Supplementary Papers, Vol. 124, July 1982. 15 Patience, G., "The Contribution of the Propeller to Energy Conservation in Ship Operation," SMM Technics Paper No. 20, May 1982. 16 Report of Performance Committee, 14th International Towing Tank Conference, Ottawa, Vol. 3, 1975. 17 M. Aucher, "Useful Points of View on the Section Drag on Propeller Characteristics," prepared for the International Towing Tank Conference, Performance Committee Meetings, Tr0ndheim, Norway, Sept. 1973. 18 H. Lerbs, "On the Effects of Scale and Roughness on Free Running Propellers," Journal of the American Society of Naval Engineers, No. 1, 1951. 19 Ferguson, J. M., "The Effects of Surface Roughness on the Performance of a Model Propeller," Trans. RINA, Vol. 100, 1958; written discussion by J. Baird.

Discussion Inserting some typical figures, the above formula gave cot service This discussion is principally concerned with the first part = 0.37 while the authors' modification gave cot serviee= of this important paper dealing with the additional loadings on 0.39. Each of the three issues raised so far indicate some overestithe propeller from the exigencies of service. The increasing use of self-polishing copolymer antifoulings mation of the effects of hull roughness by the authors, but alimplies ships whose intact paint remains weed-free and shell- ternatives are offered. Propeller roughness and its effects present especially difficult free in service. Only the damaged areas are then subject to • problems and it may be as well to repeat some of the cautions fouling penalties. The polishing paints and a general improvement in paint application procedures during the past few given in reference [11]. Average Hull Roughness (or MAA) is an inappropriate measure for propeller surfaces. The short years allow the more optimistic view of surface deterioration wavelength cutoff determined by the stylus ball diameter and to be taken from reference [4]. the long wavelength cutoff are critical to the value ok the The ITTC correlation allowance accounts not only for roughness, but also, most importantly, for length. If it is to be roughness height measure; examples are given in reference [11]. Texture measures are also likely to be of importance to the drag. used as a roughness penalty predictor, then it should be reduced The Schlichting sand grain formula quoted by the authors is also as noted in [11] and by about one half. A more up-to-date formulation accounting also for speed [20] (additional refer- difficult to apply since the drag comparability between a 100 percent dense sand grain surface and a Colebrook-White surences follow some discussions) is face, such as a propeller blade, is unknown. For the purposes of the authors' propeller design-point cal10aACF = 44 - 10(Rn) -1/a + 0.115 culation routine, however, improved and simplified methods for both measuring and defining propeller roughness are likely where Rn is Reynolds number. to be available in the next few years. Meanwhile, use can be A resulting change to the authors' formula would then be made of Musker's equivalent height measure h' in the roughness p(WS)(Cs)(O.5144Vs) a function when calculating rough section drag coefficients for PE roughness = (]50 use in a propeller design routine for KT and K O. An interesting It is worth noting that many new ships today enter service with recent example of rough propeller power penalties is given by an average hull roughness less than 100 = # m AHR; 150 #m Svensen [21]: is now a poor new finish. The authors' modification to the Ap - - × 100% = 1.107(h') 1/3 - 1.479 for h' > 8 ITTC wake scaling formula (which has a + sign missing as P printed) is open to question. An alternative approach is as It may be that the 1/3rd power law may be as useful a rule of follows, We may write thumb for rough propellers as it is for rough hulls. COTSservice = (t + 0.04) + (COTM t 0.04) R. L. Townsin, Member

- -

- -

X (1 + k)CF

trial -]-

ACF

trial +

Cs

(1 + k)CrM COTStri~l = (t + 0.04) + (WTM -- t -- 0.04) X (1 '{-

k)C F trial

4-

mE F trial

(1 + k )CFM from which, by division we obtain coTS service = (t + 0 . 0 4 ) + (coTS trial -- t - 0.04) X 1 + (1 + k)Cv trial + ACe trial"

218

Additional references 20 Townsin, R. L., "Bottom Condition and Fuel Conservation," Proceedings, VIII WEGEMT Graduate School, Gothenburg, Sweden, Aug. 198& 21 Svensen,T. E., "The Economics of Hull & Propeller Maintenance Examined in the Face of Uncertainty," Trans., NECIES, Oct. 1983. R. B. Couch, Member

I have read this paper with considerable interest. The authors obviously have put a lot of work into it. I am not prepared to discuss it in depth; however, a few comments: Presumably

Effects of Propeller Design-Point Definition

as more reliable data become available estimates of the effects of bottom roughness and propeller roughness can be made with great confidence. It probably can also be said that dry docking, bottom cleaning and propeller cleaning carried out properly and regularly will reduce the effect over the lifetime of the ship. Recently we had the opportunity of retesting some old model propellers and found that the efficiency had dropped approximately five points over a period of several years. This difference was checked independently at two laboratories. Close inspection of the model propeller did not show obvious damage or roughness. This surprising finding emphasizes the importance of propeller conditions.

layers of containers, which can make a difference in power requirements of up to 10 percent. As manufacturers also of controllable-pitch propellers, we have no doubt that a point can be made for CP propellers and it would indeed be very interesting to evaluate the pros and cons in a detailed independent study. From operations in the Great Lakes with CP propellers coupled directly to slow-speed engines, we know for a fact that the wear rate of liners, rings, etc., is considerably better than on an equivalent engine running with a fixed-pitch propeller. Also, in view of the ever-increasing use of PTO's (power takeoff's) even on slow-speed engines, a constant shaft rpm is very attractive.

Ernst P. Jung, Member

The authors have presented a thorough treatment of propeller design-point definition, and its importance in establishing meaningful performance requirements for contract specifications. The propeller absorption analysis presented in the paper appears to be a technically sound procedure for evaluating propeller design points, assuming the user can reasonably predict hull and propeller roughness characteristics and their associated performance effects throughout the ship's service life. My comments pertain to the guidance providedby the authors in these areas. Figure 28 accompanying this discussion shows the wide range of Average Hull Roughness (AHR) values that may apply to a ship at any point in its service life. Also, the fouling factors that were taken from my 1980 work (reference [2] of the paper) should have been adjusted for the present work using the original source material from logbook data analysis as contained in references [22] and [23] of this discussion, bearing in mind that foulant macroroughness will completely mask substrate microroughness. To determine the effect of hull roughness on frictional resistance, the authors use the formula recommended by ITTC-

The authors are to be congratulated for a very well written and precise paper on a timely subject. It is a pleasant surprise to note for a change that the increase in power is indeed attributed to the deterioration of the hull and propeller as well as the various operating conditions rather than to the deterioration of the diesel engine. We have indeed held very lengthy discussions on this subject during the past few years and, of course, depending upon the particular naval architects at the consultants, shipyards and owners, you get requests from practically no margin all the way up to 80 percent, mostly divided between the engine and the vessel. As very clearly documented in the paper, the actual requirements are some-• where in between, The rule of thumb of a propeller absorbing 85 to 90 percent of MCR at nominal engine rpm, which we have used for a considerable time, is generally still acceptable. However, it is vital that a greater emphasis be placed on the type of vessel and the intended service. As has been proved, even if we can run a containership on trial with full-load draft, it is an entirely different matter when the deck is stacked with four or five

John Austin Malone, Member

1200

1100

1000

E 900 :=L

,K v ~) uJ z :c

E ,o

800

700 0 I 0 n,,

600 0 n-

0 L q. o

500

uJ 0 < tel

400 0 300

20O

IO0

0

2

4

6

8

-

10

12

14

16

18

20

Y E A R S IN S E R V I C E

Fig. 23 Effects of Propeller Design-Point Definition

219

WE (KRESIC & HASKELL, RO-RO)

/

O.40-

= 0.56 ] AT 450 DAYS

w n i

/

EVENT 2

O.35-

.A

0.30- ~,,

! ! I

0.250

40

80

120

160

DAYS OF FOULING AFTER RE-PAINTING

'REF. 22 DATA FOR C4 BREAK BULK SHIP

+0.10-

~o ° Z O k-

o t~ u.

z

-KRESIC & HASKELL, RO-RO 6w=70.0Cs•

]+REF. 22 DATA FOR C7 I CONTAINERSHIP (NO . Cs) ~ +0.05 - ~ TREND, OF 6w VS.

• ~

[

0.00-

•

+e

~" REF. 22, C4 BREAK BULK SHIP 6w = 35.7C s + O.O13

Z < -TO

-0.05 0

I

I

I

I

I

I

I

I

I

I

1

2

3

4

5

6

7

8

9

10

•C s x 10 4

Fig. 24

1978, which does not account for the effects of varying sur{ace texture or roughness distribution on ACF. The authors' reference [11] provides a critique highlighting the shortcomings of the source data and associated data reduction techniques used in developing that formula. With regard to reliability, reference [11] suggests that the formula overestimates the hull roughness-induced performance penalty by 7 to 9 percent, while reference [24] herewith suggests that it underestimates the penalty by 9 percent, both based on the results of ship performance monitoring experiments. To determine the effect of hull roughness on wake fraction, the authors use an adaptation of an ITTC-1978 formula which, again, does not account for the distribution of roughness over the wetted surface with its significant influence on boundarylayer development and wake fraction. Further, Fig. 24 herewith shows that the wake fraction changes predicted for the RO/RO ship using this formula appear inconsistent with respect to full-scale data from references [6] and [22]. On propeller roughness, the authors have not addressed surface texture (a) or wavelength cutoff values (k), the importance of which are emphasized in their references [10] and [11]. Also, in compiling propeller roughness data from the literature, the authors appear to have intermixed different roughness height definitions, including MAA, Rtm, Ra, and Hmax. For the record, the ISO standards use the definition of average roughness height (Ra) with long wavelength cutoff of 0.8 mm, rather than a peak-to-valley height (MAA) as stated in the paper. Considering this confusion over definitions, and the authors' understatement on the limited extent of mea220

surement data, their propeller roughness time history must be recognized as highly speculative if not totally arbitrary. With regard to the effect of propeller roughness on propulsive efficiency, the ITTC-1978 equations do not account for variations in surface texture, for the relative roughness on the suction and pressure sides, not for its spanwise or chordwise distribution. In fact, the form of the ITTC-1978 formulation suggests that kp is not a measurable roughness parameter but the Schlichting sand grain roughness, and, to the best of my knowledge, there is no accepted relationship for establishing equivalence between measurable parameters and Schlichting sand grain roughness. In summary, shortcomings of present capabilities to reliably predict time histories of hull and propeller roughness, and to reliably determine the associated effects on ship performance, leave me with the feeling that the authors' propeller absorption analysis technique is not as practical as it appears. It is a logical approach for evaluating propeller design-point definitions, but the results of such evaluations can be no more reliable than the input roughness time histories and roughness-induced performance models. Looking ahead, as the roughness phenomenon is better understood and improved roughness-control techniques become widely implemented, the propeller design-point problem will be greatly simplified by narrowing the range of hull and propeller surface roughness conditions over which' a propeller/diesel engine system must operate in service. Additional references 22 Hamlin, N. A. and Sedat, R., "'The In-Service Roughness AI-

Effects of Propeller Design-Point Definition

lowance: Effects of Drydocking, Recoating and the Passing of Time," SNAME ShipbOard Energy Conservation Symposium, Sept. 1980. 23 Malone, J. A. and Allman, M., "Hull Performance Assessment Model," U.S. Maritime Administration Report No. MA-RD-98080015/6/7/8, Jan. 1980. 24 Gronwall, P. E. and Zink, P. F., "Containership Operator's Program of Bottom Maintenance for Reducing Fuel Consumption," SNAME Spring Meeting/STAR S37mposium, April 1982. R o b e r t E. R e i d , M e m b e r

17

-

"?, o 15 _

MCO

I

J !

NSO

I

X a_

This paper is an important contribution to the state of knowledge of the in-service performance on the ship-machinery-propeller system, and in particular its impact on propeller design-point definition. The results presented in the paper are based largely on theoretical or empirical relationships due to the lack of experimental full-scale data, as carefully pointed out by the authors. The authors also stress, as has been done in, for example, references [4] and [10], the need for both a routine hull and propeller maintenance program and the establishment of in-service performance monitoring. We at Erskine Systems Control, Inc., under contract to the Maritime Administration, have developed an onboard Ship Performance Monitoring System to provide these very data for these reasons. The system, shortly to be installed aboard the SS AImeria Lykes, measures and monitors, in real-time, many of the critical performance parameters of the ship/propeller/ engine system described in the paper, using several key measurement sensors and on-line digital processing techniques. The system is designed to identify and track the changing propeller/propulsion parameters, as a function of hull/propeller degradation and environmental and operational factors. It is

fully automatic, is self-learning and adaptive, and has the ability to both calibrate sensor errors and to detect and diagnose sensor failures. It represents the state of the art in on-line identification, estimation and performance monitoring technology. The principal stimulus for the current program is the recognition, based on the recent and ongoing research into the effects of hull and propeller degradation, of the need for on-line performance monitoring if, as so well justified in the paper, the "ship" as a system is to be operated to both meet the in-service performance expectations of .the owner and run as efficiently as possible. We hope to be able to report the results of our work to the Society before long. As a comment on the authors' assumptions regarding identification and correction of damage, reference [11] seems to show that it cannot be assumed that damage to hull or propeller will be immediately diagnosed unless, of course, an on-line performance monitoring system is in place aboard ship. The authors' conclusions on the advantages of a CPP installation are also interesting. Optimum operation of the system to a specified criterion depends on current knowledge of the characteristics of the ship, engine and CPP at that time. The selection of the correct operating point for the engine/CPP combination is still a relatively complex problem, which is best solved, we believe, using the outputs from an on-line performance monitoring system and the correct human/computer interface. The current state of the art in the relevant display technology [25} is shown in Fig. 25 herewith. We have configured our system as a distributed intelligence network aboard ship, to enable local decision-making and control at various points on the ship, and in particular to provide the means to select the correct controls for optimum operation of diesel engine and CPP propulsion systems. The authors also briefly refer to the effects of steering on resistance. Added resistance related to steering, in both calm water and waves, does not have a significant effect on ship speed/power performance, especially for a diesel-powered ship, as has been shown in [26]. The effect of wake fraction variation due to yawing results in an interaction between the diesel en-

S

3 uJ

_

2 ,/ tI

I

40

50

60 70 Propeller Speed (rpm) ( ~ ) ~ P r o p e l l e r Law Curve at Roted Propeller Pitch ( ~ , ) ~ M e i n Engine Output Limitation Curve ( ~ ) ~ M o i n Engine Speed Limitation Line @ ~ B a r r e d Speed Renge of Operation Due to Torsional Vibration (~)-----Turbo Charger Blower Surging Zone ~ ) - - - - I s o - F u e l Consumption Rate Curve (Z)----Minimum Fuel Consumption Rate Curve (~)m~lso_Ship Speed Curve (~lso-Piich Curve ----ALC Setting Curve • Current Opereting Point (~.-e---Pest Opereting Points end their Trock

~

Fig. 25

CPP propulsion system display and relationship of system operating variables

gine and steering control systems. Figure 26 [26] shows the results of full-scale tests for a diesel-powered containership using different steering gear control systems in essentially calm water. Both heading and rudder angle recordings, and measured steering and propulsion performance, are included. The changeover from proportional steering gear control to bangbang control resulted in an average decrease of 1.8 percen t in ship speed and an average increase of 0.4 percent in shaft torque over the measurement period. The wear of the steering gear, as discussed, [0r example, in [27[, will also have an impact on propeller performance. One of the additional goals of our program is to gain further knowledge of the complex dynamics involved in this particular problem, which has been shown [26] to result in losses equivalent to between 2 and 4 percent of normal full power. Additional references 25 Yamashita,F., "Advanced Propulsion Operation System--New Operation System in NKK's New Energy-Saving Ships," Proceedings, 4th IFAC/IFIP International Symposium on Ship Operation Automation, Genoa, Italy, Sept. 1982. 26 Reid, R. E. et al, "Energy Losses Due to Steering Gear Installations on Merchant Ships: Theory and Practice," SNAME Ship Costs and Energy Symposium, New York, Sept. 1982. 27 K~illstr/Jm,C. G. and Norrbin, N. H., "Performance Criteria for Ship Autopilots--An ~-nalysis of Shipboard Experiments," Proceedings, Symposium on Ship Steering Automatic Control, Genoa, Italy, June 1980.

Effects o f Propeller Design-Point Definition

221

Type 4 System

I"

q"

Switching Point

Type 2 System

q

~£

0

12

Time

Rudder Rudder Angle Serve Variance Type {~;2 -- 32}

0-12 Tin Type 4 24-36 Tin Type 2 Fig. 26

0.188 deg 2 1.783 deg 2

Yaw Rate Variance

Mean Square Average Speed Average Shaft Heading Error Through Water Torques

{~'2- r~

{{~'}

{7}

{O}

0.0037 (deg/sec) 2 0.0066 (deg/sec) 2

0.033 deg 2 0.495 deg 2

18.77 knots 18.43 knots

72.85% of max 73.15% of max

Measured steering and propulsion performance: proportional versus bang-bang steering systems, 610-ft-long containershi p in calm weather, North Atlantic, Feb. 1982

R. Latorre, Member

The authors have presented an interesting and timely paper on the influence on the propeller design-point definition. They have succeeded in making a clear presentation in Figs. 8 through 21 of the consequences of the different design points. Since additional studies will be made following the authors approach, it would be worthwhile if the authors would extend Table 4 to indicate the Blade No. Z for the R O / R O and OBO. In selecting the optimum propeller at each design point, I to VII, did the program examine different bladed propellers? If so, how was Ae/Ao or the resulting blade number influenced by the design-point definition? In references [28] and [29] herewith Jiang and Cui made a study to the influence of the design point on the optimum propeller diameter, P/D, and area ratio, Ae/Ao. Table 12 illustrates their results for a single-screw bulk carrier with a 12 000-bhp MCR diesel at 122 rprn. They used the Japanese AU propeller charts for the propeller design. Did the authors notice a similar trend in P/D and Ae/Ao corresponding to Nos. 4, 5, and 6 where the rpm n and power is fixed and the propeller diameter D is also constant? In this case a 5 percent addition in the resistance results in a 0.3 percent drop in the open-water propeller efficiency at the match point. The authors are to be congratulated on their study. It is hoped that they will continue it to examine the sensitivity of the propeller design variables in the context of extended operation over a period of years. Additional references 28 Jiang, W. and Cui, C., "On the Propeller Design Point of Diesel Powered Ships," Trans., Chinese Society of Naval Architecture and Marine Engineering, No. 74, July 1981, pp. 23-33 (in Chinese). Table 12

222

56

24

Time (Tins)

No.

Power

1 2 3 4 5 6

100% 90% 100% 100% 100% 100%

29 Latorre, R., Luthra, G., and Tang, K., Report No. 249, Department of Naval Architecture and Marine Engineering, University of Michigan, Ann Arbor, Sept. 1982, pp. 35-44. Mark F. Nittel, Member

This paper presents an excellent overview of factors which must be considered during the design of a propulsion system. The consideration of hull roughness, propeller blade roughness, and changes in the propulsion components as a function of service life is a welcome approach. The formulas for estimating future changes in effective horsepower and wake fractions are useful tools for propulsion plant designers. The seven propeller design-point definitions provide excellent examples of the implications of service margins selected during the design process. The authors state that the controllable-pitch propeller is perceived by some shipowners as a device with "doubtful reliability." The perception is not necessarily correct, as examination of the 34 ships of the DD 963 and DD 993 classes shows that the CP propellers have indeed proven to be highly reliable. Of all the incidents reported by these ships, only 0.75 percent relate to the CP propeller system. Further, there has been, to the knowledge of Bird-Johnson Company, only one occasion when a ship reported not ready for sea as a result of CP system difficulties in the more than 130 equivalent ship-years of operation examined. It is hoped that the authors will consider this in their future work.

Authors' Closure The authors would like to thank all the discussers for their comments. We found them most interesting and regard them as valuable additions to the paper. We would particularly like to thank Dr. Townsin for his

Propeller design for a single-screw bulk carrier, 12 000 bhp MCR, 122 rpm

Design Point rpm Resistance 100% 100% 103.5 100% 100% 100%

100% 100% 100% 110% 115% 120%

Dia, m

Propeller

P/D

Ae/Ao

Open-Water Efficiency

5.43 5.31 5,28 5.43 5.43 5.43

0.881 0.881 0.883 0.87 0.865 0.860

70% 74.3% 70% 73% 75% 77%

56.5% 55.4% 55.1% 55.0% 54.6% 54.3%

Effects of Propeller Design-Point Definition

references for roughness allowance coefficient and propeller deterioration. As Dr. Townsin states, the increased use of self-polishing copolymer (SPC) antifouling paints is helping to decrease the effect of roughness on ships. However, in the paper we have considered only the use of conventional antifouling paints, which are still the most prevalent paint type being used. The effects of using the SPC paints could easily be ex~imined with our method by using the appropriate hull roughness factors. Based on Dr. Townsin's comments on our in-service effective wake fraction formula we have to agree somewhat with his modification. There are two differences between his formula and ours: the keeping of the 0.04 term in the formula and the denominator of the last part of the formula. We do agree with the addition of the roughness allowance coefficient in the denominator of the formula as Dr. Townsin suggests, owing to the fact that this should be the total frictional resistance on trials including roughness resistance. However, we still feel that the 0.04 term should not stay in the formula. As stated in this paper, this term is meant to take into account the effect of the rudder in going from the model to full scale. However, when predicting the wake from trial to service, it is no longer necessary to carry along any correction for the rudder. The modified formula would be

WT service= t + (WT trials t) -

-

+

cs

]

(1 + k)CF + ACF trials

The difference in calculated wakes between this modified formula and the one given in the paper is only about 1 to 2 percent. The large differences Dr. Townsin quotes are mainly from the retention of the 0.04 term. Dr. Townsin indicates that the ITTC-78 ACF formula when used as a roughness penalty predictor overestimates ACF, which would lead to a higher Pe roughnessthan would actually occur. However, since finishing the paper we have had an opportunity to apply the method to an actual ship and found PE roughn~ was underestimated. We found that the underprediction resulted mainly from underprediction of the hull fouling effects. This appears to be supported by Mr. Malone's comments, which suggest that the fouling factors taken from his paper should be reanalyzed. Using the fouling rates of the paper will lead to conservative results, which are somewhat compensated for by using the ITTC-78 ACv formula. To date we have been unable to find a satisfactory replacement for the fouling factors given in the paper, or to reanalyze Mr. Matone's data. But for guidance we have found work in the Russian literature, based on statistical analysis of ships of the Russian fleet [30], which shows that the total resistance will increase between 30 and 40 percent in a 24-m0nth docking period for conventual antifouling paints and depending upon the fouling severity. The impact of using higher resistance on the propeller absorption curves displacement would be up and to the left in the diagram, creating even worse situations than are shown in Figs. 8 through 15. Therefore, problems such as overloading, engin e performance deterioration or speed loss will occur sooner in the life of the vessel and be intensified. Dr. Townsin and Mr. Malone are certainly correct in their cautions on the problems of measuring the roughness on the propeller. This probably was the area in which we had most difficulty gathering information. However, we feel that Fig. 2, which compares test KT and KQ with those calculated, demonstrates that the method is giving reasonable results. There is room for improvement in the area of estimating the rate and radial distribution of roughening of the propeller.

There is certainly need for more statistical data from the fullscale measurements. Much of Mr. Malone's discussion focuses on the fact that we have not taken into account the variation of roughness and texture over the ship or propeller. We were aware of the work in this area; however, that work has not yet been able to quantify these effects to the point of predicting changes in ACF or ACo. Therefore, we chose to ignore these details of the roughness in favor of the overall roughness measurement. We feel that Mr. Malone's Fig. 23 actually supports our hull roughness predictions. Our hull roughness curve is within the bounds of the roughness envelope through 16 years, which seems to indicate that the hull roughening prediction is reasonable. Mr. Malone's comments on the effective wake fraction formula indicates that we were not as clear on its use as we could have been. Our wake fractions formula predicts service wake from trial wakes based on changes in the total hull roughness. Mr. Malone's interpretation of the effective wake, as shown in his diagram illustrating the wake unchanging for a period of time, would be correct if roughness allowance coefficient was influenced only by fouling. However, ACv is influenced by many factors before fouling becomes significant, as pointed out in the paper. Therefore, ACF begins increasing the moment the ship returns to the water and consequently the wake is also increasing steadily. Also, Mr. Malone's comparison of the changes of wake compared with those published for the C4 break-bulk ship is not a valid comparison. His reference [22] indicates thats Aw is the change from model tests to service; however, the Aw taken from our paper is the change from full-scale trials to service. By comparing these curves on the same basis, a consistent trend is shown. We were not aware of the ship performance monitoring system which Dr. Reid is implementing. The type of statistical data that could be produced by this type of program will be very useful, though the full benefit of such a program can be realized only if the monitoring is coupled with the tracking of the ship and propeller roughness by measurement. Only with both kinds of information would it be possible to Validate the present correlation methods between roughness and its ultimate effects on power absorption, revolution, and ship speed. We look forward to the results of this study and hope more programs of this type will be undertaken in the next few years. In response to Dr. Latorre, for both ships analyzed in the paper, only four-bladed propellers were used. Because the choice of the number of blades is determined most often by vibrational considerations rather than by efficiency, we chose only to investigate four-bladed propellers for this study. The diameter, pitch and blade area were optimized to provide the highest open-water efficiency; however, for our ships, as is often the case, the maximum diameter was restricted by physical and operational constraints and did not permit the use of the optimum propeller diameter. In Tables 13 herewith we have given the propeller characteristics for each of the propeller design definitions, along with the open-water efficiency, for beth the RO/RO and OBO ships described in the paper. Also given is the optimum propeller diameter for each definition. As can be seen, in all cases the optimum diameter was quite a bit larger than permitted. We were not surprised by Mr. Jung's comments that he is sometimes involved in designs where no margin is requested. This shows tbe necessity of educating people on the consequences of inadequate margins. We have been involved in similar cases where the use of insufficient margins led only to headaches for the shipowner. We hope this paper will help to convince shipowners and ship operators of the importance of

Effects of Propeller Design-Point Definition

223

Table 13

Definitions

RO--O/R

OBO

I II III IV V VI VII I II III IV V VI VII

D, m

P/D

Ae/Ao

770

DOPT,m

6.30 6.30 6.30 6.30 6.30 6.30 6.30 7.62 7.62 7.62 7.62 7.62 7.62 7.62

0.916 0.880 0.899 0.891 0.857 0.835 0.820 0.916 0.879 0.895 0.891 0.846 0.821 0.803

0.654 0.639 0.667 0.674 0.678 0.687 0.693 0.518 0.503 0.533 0.536 0.542 0.551 0.558

0.615 0.618 0.597 0.588 0.623 0.625 0.626 0.520 0.524 0.485 0.479 0.526 0.527 0.528

6.78 6.76 6.82 6.82 6.59 6.50 6.46 8.57 8.46 8.62 8.64 8.35 8.28 8.20

having the correct margin for their designs, and that a thorough propulsive analysis of a ship at the early stages of design is important. Mr. Jung is right that diesel engines are often unjustly the first to be blamed when ships start to perform poorly, when there are other factors involved. However, we are sure Mr. Jung will agree that there is a deterioration in performance of the engine with time. We did not include this aspect in the method;

224

however, we plan to try to research it further in an effort to also take this phenomenon into account. We would like to thank Professor Couch for his comments on the change in efficiencies of the stock model propellers. This is another illustration of the importance of keeping propellers clean at all times. Even very small increases in roughness that are unnoticible by visual inffpection will cause degredations in performance. We would like to thank Mr. Nittel for his comments on CP propellers. We do agree that, today, CP propeller are generally no less reliable than fixed-pitch propellers and offer many benefits. Unfortunately, many shipowners believe, correctly or uncorrectly, that CP propellers are less reliable than their fixed-pitch counterparts along with being more expensive. It appears that the manufacturers of CP propeller systems must do more to dispel this notion, through studies using the method shown in the paper. In closing, we hope that the paper has given a better appreciation of the influence of roughness on the performance of ships and the importance of a good hull maintenance program and adequate propeller design margins. Additional reference 30 Dityatev, S. G. and Katsman, F. M., "Prediction of Ship Propulsive Performance in Service," Sudostroenie, No. 11, 1982.

Effects of Propeller Design-Point Definition