Hoyle_J_W.Bulbous_Bow_Design_M.1986.TRANS.pdf

SNAME Transactions, Vol. 94, 1986, pp. 31 -56

A Bulbous Bow Design Methodology for High-Speed Ships

Jeff W. Hoyle, 1 Associate Member, Bill H. Cheng, 2 Visitor, Bruce Hays, 3 Associate Member, Bruce Johnson, 4 Member, and Bruce Nehrling, 5 Associate Member

A methodology for designing bulbous bows for high-speed, fine-form ships is proposed. Using the Kracht bulbous bow design curves developed for low-speed, full-form ships as a starting point, a series of bulb forms is developed and analyzed using a combined numerical and experimental approach to ascertain resistance and seakeeping characteristics. This study was performed using the FFG-7 class of naval frigates as the reference hull form. Nine variations in bulb design (including one similar to that found on the Italian frigate Maestrale) plus the bulbless hull form were analyzed using the DTNSRDC's XYZ Free Surface Program. Five of the bulb variations were appended to a model of the FFG-7 and tested in the 116-m (380 ft) towing tank at the U.S. Naval Academy. The results from the computer predictions and the calm-water towing tank tests show remarkably similar trends while the relative rankings of the bulb forms derived from these analysis procedures were identical. Furthermore, the addition of a bulbous bow to the FFG-7 hull form appeared to only marginally degrade the ship's seakeeping characteristics.

Introduction THIS is an exciting time to be working in the field of

experimental naval architecture. When numerical hydrodynamic flow code predictions are integrated with experimental towing tank procedures, the ship design process is enhanced. The ability to numerically predict flow patterns around a hull and ship motions in a seaway enables the naval architect to perform a computer-based order-of-merit ranking of hull form candidates before starting a model test program. Tra- ditionally, the early stage of hull form design has been based on variations of existing hull forms whose hydrodynamic characteristics are stored in a data base accessed by regression analysis. These regression analysis methods are adequate for the concept design stage so long as the hull form variates lie within the limits of the data base.

Today, the possibility of calculating these hydrodynamic characteristics directly allows us to investigate hull forms not included in a data base [1]. 6 Tank testing is still required, however, since the flow calculations do not accurately predict the actual flow fields and viscous effects. Nevertheless, the flow codes are sufficiently accurate in a relative sense to attempt to optimize the hull form before model testing begins. This approach to hull design was successfully used with

1 Ensign, USN; Nuclear Power Program. 2 Physical scientist, David W. Taylor Naval Ship Research and

Development Center, Bethesda, Maryland. 3 Naval architect, Design Systems & Services, Inc., Annapolis,

Maryland. a Director, Hydromechanics Laboratory, U.S. Naval Academy,

Annapolis, Maryland. 5 Professor and Director of Naval Architecture, U.S. Naval Acade-

my, Annapolis, Maryland. 6 Numbers in brackets designate References at end of paper. Presented at the Annual Meeffng, New York, N.Y~, Noverfifier 19-

22, 1986, of THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERS.

The views expressed herein are the opinions of the authors and not necessarily those of the Department of Defense or the Department of the Navy.

dramatic results by those Dutch scientists and naval architects who assisted in the 1983 Australian challenge for the Ameri- ca's Cup [2].

An area in which relatively small changes in hull form geometry can lead to significant changes in ship resistance is in the use of bow bulbs. Since the turn of the century, naval architects have realized that a reduction in the overall resistance of a ship can be achieved by the addition of a bulbous bow to the hull form. For instance, Admiral David W. Taylor recognized these effects when he fitted a bulbous bow to the battleship, Delaware, to increase her speed at constant power [3]. Today, nearly 80 years later, the bow bulb is utilized routinely in the design and construction of surface ships. However, most research concerning optimal design~ and power prediction for ships with bulbous bows, such as Kracht's [3], has concentrated primarily on low-speed, full- form ships such as merchant ships, naval auxiliaries, and amphibious ships. Thus, bulbous bow design criteria for high-speed, fine-form vessels such as destroyers and frigates are relatively unknown. To date, the only full-scale application of a large bulbous bow to such a hull form is that found on the Italian frigate Maestrale. In this paper, the results are presented of an investigation of bulbous bows for a typical naval frigate (FFG-7 appended with 15-deg stern wedge). The study was conducted at the U.S. Naval Academy (USNA) in connection with the David W. Taylor Naval Ship Research and Development Center (DTNSRDC) and the Naval Sea Systems Command (NAVSEA).

Objectives The objectives of this research project were threefold: 1. Verification of the usefulness of numerical hydrody-

namics in bow bulb design by direct comparison of computer predictions of resistance and seakeeping characteristics of the FFG-7 appended with various bow bulbs to actual model test results.

31

(~b) ( I )

/~ -TYPE O-TYPE ~ - T Y P E

Fig. 1 Bulbous bow types

2. Determination of the applicability of the NAVSEA interpretation of the Kracht bulb design charts to fine-form, high-speed vessels such as the FFG-7.

3. Investigation of the effects of changes in bulb length and breadth on the resistance and seakeeping performance of the FFG-7 hull form.

Bow bulb theory The reductions in drag achieved by the addition of a bul-

beus bow to a hull form are derived primarily by the lowering of wavemaking resistance through attenuation of the ship's bow wave system. Furthermore, there is reason to believe that a bulb acts to reduce viscous resistance as well, by smoothing the flow around the forebody [3]. Since the beneficial action of a bulbous bow depends heavily on the waves it produces and the flow around it, it is quite obvious that the size, position, and form of the bulb body will have a marked impact on the resistance characteristics of the ship.

For slender, fast, hull forms such as the FFG-7, the primary reduction in resistance is due to the reduction of the free wave system of the ship. This reduction of the free wave system is accomplished by cancellation; that is, depending upon the phase and amplitudes of the waves created by the bulb and the ship, the two may cancel totally. The phase difference of the two wave systems is determined by the location of the bulb, and the amplitude of the bulb's wave is determined by bulb volume [3].

Recognizing the importance of the bulb form in reducing a ship's resistance, it is necessary to classify bulbs according to some geometric parameters. Kracht [3] differentiated bulbs into three main categories according to the shape of the bulb's cross section at the forward perpendicular. These three class- es are presented below and are depicted graphically in Fig. 1.

(a) A-Type: Figure l(a) shows the drop-shaped sectional area of the delta type with the center of area in the lower- half part. This shape indicates a concentration of the bulb volume near the base. The Taylor bulb and the pear- shaped bulbs belong to this type. (b) 0-Type: This type, shown in Fig. l(b), has an oval sectional area, a center of area in the middle, and a central volumetric concentration. All the circular, elliptical, and lens-shaped bulbs as well as the cylindrical bulbs belong to this type. (c) V-Type: The Nabla type, shown in Fig. l(c), also has a drop-shaped sectional area. However, its center of area is situated in the upper-half part, indicating a volume concentration near the free surface. Because of its favorable seakeeping properties, this type is the most common bulb in use today [3].

Bow bulb design Presently, the complexities of the hydrodynamic interac-

tions between a bulb and the main hull prevent a completely analytical approach to bow bulb design. Therefore, the development of an optimum bulb for a ship is an empirical and iterative process. Fortunately, the use of numerical tools to predict the hydrodynamic performance of candidates in a bulb design study affords the designer an opportunity to evaluate many different bulb forms before committing to a series of expensive and time-consuming model tests. In this manner, an optimal bulb form can be developed efficiently and expeditiously. However, to begin the iterative process of optimizing a bulb for a particular hull form, an acceptable initial design must be develped which can then be modified in an attempt to enhance overall performance. In order to provide this starting point, the authors chose to apply the NAVSEA interpretation [4] of the Kracht design charts [5]. These charts were developed by Alfred M. Kracht from a large number of model tests of ships with and without bulbs and having block coefficients ranging from 0.56 to 0.82. Since the goal of this project was to design bow bulbs for the fine, fast FFG-7 hull form (block coefficient equal to 0.45), the necessity of verifying the applicability of Kracht's design charts became readily apparent. To accomplish this task, a trend study was initiated which attempted to detect changes in the optimum value of the major bulb parameters defined by Kracht as a function of changing block coefficient and Froude number. This trend study is presented graphically in Appendix 1. The results of this study were encouraging since it indicated that most of the optimum parameter values vary

Kracht also divides bulbs into two types according to their lateral contours:

(a) Those bulbs which do not change the outline of the stem of the ship.

(b) Those bulbs which protrude forward, thereby altering the ship's stem profile.

This classification is unnecessary however, since bulbs such as the Taylor bulb which do not ciaange the stem outline do not have favorable properties and are no longer in use. Lastly, bulbs can be described as "additive" or "implicit." An "additive" bulb is one which is added to an existing hull form, thus increasing the volumetric displacement of the ship. For an "implicit" bulb application, a portion of the ship's volumetric displacement is shifted forward to create the bulb [3]. All bulbs designed during the course of this study can be described as "additive" and are of the Nabla type.

In addition to reducing the resistance of a hull form, bulbous bows also influence other properties of a ship. For instance, model tests have shown that bulbous bows can influence the quasi-propulsive coefficient, wake fraction and thrust deduction fraction. However, it is not certain if these bulb effects are present in the full-scale ship because of the importance of scale effects on the expansion of these model test results. Bulbous bows do not seem to significantly influence course stability or maneuverability, and model tests in regular waves tend to indicate that the "bulbous ship is the best ship regardless of seakeeping aspects" up to a wavelength to ship's length ratio of about 0.8. For ice navigation, ships equipped with bulbous bows have a definite advantage. The bulbs tend to tip the ice floes so that they slide along the ship's hull on their wet side, which has a smaller friction coefficient. Thus, the speed loss of a ship equipped with a bulbous bow in ice is less than that of the same ship without a bow bulb. As these factors present no reasons sufficient to prevent the utilization of a bulbous bow, it appears that bulb design may be based solely on calm-water characteristics [3].

32 Bulbous Bow Design

only slightly with block coefficient. Thus, it was concluded that the range of the Kracht bulb design charts could be extrapolated downward. The design charts corresponding to 1. the smallest block coefficient (CB = 0.56) were used to develop initial bow bulb parameters for the FFG-7 hull form. These bulb parameters, described by Kracht [8,5], are defined as follows:

1. Breadth parameter (CBB)--The maximum breadth (BB) of bulb area (ABT) at the forward perpendicular divided 2. by the beam of the ship at amidships (BMs): CBB = BB/ BMS

2. Length parameter (CLPR)--The protruding length (Lea) divided by the length between perpendiculars (Lee) of the ship: CLe~ = LpR/Lpp

8. Depth parameter (CzB)--The height (ZB) of the fore- 3. most point of the bulb over the baseline divided by the draft (TFp) at the forward perpendicular: CZB = ZB/ TFp

4. Cross-section parameter (CAax)--The cross-sectional area (ANT) of the bulbous bow at the forward perpendicular divided by the ship's midship section area (AMs): CABT = ABT/AM S 4.

5. Lateral parameter (CABL)--The area (ABL) of the protruding bulb in the longitudinal plane divided by the midship-section area of the ship (A~s): CABL = ABL/ AMS

6. Volumetric parameter (CvPa)--The volume (VpB) of the protruding part of the bulb divided by the volume of 5. displacement (VwL) of the ship: Cvra = VpR/VwL [4]

NOTE: Protruding is used here to mean that part of the bulb which extends forward of the forward perpendicular.

A graphical representation of these bulb parameters can be found in Fig. 2 [4]. 6.

Utilization of the Kracht design charts to derive near-optimum values of the parameters defined above is the goal of the NAVSEA method. Each design chart shows the residual power reduction coefficient 7 (ACevR) as a function of Froude number and one of the bulb shape parameters defined above for a particular block coefficient. This gives a total of six charts for each value of block coefficient. A representative design chart showing the residual power reduction coefficient (ACpvR) as a function of the length parameter (CLpR) is shown in Fig. 8 [4]. The wavy shape of this curve is typical of all the design charts. Thus, a given value of the residual power reduction coefficient can be achieved at more than one value

7 The residual power reduction coefficient is a measure of the percent reduction in power necessary to drive a ship equipped with a bulbous bow as compared to the same ship without a bow bulb. It considers only that power which is necessary to overcome the residuary resistance of the hull form (total resistance minus frictional resistance) and is quantitatively defined in [4] as:

ACpvR = [Cpvao(without bulb) - Ceva(with bulb)]/CevRo

where CpVR ---~ P D/[ (P / 2) VS( V W2L) O'&2G] - - CFS /[ 170( v 2L)0 333]

PD = delivered power p = mass density of water V = speed

VWL displacement volume CF = frictional resistance coefficient

ITTC standard: CF = 0.075/(log(RN) -- 2.0) 2 S = wetted surface

~/o = quasi-propulsive coefficient

Maximizing the residual power reduction coefficient is desirable since a large ACevR indicates a large reduction in residuary resistance as a result of the addition of a bulbous bow.

BREADTH PARAMETER

Cs8 = BB/BMS

LENGTH PARAMETER G_• = Lm/t-~

I

Lpp

DEPTH PARAMETER

CzB = za/TFp

CROSS-SECTION PARAMETER

CABT = Asr/AMS

LATERAL PARAMETER

C^~_ = ABLIAMs

i FP

VOLUMETRIC PARAMETER CVpR = Vm/ VwL

Fig. 2

i

Bow bulb parameters

of the bulb parameter. These multimodal functions further complicate the problem of finding near-optimum values for each of the six parameters. For the purposes of this design, the guidelines set forth in [4] were followed to define the parameters of the first bulb produced in this study. An outline of the chart utilization procedure (taken from [4]) together with notes peculiar to this project is presented in Appendix 2. For all subsequent bulb forms, methodical variations to the shape of the Bulb No. 1 were made by assuming specific values of the length parameter (CLpR) and the breadth parameter (CBB). These variations were made based on the bulb design matrix presented in Table 1, where the numbers inside the table correspond to the number assigned to the bulb which had that particular length and breadth. The design methodology put forth in [4] was then utilized to develop a bulb form with the required parameters. A com- plete listing of the parameters for each bulb designed, as well as the values for the existing Maestrale 0-type bulb, is presented in Table 2. Figure 4 shows a computer-generated profile view of each bulb.

Again, the numbers in that table correspond to bulb number, with Bulb 0 being the existing Maestrale 0-type bulb.

Despite the desirability of a systematic approach to bulbous bow design, the method outlined in reference [4] does have significant limitations. Perhaps the most important disadvantage of this method is that it fails to provide a means for fairing the bulb into the rest of the hull. As is apparent from

Bulbous B o w Design 33

5

C I

0 ol

I0 ~ C p~7R I f -

C8 = 056 obere Orenze / . ' / 70 " 070 (uPP+:a LIMz~ ~ I ' / /

/ ' i "'..+ ," / / / /oz+ 53o //

-~'~" I H~- ~ / : I I Y / ] .

//. '- ,xJ ,&TJi -

~ _~ J o.32 o3 .... p :~.. o.z6.~

0 02 0 03 CLp R 04,

Fig. 3 Representative Kracht design chart

the definitions of the bulb geometry parameters, the design charts are concerned only with that portion of the bulb forward of the forward perpendicular. The integration of the bulb into the ship's hull is left entirely to the designer's discretion, although Hagen and Fung do offer some guidelines on this topic taken from "various experimental investigations." Their report suggests that "'simply continuing the bulb aft using longitudinal elements parallel to the ship's axis is proba- bly as good a solution to the problem as any" [4]. In light of this recommendation and in the interest of keeping other form factors constant throughout this study, all bulbs designed for this project maintain a constant cross section aft of the forward perpendicular until they intersect with the main hull.

C o m p u t e r p r e d i c t i o n s

Two numerical tools were utilized to predict the hydrodynamic performance of the candidate bulbs. First, the XYZ Free Surface Program was used to assess the calm-water resistance characteristics of the FFG-7 configured with and without bulb forms. Then, the Navy Standard Ship Motions Program (SMP) was run to predict their seakeeping performance. An overview of the operation of these two programs is presented here. For a more detailed description of the

Table 1 Bulb design matrix

Bulb Length (fwd of FP) Maximum 1.22 m 2.44 m 3.66 m 4.27 m 4.88 m

Bulb Width (4 ft) (8 ft) (12 ft) (14 ft) (16 ft)

2.13 m (7 ft) 8 7 2 1 3 2.79 m (9 ft) . . . . . . 5 4 6

Table 2 Kracht bulb parameters

Bulb No. CLp a CBB Cz B CABL CABT CVPR

0 0.011 0 . 1 9 4 0 . 2 9 3 0 . 0 6 4 0 . 1 2 5 0.0014 1 0.034 0.165 0.46 0.174 0 . 0 8 6 0.0028 2 0.030 0.165 0.46 0.165 0 . 0 8 8 0.0030 3 0.040 0.165 0.46 0.219 0 . 0 8 8 0.0039 4 0.034 0.200 0.46 0.174 0 . 1 0 6 0.0035 5 0.030 0.200 0.46 0.165 0 . 1 0 6 0.0036 6 0.040 0.200 0.46 0.219 0 . 1 0 6 0.0047 7 0.020 0.165 0.46 0.110 0 . 0 8 8 0.0020 8 0.010 0.165 0.46 0.056 0 . 0 8 8 0.0010

computer programs and their operation, consult references [6, 7].

R e s i s t a n c e

The XYZ Free Surface Program (XYZFS) was used to evaluate resistance performance in order to identify promising design candidates before model construction was begun. XYZFS computed the wave resistance, based on the integration of surface pressure on the hull, for the nine FFG-7 bulb variations presented in Fig. 4. These computations were done at sinkage and trim positions determined by experiments on the Maestrale 0-type bulb at ten speeds corresponding to a Froude number range from 0.177 to 0.442. After wave resistance predictions were made, residuary resistance coefficients were obtained by adding an estimated form drag to the wave resistance coefficients for each of the ten speeds. An estimate of the total resistance coefficient was obtained by adding the International Towing Tank Conference (ITTC) 1957 frictional resistance to the residuary resistance. Finally, an estimate of effective horsepower was obtained from the total resistance coefficient. The powering predictions obtained from the XYZFS Program for those bulbs which were

0

m m m m "R +|-+++++

1 2

3 4 5

/ /

Fig. 4

i

7 8 Scope of bulb designs


1. 121

1 . 1 8

I . OB

1. f16

I . 84

I . 82 o H V- 1.012 n" n¢

0. .gE I" h:

.gE

.q~

.92

.90

.BB

.81

EFFECTIVE HORSEPOHER RFITIO ( EHP HITH BULB RND HEDGE )

EHP HITH HEDGE ONLY C t = METHOD: EHPa = ( C t = ) a * Sa

EHPb ( C t s ) b Sb

LEGEND:

- - B U L B O -----BULB 1 - - - -BULB 4

t BULB 6 ~ I--BULB 8

SHIP SPEED (KTS)

Fig. 5 Effective horsepower ratios from XYZFS output

eventually chosen for model testing are presented as effective horsepower ratios in Fig. 5. EHP ratio was defined as the effective horsepower of the FFG-7 appended with a bulbous bow divided by that of the bulbless hull form. Thus, whenev- er the EHP ratio for a particular bulb is below 1.00, the bulb is lowering the resistance of the FFG-7 hull form.

For each bulb configuration, the following steps were executed sequentially in order to obtain resistance predictions from the XYZ Free Surface Program:

• S t e p / - - T h e bulb form, as specified by a body plan and centerline profile, was digitized at the U.S. Naval Academy's Computer Aided Design Interactive Graphics (CADIG) facil- ity and appended to the digitized FFG-7 hull form.

• S tep 2--The digitized data were transferred to the Da- vid W. Taylor Naval Ship Research and Development Center. A Tektronix terminal received and forwarded the digitized data to a Cyber 176 computer at DTNSRDC.

• S tep 3--The digitized data were used by a B-spline program [8] running on the Cyber 176 computer to generate a surface grid of the hull and of each bulb. The hull, when appended with a bulb, was subdivided into two sections. The first section consisted of a bulb plus that part of the hull up to station 2; the rest of the hull, including the 15-deg stern wedge, formed the second section, which was kept constant during the course of this study. The surface grid, or panel-

IIII / / / I 1~~HI / I I I I/11 /1111 / I\ \ \ \ \ \\\\~ , , , , •

Fig. 6 Panelized representation of the FFG-7 with stern wedge and Bulb 1

Bulbous Bow Design

ized representation, of the FFG-7 hull form appended with a bulb and stern wedge is shown in Fig. 6. Three views of a panelized bulb are shown in Fig. 7,.

• S t ep 4--The hull and bulb data created in Step 3 were carefully checked using numerical tests as well as computer graphics. The numerical model of the FFG-7 hull as repre-

ToP V I E n l

,=PONT Y,'/14/

.vIEW

Fig. 7 Panelized Bulb 1

35

sented by this data was then transferred from DTNSRDC to a Cray supercomputer in Seattle, Washington.

• Step 5--The XYZFS Program was run to analyze each of the bulb variations. The panelized hull and bulb data were used as input to XYZFS. The output gave predictions of. wave resistance, residuary resistance, total resistance, and effective horsepower.

• S t ep 6- -XYZFS results were analyzed at both DTNSRDC and USNA. The surprisingly good performance of wider bulbs at speeds from 18 to 25 knots was first noticed at this phase. As a result of superior performance, as predicted by XYZFS, Bulb Nos. 4 and 6 were recommended for model construction and testing at USNA.

Seakeeping

The Navy Standard Ship Motions Program (SMP) [7] was used to predict the seakeeping characteristics of the FFG-7 hull form in four different configurations: bulbless and appended with bulbs zero, one, and four. Each configuration also included the 15-deg stern wedge. SMP calculated significant pitch and heave amplitudes and the probability of slam- ming for each hull variant as a function of full-scale ship speed ranging from zero to 30 knots. Data were analyzed for the long-crested, head seas condition, with a modal period of nine seconds and significant wave height of 8.7 m (12.14 ft). These values were determined from operator feedback to be conditions at which the full-scale ship begins to slam. The ship motions program was run on a VAX computer at the Naval Sea Systems Command via a telephone link from the Naval Academy. Results were printed at NAVSEA and sent to USNA for analysis. These results, presented graphically in Figs. 8 through 10, indicate very little difference in the seakeeping performance of the FFG-7 with and without bow bulbs, and even less difference between two bulb configurations. Figures 8 and 9 do not show the results associated with Bulb 0 since its predicted significant heave and pitch amplitudes coincided almost exactly with those of Bulb 1. Conse- quently, for reasons of clarity, they were not plotted. Since the SMP results for these three bulbs were only slightly different, the program was not executed for the remaining bulb forms.

M o d e l t e s t p r o g r a m

In order to experimentally determine the resistance and seakeeping characteristics of the FFG-7 when appended with a stern wedge and various bulbous bows and to verify the results obtained in the computer predictions phase of this project, an extensive series of model tests was performed in the USNA Hydromechanics Laboratory's 116-m (380 ft) towing tank. The test program consisted of three tasks: (1) bulb construction and model preparation, (2) effective horsepower testing to determine resistance characteristics, and (8) head seas testing in irregular waves to assess seakeeping performance. Each of these tasks will be discussed separately.

Bulb construct ion and model preparation

Four of the nine bulb candidate designs developed for this study were built. These bulbs were constructed from a high- density foam by the Technical Support Department of the U.S. Naval Academy. Careful consideration of the program's objectives resolved the issue of which bulbs to produce. In keeping with the combined numerical/experimental design methodology, Bulbs 4 and 6 were constructed based on their superior resistance characteristics as predicted by the XYZ Free Surface Program. In addition, Bulb 1 was chosen for construction to further validate the extrapolation of the Kracht design charts' range to lower block coefficients. Since

36 Bulbous Bow

the length of Bulb 8 was nearly identical to that of the Maes- trale-style bulb it was built in order to directly compare the effects of an 0-type bulb and V-type bulb of similar length.

Before testing began, all bulbs were appended to a 1//24.75 scale model of the FFG-7 fitted with a removable bow section and a 15-deg stern wedge. This stern wedge was used because of its favorable effects on the resistance characteristics of the FFG-7 hull form as reported in [9]. Each model configuration was ballasted to a full-scale draft of 4.87 m (14.35 ft) with 0-deg trim. The longitudinal gyradius was set to 25 percent of the length between perpendiculars. Turbu- lence was stimulated with studs placed near the bow. For ease of rigging, all model configurations were towed from the same point. This towing point was at the longitudinal center of gravity of the model when it was configured with the Maestrale-style bulb. While each model had a slightly different LCG, the variations were extremely small. Experi- ence has shown that very small variations between the towing point and the actual LCG do not produce observable differences in the test data. The particulars of each configuration are compared in Table 3.

Resistance tests

Calm-water resistance tests were performed on each configuration to determine the effective horsepower necessary to propel that ship at full-scale speeds ranging from 12 to 80 knots. The variables measured during these tests included speed, drag, sinkage, and trim. The dynamometer used for both the effective horsepower and subsequent seakeeping tests was a Netherlands Ship Model Basin air-bearing rig with a single heave post. The model was free to pitch and heave, and the other motions (yaw, roll, sway, and surge) were locked. Table 4 provides information about the sensitivity of the transducers utilized during these experiments. Results of the effective horsepower tests are presented as effective horsepower ratios in Fig. 11. Again, an EHP ratio of less than 1.00 indicates that the bulb is lowering the ship's resistance. The EHP ratio of Bulb 0 was derived from data that were manipulated to subtract the effects of a calibration error, while all other curves come directly from model test data.

Seakeeping tests

In addition to calm-water resistance testing, head-sea tests in irregular waves were performed on all model configurations in order to assess seakeeping performance. Waves used for this study were periodic encountered irregular waves as described in [11]. Three waves were constructed according to the methods described in that paper, corresponding to ship speeds of 15, 20, and 25 knots. In this manner, each model configuration was tested in three identical irregular wave trains. For most configurations, the model was towed at each speed twice to collect data on resistance, pitch, heave, bow acceleration, and encountered wave height. Unfortunately, the amount of data collected was not sufficient to provide statistically significant results. However, videotapes made of all configurations and speeds tested allowed for a subjective comparison of seakeeping performance. From these videotapes, it was apparent that the addition of a bulbous bow to the FFG-7 hull form tended to degrade its seakeeping performance to a small degree. This fact was evidenced by the more extreme motions and greater amount of water taken onto the deck when the FFG-7 model was configured with a bow bulb. On the other hand, the videotapes also indicated that the bulbless hull form rose higher out of the water in response to the wave action. While the hull seemed to take more water over the deck when configured with a bow bulb, the amount of water taken on appeared to decrease with increasing bulb length. Overall, the small degradation in

Design

2- L

E c__ z o- (IZ o c 4 (n bJ

O (E Ld S

CD (D

,.~- Z3

~D

b_ H z ~2 t~

CO -.j

0 .

SHIP MOTIONS PROGRRM:

PREDICTION OF SIGNIFICRNT

HERVE RMPLITUDE FOR

FFG-? HULL FORM (NO TRIM)

-Long C r e s t e d Waves -Head Seas C o n d i t i o n -Modal Perlod-g Seconds -S|g. Have H e i g h t - 1 2 . 14 £t

= 3.7 meters

f LEGENB: FFG-7 WITH STERN WEDGE AND NO BULB

FFG-7 WITH STERN WEDGE RND BULB i

[ ] FFG-7 WITH STERN WEDGE ~ D BULB 4

5 10 15 20 25 SHIP SPEED (KTS)

Fig. 8 Signi f icant heave amplitude versus ship speed

f ,

3 0

w Q

E

oz

I u

n

t--- z ff] U a

b_

Z k9 H u1

3 • 0,

2.~

2.0

I.

i.

0 .


PREDICTION OF SIGNIFICRNT

PITCH RMPLITUDE FOR


- L o n g C r e s t e d Waves -Head Seas C o n d i t i o n -Moda l P e r i o d - 9 Seconds -Sig. Nave Helght-12.14 Ft

= 3.7 meters

LEGEND: A FFG-7 WITH STERN WEDGE

RND NO BULB

Q FFG-? WITH STERN WEDGE RND BULB l

[ ] FFG-7 WITH STERN WEDGE RND BULB 4

Fig. 9

.... l~ .... /5 .... 2'~ .... ~'5 S H I P S P E E D ( K T S )

Signi f icant pitch amplitude versus ship speed

E;} E~

>- I-- H I

m 02 m O Od Q_

ID Z t-4 >-

(]: _J U]

?.

6.

5.1

4 .

3.

2 .

1.1

61 Fig. 10


PROBRBILITY OF SLRMMING

AT STRTION TWO FOR


- L o n g C r e s t e d H a v e s - H e a d S e a s Conditlon -Modal Period=9 Seconds -Slg. Have Helght=12.14 Ft

=3.7

FFG-7 WITH STERN WEDGE AND NO BULB

Q F F G - 7 WITH STERN WEDGE AND BULB 1

[ ] FFG-? HITH STERN WEDGE RNB BULB 4

F F G - 7 H I T H STERN HEDGE RND BULB B

5 1'6 1'5 26 2~ ' 3 ~ SHIP SPEED (KTS)

Probabili~ofslammingatstation2versusshipspeed

1 . 1 4

1.12 E F F E C T I V E HORSEPOWER RRTIO

1.11~ 1 / EHP HITH BULB RND HEDGE x I \ ) ~. EHP HITH HEDGE ONLY

I .SB ~- Ct= METHOD: EHPm - ( C t s ) a * Sa

\ EHPb fCt= )b Sb 1.86 \

\

\ 1.04 I ~ \ /

\ / 1.82 ~ x, //

I- I . B O

.91: EN

. 9 4 - - B U L B 0 - - - - - B U L B 1

• 92 - - - - ' B U L B 4 ' = 8UL8 6

---- ~ULB 8

.BB

.o~ , i t , '

Flg. 11

1'. ' l ' . ' 2b ' 2~ 2t' ' 2'. SHIP SPEED (KTS)

Effective horsepower ratios from model test results

' 2'. ' y~


Table 3 Model particulars of FFG-7 configurations

Parameter Units No Bulb Bulb 0 Bulb 1 Bulb 4 Bulb 6 Bulb 8

Waterline length (m) 5.03 5.03 5.03 5.03 5.03 5.03 (ft) 16.5 16.5 16.5 16.5 16.5 16.5

Wetted surface (m 2) 2.693 2.774 2.778 2.793 2.808 2.756 (ft 2) 28.99 29.86 29.90 30.06 30.22 29.67

Displacement (N) 2137 2204 2206 2214 2217 2201 (lb) 490.5 495.5 495.9 497.8 498.4 494.7

LCG (+ fwd midship) (cm) -5.13 6.42 3.53 3.12 3.12 3.53 (in.) -2.02 2.53 1.39 1.23 1.23 1.39

Tank temperature (°C) 15.6 15.6 14.4 15.6 15.6 14.4 (°F) 60 60 58 60 60 58

NOTE: A correlation allowance of 0.00045 was used to expand all effective horsepower results to full scale. This correlation allowance was suggested for the FFG-7 hull form in reference [10].

Table 4 Model test sensors

Quantity Measured Transducer Description Tolerance

Drag block gage Hydronautics variable- linearity (% design reluctance modular force load) +0.25 gage (50-1b design load)

Pitch potentiometer 10-turn, 10 K-ohm tolerance +10% full- load; linearity +5%

Acceleration accelerometer Schaevitz 10 g linear servo linearity 0.05% full- accelerator (+ 10 g full load; repeatability load) 0.01% full load

Encountered sonic probe Wesmar LM4000 ultrasonic resolution (0.5% wave height pulsed sonar system (30-in. measured range);

measured range) (60-in. full linearity (0.5% full load) scale)

seakeeping performance resulting from the addition of a bulbous bow to the FFG-7 hull form did not seem sufficient to override the resistance advantages provided by the bulb.

Comparison of computer predict ions with model test results

As previously stated, one of the primary objectives of this project was to evaluate the usefulness of numerical hydrodynamics in bow bulb design. Therefore, the computer-predicted resistance and seakeeping characteristics of the FFG-7 configurations were compared directly with the results obtained from actual model testing.

Resistance The comparison of the resistance predictions of the XYZ

Free Surface Program and the results obtained from calm- water tank testing of the FFG-7 model can be made on two separate bases. First, in an absolute sense, the XYZ Free Surface Program predicted, by approximately 10 to 15 percent, lower resistances than the actual model tests did for all configurations at all speeds. This is evident on the plots of total ship resistance coefficient versus ship speed presented in Figs. 12 through 17. The multiplicity of points in the 18-20 knots speed range on the XYZFS curve results from the execu- tion of two separate algorithms within the program. One algorithm assumes a wet transom stern while the other assumes that the transom is completely dry. Since the model tests showed that the transom stern of the FFG-7 was dry at all speeds above 16 knots, the "wet-algorithm" points above that speed were ignored for the purpose of fairing those curves. The dashed line between 14 and 18 knots on the XYZFS curves represents the uncertainty as to when the transition from a wet transom to a dry stern occurs. The EHP ratio

curves shown in Fig. 5 were derived from these curves. As was the case with those EHP ratios, the total ship resistance coefficient curve from model test data for Bulb 0 represents refaired data as a result of a calibration error. A comparison of EHP ratios derived from XYZFS output and model test results for each bulb can be found in the graphs of Appendix 3. Although similar trends are evident from these curves, the quantitative results still differ somewhat. Many plausible explanations can be advanced for the absolute differences between XYZFS predictions and model test results. First, when the FFG-7 model was tested, a skeg was present. This skeg was not added to the hull form when it was panelized for the XYZFS Program. Thus, the drag of the skeg is not included in the computer predictions but does add to the overall resistance of the model. Second, there were small differences between the running draft at the forward and after perpendiculars which were used as input to the XYZ Free Surface Program and those observed during model testing. Other possible sources of the offset between XYZFS predictions and model test results include the questionable ability of numerical hydrodynamic codes to predict form drag and model wave-breaking resistance. This latter difficulty could become a factor at high speed. Any or all of these explanations, acting together or independently, could account for the absolute differences occurring between the ship resistance coefficient curves obtained from computer predictions and those derived by experimental testing.

Irrespective of the absolute quantitative differences between the XYZFS predictions and the model tests, a relative comparison of results as shown in Table 5 reveals identical rankings from both sources. These rankings were developed from the EHP ratio curves presented in Figs. 5 and 11 and are based on the 18-25 knot speed range. They are arranged from best to worst, with best being that bulb which had the

Bulbous Bow Design 39

p 3 4 - ! <

Wc

h W 0 U 3 . E

hi U Z E F- 3 .4

W ¢'v 3 . E

. J

f_. ( 0

I-- ;) . t i n H T e (/3

Fig. 12

i- t B

,~ ,'6 ,'8 ;)'e ;)'z SHIP SPEED (KTS)

Ship total resistance coefficient versus ship speed:

LEGEND= e HODEL TESTS B XYZFS:NET ~ O H & XYZFS:DRY TRFINSOH

FFG-7 wi th w e d g e and no bulb

5 .1 t

(T) 4. (

Cg

:¢

4.;

LL W 0 U 3 . ~

W U Z QZ I-- 3 .4 U) H U3 W rY

3 .e J E: F- O F" 2 ~1

[1. H I O3

/ Note: Model Test Points /

represent refaired.data. ~ ~ / A

'~ ~ E] ® MODEL TESTS / / I:1 M XYZF'S = NET TRRNSOM

______.._._._ffi ~ In & XYZFS= DRY TRRNSOM

A 1'6 ,'o 2'. 2'2 S H I P S P E E D ( K T S )

Sh ip total res is tance coe f f i c i en t versus ship speed:

2 - ~ 2

Fig. 13

2~ 2~ ' 2~ 3`"

FFG-7 w i th w e d g e and Bulb 0

4 0 B u l b o u s B o w D e s i g n

2.

<

v 4.2

b_ W O U 3 . 8

W O Z (E i - - 3 . 4 03 i-,4 03 W n."

3.~ J ( I I-- O I--- ~.. Q. i-i "r" 03

~ _ - - - - - - < F ~ LEGEND: / ® MODEL TESTS

/ B [] X Y Z F S : N E T TRRNSOH 7 / J [] A X Y Z F S : D R Y TRRNSOM []

SHIP SPEED (KTS)

5.12

Fig. 14 Ship total res is tance coe f f i c ien t versus ship speed: FFG-7 with wedge and Bulb 1

5 . 0

p 3 4 . 6 <

v 4 . 2

LL W O (..) 3.1B

W f j Z I1- I.- 3. U3 I-.4 U3 W IZ

3. _J CE I-- O I--

2 . ! O.. i.-.4 1- 03

2

Fig. 15

S / /

Q MODEL TESTS j B B X Y Z F S : N E T TRRNSOH

~ . / [] A X Y Z F S : D R Y TRRNSOH

SHIP SPEED (KTS)

S h i p t o t a l r e s i s t a n c e c o e f f i c i e n t v e r s u s ship speed: FFG-7 with wedge and Bulb 4

B u l b o u s B o w Des ign 41

Fig. 16

5.0

® MODEL TESTS

I I i l j [ ] []XYZFS:WET TRRNSOM

I [] a XYZFS:DRY TRRNSOM

S H I P S P E E D ( K T S )

Shiptotalresistance coefficientversus ship speed: FFG-7 with wedge and Bulb 6

5,~

( -04 .6 < Eg

4 .2

t2 t.u W O U 3 .

W U Z QS ~ 3 . , U')

ffl W n~

3 . e ..J 122 I--- O

2.( n M T U'} ~

Fig. 17

cq4.G < Q

x< 4.2

u2 h W O r..) 3 . e

W U Z

I--3.4 O3

W n~

3 ~

F ' - 2 . 6

T O3

2.

J B I

~ LEGEND : j. B ® MODEL TESTS

/" {3 / ~'1XYZFS: WET TRRNSOM

/ 1 3 B & XYZFS: DRY TRRNSOM

,'. ,'6 ,'B 2'~ 2'2 2~ z'~ 2~ 3'B SHIP SPEED ( K T S )

Ship total resistance coefficient versus ship speed: FFG-7 with wedge and Bulb 8


lowest EHP ratio over the speed range considered. These identical relative rankings indicate that the XYZFS Program can be used with confidence to select the best alternative from among competing bulb forms if minimizing a ship's total resistance is of primary concern.

From this relative ranking of bulb forms, an important observation can be made about the effects of bulb geometry on performance. In general, the resistance advantages derived from adding a bulbous bow to the FFG-7 hull form seemed to increase with increasing bulb volume. One note- worthy exception to this rule is the superior performance of the relatively small Maestrale 0-type bulb, which has a lower effective horsepower ratio than Bulb 1 over the majority of the relevant speed range, despite having less volume. With this exception, larger bulb size seemed to enhance a bulb's resistance reducing effect. Bulb 6 (the longest and broadest of all bulbs tested) illustrated this point by having the lowest EHP ratio over the 18-25 knot speed range. This observation seems to indicate that bow bulbs for fine-form, high-speed ships should be made as large as possible within the practical constraints associated with ship handling.

Seakeeping In an attempt to avoid the many runs required to obtain

statistically significant results from irregular wave tests, a "single pass" method was used. This periodic irregular encountered wave technique [11] should have produced a relatively smooth transfer function after a single tank run. How- ever, this method inadequately accounts for wave-wave inter- actions which appear to cause frequency shifts sufficient to make the encountered wave nonperiodic [12]. Nonperiodic encountered waves require long records to give statistically significant results. Even though several runs were made at each speed, there were not enough data produced to provide the required degrees of freedom in the analysis. Therefore, as previously stated, it was not possible to make a direct comparison between the seakeeping model tests and the results obtained from the Ship Motions Program.

Table 5 Relative bulb rankings

XYZFS Model Tests

Bulb 6 Bulb 6 Bulb 4 Bulb 4 Bulb 0 Bulb 0 Bulb 1 Bulb 1 Bulb 8 Bulb 8

project, future studies of bulbous bows for high-speed, fine- form vessels should consider the following:

1. Further optimization of the bulb form parameters using the XYZFS Program to rank competing designs.

2. The impact of practical shiphandling factors such as anchoring and dry docking on the bulb design parameters.

8. An investigation of the effect of variations on the 0-type bulb form.

4. Alternative methods for fairing the bulb into the hull.

Acknowledgments The success of this research project can be attributed direct-

ly to many individuals associated with several organizations. In particular, the authors gratefully acknowledge the support of Dr. David Moran, Dr. John Jayne and Mr. Ron Miller of the David W. Taylor Naval Ship Research and Development Center and Mr. Reilly Conrad of the Naval Sea Systems Command. Mr. Tom Price, a Naval Academy model maker, did an excellent job of building and installing the bulbs. A very special thanks is extended to John Hill, John Zseleczky, Louise Wallendorf, Steve Enzinger, Donald Bunker, Norm Tyson and Darlene Batten of the Naval Academy Hydrome- chanics Laboratory. Finally, but certainly not least, the authors thank Mary Palombo for her patience and skill in typing this manuscript.

Conclusions On tile basis of the work outlined in this paper, several

conclusions become evident concerning the application of bow bulbs to fine-form, high-speed ships. First, the XYZ Free Surface Program did provide an accurate relative resistance ranking of the bulbous bow configurations, allowing the combined numerical/experimental design methodology to successfully develop bulbous bows which improved the resistance characteristics of a high-speed ship. Second, the Kracht bulb design charts did yield an acceptable initial design. However, this design was not optimum, since increases in bulb breadth and volume tended to minimize the resistance characteristics of the ship. This trend indicates that bow bulbs should be made as large as the practical constraints associated with shiphandling will allow. Third, it is important to note that the existing Maestrale 0-type bulb possesses beneficial resistance characteristics as well as the practical advantages of short length and ease of manufacture. Finally, the bulbous bows did tend to qualitatively degrade the seakeeping performance of the hull form, but only to a small degree. This degradation did not seem sufficient to override the resistance reductions attained when the hull was configured with a bow bulb. These resistance reductions, while not substantial enough to warrant retrofitting existing ships of this type, do indicate that serious consideration should be given to installing a bow bulb on future ships.

As a follow-on to the research conducted during this

Bulbous

References 1 Salvesen, N., Von Kerczek, C. H., Scragg, C. A., Cressy, C. P.,

and Meinhold, M. J., "Hydro-Numeric Design of SWATH Ships," TRANS. SNAME, Vol. 98, 1985.

2 Van Oossanen, P., "The Development of the 12 Meter Class Yacht Australia II,'" Proceedings, Seventh Chesapeake Sailing Yacht Symposium, Annapolis, Md., Jan. 19, 1985.

8 Kracht, A. M., "Design of Bulbous Bows," TRANS. SNAME, Vol. 86, 1978.

4 Hagen, G. B. and Fung, S., "A Guide for Integrating Bow Bulb Selection and Design into the U.S. Navy's Surface Ship Hull Form Development Process," Naval Sea Systems Command Techni- cal Note No. 885-55W-TN0001, April 1983.

5 Kracht, A. M., "Weitre Untersuchungen Uber Die Andwen- dung Von Bugwulsten," VWS Bericht No. 811/78, 1978.

6 Cheng, B. H., Dean, J. S., and Jayne, J. L., "The XYZ Free Surface Program and its Application to Transom-Stern Ships with Bow Domes," David W. Taylor Naval Ship Research and Develop- ment Center, Bethesda, Md., 1984.

7 Meyers, W. G., Applebee, T. R., and Baitis, A. E., "User's Manual for the Standard Ship Motions Program (SMP)," DTNSRDC/SPD-0986-01, David W. Taylor Naval Ship Research and Development Center, Bethesda, Md., Sept. 1981.

8 McKee, J. M. and Kazden, R. J., "G-Prime B-Spline Manipu- lation Package Basic Mathematical Subroutines," DTNSRDC Report 77-0086, David W. Taylor Naval Ship Research and Development Center, Bethesda, Md., April 1977.

9 Zseleczky, J. and Johnson, B., "The Effects of a Bow Bulb and Various Stern Wedges on the EHP of FFG-7 Class Frigates," U.S. Naval Academy Hydromechanics Laboratory Report EW-8-84, Feb. 1984.

10 Woo, E. L., Karafiath, G., and Borda, G., "Ship Model Corre-

Bow Design 43

lation of Powering Performance on USS Oliver Hazard Perry, FFG-7 Class," Marine Technology, Vol. 20, No. 1, Jan. 1983.

11 Johnson, B., Anderson, C. H., Clark, A. J., and Lund, R., "'Single Pass Seakeeping Tests Under the Periodic Irregular Encoun- terea Wave Technique," Proceedings, 19th General Meeting of the

American Towing Tank Conf., Ann Arbor, Mich., Vol. 1, July 1980. 12 Johnson, B., Wallendorf, L., and Dalzell, J., "On the Genera-

tion of Complex Periodic Irregular Wave," Proceedings, 17th Inter- national Towing Tank Conference, Goteburg, Sweden, Vol. 2, Sept. 1984.

Appendix 1

Bulb parameter trend study

5 . 0 E

~" 4 . 0 ~ l..d I -

~_ 3 . 0 ~

m J m

2 . 0 0

i .00

0.00

HRGEN BULB DESIGN PROCESS F N = O . 2 8

"1~ C Lp R , 1 0 ^ 2

A

G w ~

.'s, . ~ .~8 .'~o

CRBL*IO

O czB

BLOCK COEFFICIENT

Fig. 18 Trends in bulb parameters with block coefficient

5 . 0 0

cz 4.00 Ld

W

n~

3.00

Ixl

2 . 0 ~

I . 0~

0 I 0~

HRGEN BULB DESIGN PROCESS FN=0.28

0 _ _ . ~I 0

FIC ~pR *10^3

l i .s4 .56 .~8 .~o

BLOCK COEFFICIENT

Fig. 19 Trends in bulb parameters with block coefficient

--a *10 CBB

® CRB T .10

.162


5 .E~E

b3 4 . 0 E

Ld F- W

n- r r

3 . 8 E m .3

m

2 .OE

1 .EIE

HRGEN BULB DESIGN PROCESS CB=O.56

[] D ~ czB*io

/

/ i ~ '-aCLPR. 1 O ̂2 / a

0 -

~L

- - 0 - ~ b ' ~ - - E) CRBL*1E]

e.% . ~ .'z8 .~e .3z /34 FROUBE NUMBER

Fig. 20 Trends in bulb parameters with Froude number

5 . 0 ~

u3 4 . 0 E

bJ i-- W ~Z QZ rY

n ~ 3 . 0 E

m

2 . ~ E

1 . ~E

HRGEN BULB DESIGN PROCESS CB=E~.56

[] . CVpR "10^9

A . . . . . -~ECB B b . . . . ~ . . . . "~ . . . . .18

(~ - - ~ - - 0 O - - 0 CRBT *IE]

0 . I~.~ .J26 .~2B .13E~ .132 .134 FROUDE NUMBER

Fig. 21 Trends in bulb parameters with Froude number


Appendix 2

Detailed outline and notes on bulb design methodology

Step (a)

Note (a)

Step (b)

Note (b)

Step (c)

Note (c)

Step (d)

Note (d)

Step (e)

Determine the block coefficient and the design Froude Note (e) number for the candidate bulbless ship. Select that set of design charts from [5] which is appropriate for the block coefficient. Since the block coefficient of the FFG-7 hull form is considerably smaller than the range of block coefficients found in [5], the smallest block coefficient was selected. A Step (f) Froude number of 0.30 (a ship speed of 20.4 knots) was chosen as the design Froude number for this project. This decision was based on the mission profile of the FFG-7 which reveals that 20 knots is the most common ship speed. In each of the six design charts, locate the points on the appropriate Froude number curve where the maxima oc- Note (f) cur. Record the values of ACpva associated with each of the maxima. It was assumed that the authors of [4] intended that only the largest maxima from each of the design charts be Step (g) recorded. This assumption was necessary in order to un- derstand the intention of Step (c). Without this assumption, one might select a suboptimal design value of ACpva. Select the smallest of the recorded values of ACpva as the Note (g) design value for the bulb. From each of the six charts determine the value of the bulb parameter which corresponds to that value of ACpvR. Several guidelines are presented at this point to aid in parameter value selection since several values for a given bulb parameter will proba- bly exist for a single value of ACpvB. The ACpvR chosen as the design value for the FFG-7 was 0.29, and the guidelines for resolving ambiguities arising from the undulations of the curves were closely followed. Step (h) Determine the height of the bulb at the forward perpendicular. Both the selected height and the distance trom the bottom of the bulb to the baseline are matters of judgment. However, the height of the bulb is constrained by at least two requirements: (1) It must be large enough to enable the required cross-sectional area (ABT) to be developed, and (2) the top of the bulb must be an appropriate distance below the design waterline. A tentative value for bulb height can be obtained using the following formula:

Note (h) HB ffi (4ABT)/(rBB)

Step (i) By using this formula, the full-scale height of the candi- Note (i) date bulb was determined to be 2.88 m (9.27 ft). An arbitrary value of 0.17 m (0.56 ft) was assigned as the distance from the baseline to the bottom of the bulb giving an effective bulb height of 2.83 m - 0.17 m = 2.65 m (8.71 Step(j) ft). These values were held constant for all bulbs produced for this study. The effective bulb height was used in all further calculations required by this design methodology. La~, out the upper portion of the longitudinal profile of the bulb by joining the point at the forward perpendicular (at height Ha above the bottom of the bulb) to the point at the Note (j) nose (at height ZB above the baseline) with an arbitrary concave curve having the ~eneral shape of an ellipse or parabola with its vertex at the nose.

Quite obviously the discretion of the designer impacts heavily on the choice of the "arbitrary curve" mentioned above. Thus, it is readily apparent that two designers utilizing exactly the same bulb parameters could develop completely different bulb forms.

Lay out the lower curve of the longitudinal profile by computing distances y(x) below the upper curve at longitudinal distances, x, forward of the forward perpendicular according to the following formula:

y(x) = [ H~ - x2( HB/ LpR)2] 0'~

This step is relatively straightforward and easy to imple- ment; however, no mention is made in [4] as to how this formula was derived.

Integrate to determine the area ABL for comparison with the design chart. Make minor adjustments to the longitudinal profile to obtain approximate agreement with one of the values selected from the design chart.

The value of CABL developed by this method was approximately twice that selected as the near-optimum value from the design chart. Presumably, this occurred because of the appreciable downward extrapolation from the range of block coefficients in the design curves. It was necessary to disregard the value of CABL taken from the design charts in order to attain the correct value of other bulb parameters.

Compute the approximate transverse areas of the bulb at selected longitudinal stations by means of the following equation:

AT(x) ffi y2(x)A'BT/ n~

where A'BT is the actual designed transverse area of the bulb at the forward perpendicular. In general, A'BT is likely to be equal to, or very nearly the same as, the design- chart value.

Again, no mention is made in [4] as to how this formula was derived. Integrate to determine the bulb's volume. A straightforward application of Simpsen's rule to the transverse areas computed in Step (h) was utilized to com- plete this step.

Starting with these approximate parameter values, develop a faired bulb form. Compute the values of its geometric parameters and compare them with those selected from the design charts. Although it is not expected that exact agreement will be achieved, iterations on the design can be made in order to bring the actual values into better agreement with the design chart selections.

The discretion of the individual designer again plays a major role in the implementation ot this step. After numerous iterations the final bulb form produced by this design methodology was designated as Bulb No. 1.


Appendix 3

EHP ratio comparison from XYZFS predict ions and model test results

1.14

1.12

1.18

I . 88

I . 86

I . 84

I. 82

0 H ;- 1.88 O~

o._ .98 T Ld

.96

.94

.92

. 88

.8

EHP RRTIO FOR BULB NO. 0

[/\EH~-H;~-HHPN~-UGL-E-BRNNOD'-N~-BULB'~ x']

/ --.. ~ ~ ~" ~ ~ \ \ f ~ -~ XYZFS

I~ l~ 16 2~ 2~ 2~ 2~ 26 3~ SHIP SPEED (KTS)

Fig. 22 EHP rat io for Bulb 0

1.14

1.12 EHP RRTIO FOR BULB NO. I

I . IE ~ . EHP BULB NO. 1 _~

\E / 1.8E HP HITH HEDGE RND NO BULB

I . 8E

1.84

I. 8~

o H

I.BE *.. i . nf ~ . j j _

,.~l"n .92"94"96"98 ~ --- . . . .

. 8 8

8 ,'4 I'G ,'8 26 2'~ 2'4 SHIP SPEED (KTS)

×YZFS

MODEL TEST

Fig. 23 EHP ratio for Bulb 1

2~ a~ 36


14

12

1E

88

8 6

84

B2

8E

98

. 9 6

.94

.92

,9E

.88

.e~ f4


HP HITH HEDGE RND NO BULB//

.I 7 " / \ \ XYZFS

MODEL TEST

SHIP SPEED (KTS)


1.14

. 1 2

. I E

• 8 ~

• 0 6

• 04

• 82

n,,

96

94


~E EHP BULB NO. 6

HP H I T H HEDGE RND NO BULB

j /

92

92

8B

\ XYZFS

11?- - -

MODEL TEST

S H I P SPEED ( K T S )

Fig. 25 EHP.ratio for Bulb 6


1.14

1.12

I . 08

I . 8 6

1 . 8 4

I . 8 2 o H I-- 1 . 0 ~ n" n, a. . 9 8 "I- Ld

. 9 6

. 9 4

. 9 2

. 9 0

. 8 8

.o%


HP HITH HEDGE RND NO BULB/) MODEL TEST

XYZFS

SHIP SPEED (KTS)


2'. .b

Discussion

William T. Lindenmuth, Member

[The views expressed herein are the opinions of the discusser and not necessarily those of the Department of Defense or the Department of the Navy.]

I commend the authors for this clear and well-documented study. They have shown that the XYZFS program can serve as a useful design tool regarding the relative effect of various bulb designs on ship resistance. I found a similar utility for XYZFS, in a study of streamlines along the forebody of a ship with and without bulb, to help assess the bulb's effect on the so-called bubble sweepdown problem. Here too, comparison with observations from scale-model tests indicate that the numerical predictions are not particularly accurate on an absolute scale; but they did provide nominal relative trends for the cases that were studied.

The authors' may have overstated the range (18-25 knots) for which the rankings are identical in Table 5. It appears to me that they are identical only from 22 to 25 knots.

Since the XYZFS code is actually predicting wavemaking resistance, correlation with model test data derived from "wave cuts" may prove enlightening, particularly with regard to high-speed wave-breaking resistance phenomena; wave cuts may "miss" this component of wave drag since it has been transformed to kinetic energy within the control area defined by the cut. XYZFS may predict wave spectral energy sufficiently, but I would not expect phase relationships to be accurate enough to re-create a longitudinal wave cut.

Recent wave cut experiments at DTNSRDC have shown a significant increase in the amplitude of the transverse waves when the model is propelled compared with when it is towed as a drag body. I suspect that this added wave drag is ordinarily accounted for in the thrust deduction factor. Can the authors comment on the feasibility of including propeller

effects in the XYZFS program, say with some sort of simple actuator disk?

The authors suggest that the degradation in seekeeping performance is not sufficient to override the desirability of a bulb for reduced resistance. Would increased ship motions, when operating in a seaway, create an incremental added resistance sufficient to negate the modest calm-water benefit?

It is clear that numerical hydrodynamics hasn't obsolesced the towing tank. Yet work, such as presented by these authors, advances the case for numerical design tools with great potential for naval architects.

Gabor Karaflath, Member


The basic design method proposed by the authors is to use the Kracht parametric bulb design method for an initial design, Bulb 1, and then to refine the design according to guidance from alternative systematic design variations evaluated by the XYZFS computer program. The accuracy of the XYZFS computer program to evaluate changes in bulb shape and size on performance is crucial to the success of the proposed design method.

In Table 5 the authors show that for the 18-25 knot speed range the relative bulb performance ranking according to XYZFS predictions is identical to the ranking based on model test predictions for all five bulbs tested. Although there is a general similarity in the ranking, the ranking is not identical since at 20 knots ship speed the ranking of the model test results as obtained from Fig. 11 differs significantly from the ranking according to XYZFS predictions. Nevertheless for the 18-25 knot speed range the XYZFS computer program


td n~

..J

Z 0

W

~0 0 n~ (_)

~n

I n

w

H k -

._1 W

~ J m /

° A - J ~ /2

- t l . 0 8 , - . e . 0 0

/

C O N T O U R S O F P E ( B U L B ) / P E ( S T E H )

•

d _ i111 - - ~

RELATZVE BULB LOCATTON O = x / L

Ship with Moderate Kracht Bulb "A" in Present Location

Fig. 27

. < t D W

.-I

Z 0 H

tM t - U t.,a (t)

- - O e ~

• ~

h i

H t "

- - I

.d 8 . 8 0

predicted the performance of Bulbs 0, 4 and 6 relative to that of Bulb 1 with an accuracy that was adaquate for making design decisions. For other speeds and bulbs, however, the XYZFS predictions were significantly in error. For example, Bulb 8 was a much poorer performer throughout the speed range than the XYZFS predictions indicated and both Bulbs 4 and 6 had much better performance relative to Bulb 1 in the 26-27 knot speed range than predicted by the XYZFS computer program. This high speed is of great interest to FFG-7 powering. Can the authors comment on the reasons for these erroneous predictions?

The greatest XYZFS prediction error occurs in the prediction of bulbous bow performance relative to the no-bulb performance. Figure 5 shows the XYZFS predicted powering to be within 4-2 percent of the no-bulb performance whereas Fig. 11 shows predictions based on model test to be within 4-8 percent. Decisions to incorporate a bulb are usually based on desired improvements in powering of at least 4 to 6 percent. Hence a potential improvement in powering could be missed if only the analytical prediction is followed.

Laying aside the question of XYZFS prediction accuracy, the proposed design method is limited in that, after the initial bulb is designed according to the Kracht method, there is no substantive guidance provided to the designer for further optimization. In contrast, the traditional longitudinal wave cut bulb analysis method described by Sharma [13] (additional references follow some discussions) predicts the effect of bulb section area changes and bulb longitudinal changes on resistance. Figure 27 accompanying this discussion shows a typical contour graph from a longitudinal wave cut analysis which can be used to evaluate the design and provides design guidance for optimizing it. This longitudinal wave cut method has been used at the David W. Taylor Naval Ship Research and Development Center to successfully evaluate the effect of longitudinal sonar dome position on resistance and to evaluate the performance of numerous bulbous bows.

The disadvantage of the longitudinal wave cut method is that it is an experimental technique for which adaquate time needs to be scheduled during the ship design. In contrast, the

advantages of the analytical method with regard to quick evaluation of many bulb shapes cannot be ignored even if good judgment and experience are required to correctly use and interpret the analytical predictions. The authors are to be congratulated for correctly using the XYZFS predictions to design a bulb which at high speeds exceeds the performance of the initial Kracht bulb by 3 percent. Further research and development of analytical tools for the evaluation of hull form Changes on powering performance is highly encouraged.

Additional r e f e r e n c e

18 Sharma F. D., "'An Attempted Application of Wave Analysis Technique to Bow Wave Reduction," Sixth ONR Symposium on Naval Hydrodynamics, 1966.

John J. Slager, Member

It was my distinct pleasure to have been able to review, with Dr. Johnson and Mr. Hoyle, the bow bulb designs and the initial results of the model tests and the XYZFS computations. It was also my pleasure to have been involved with Mr. Hagen and Mr. Fung in the development of the bow bulb design .guide listed as reference [4] of the paper. Therefore, I am honored and pleased to be given the opportunity to comment on this paper.

The Introduction notes that bulbous bow design criteria for high-speed, fine-form ships such as destroyers and frigates are relatively unknown. I agree that such criteria are scarce, but wish to note that at least some bow bulbs for relatively fine, moderately high-speed hull forms have been designed and model tested and the results have been documented. For instance, there is information to be gleaned from "Some As- pects of Hydrodynamic Design of High Speed Merchant Ships" by Michelsen et al, 1968 SNAME TRANSACTIONS; also, although the bow bulbs discussed were not of the protruding type, there is, I believe, useful information relative to bow bulb design presented in "Ships with Bulbous Bows in Smooth Water and in Waves" by Dillon and Lewis, 1955 SNAME TRANSACTIONS. Further, and as you are well aware, large sonar domes have been installed on U.S. Navy


frigates for some time now. Admittedly, these are a special type of bow bulb, but they affect hydrodynamic performance in a manner which is similar to the manner in which large "normal" bow bulbs affect performance. Although we have no formal "design criteria" for such bow domes, we at least have developed reasonable preliminary-design type means for estimating the residuary resistance of a relatively large range of sizes of destroyer-type hull forms with relatively large sonar domes installed. I 'm sure the authors are aware of this information. My main point is that there is at least some information available which the practical hull/bulb designer can make use of.

Under "'Bow bulb design," it is noted that the method outlined in reference [4] "fails to provide a means for fairing the bulb into the rest of the hull." "Fails" is perhaps a bit strong, since the paper took note of the rough guidelines given by reference [4] in this regard and, in fact, adopted these guidelines for the bulb designs developed by the present authors! But allow me to be the first to agree with Messrs. Hoyle et a] that the design guide presented in reference [4] has limitations.

I would like to suggest that, in addition to the effective power ratios presented in Fig 5, plots or tables of the absolute differences in effective power (due to the bow bulb) would provide worthwhile information for naval architects. Some- times it is useful to know the magnitudes of the actual power differences; for instance, these data might be useful in making a quick assessment of the cost-effectiveness of installing a bow bulb. In this regard, the numerous economic analyses we have carried out have indicated that, although naval ships normally spend only a small proportion of their underway time at high speeds, the absolute power savings, due to a bow bulb, at the high speeds are sometimes sufficient to offset the power losses, due to the bulb, over much of the remaining speed range; these high-speed power savings may be sufficient to justify the use of a bulb. Thus, tabulations of absolute values of power savings and losses at the various speeds can be meaningful for the experienced hull/bulb designer.

I agree with the conclusion that serious consideration should be given to installing bow bulbs on future high-speed ships. At least a reasonable amount of consideration is, in fact, being given in feasibility and preliminary design studies carried out for new U.S. Navy ships. Appendix B of reference [4] presents procedures for calculating fuel savings (or losses) for a so-called "single-draft" ship (such as frigates and destroyers). These procedures take into account the following information:

- - the speed/power relationships for the ship in calm water, both with and without a bulb;

- - the total steaming time (usually expressed as hours per year);

- - the speed/t ime distribution for the ship as it is deployed; and

- - the specific fuel consumption (usually expressed as pounds per shaft horsepower per hour) over the range of operating speeds.

In addition, preliminary designs for the bow bulb structure have been developed in one or more instances, thereby en- abling bulb fabrication cost increments to be developed for use in cost-effectiveness studies.

In the Conclusions it is suggested that future studies of bow bulbs for high-speed ships should consider "The impact of practical shiphandling factors such as anchoring and dry docking on the bulb design parameters.'" I would like to point out that such impact studies are already a part of our design approach. Typically, these studies (in addition to assessing the powering and fuel-consumption impacts noted

Bulbous

above) assess the impact of the bow bulb on the following elements of design:

--structural, auxiliary systems and outfit weights; - - t r im and stability (of major importance for naval ships,

especially when a relatively large bulb is being considered),

- -anchor handling, and - -maneuver ing performance. Finally, I would like to emphasize that we still have a great

deal to learn about bow bulb design, especially bow bulbs for high-speed ships. For instance, it is noted in the paper that the performance of the Maestrale-type bulb did not seem to completely fit the emerging trends of the authors' research results. Similarly (and just as one might expect), in our development of the material for reference [4] and in the design and model testing of individual bow bulbs, we have found examples of bulb performance which did not seem to fit the basic trends we thought were emerging. I sincerely hope that further bow bulb research and model testing can be carried out in the near future. Clearly, we need a lot of drops of knowledge to fill the bow bulb design information bucket; to continue the metaphore, I am convinced that this paper is certainly a very significant drop in this bucketT

Alfred Kracht, e Visitor

I have respect for Professor Johnson's work, which was necessary to utilize the design charts mentioned in [5] for the bulbous bow design methodology presented. In this way the range of bulb parameters being considered can be restricted to a minimum in the ship design process, and model tests which are still required may be performed carefully and, therefore, economically.

It seems to me that the expression "optimization" in connection with the design charts is not quite correct. In the waveless theory an optimum bulbous bow reduces the wave resistance of a ship-bulb configuration at design speed to a minimum. This question cannot be answered by the method presented by the present authors because the design charts used in this method were derived from an analysis of routine test results of ships which were fitted with moderate, non- optimum bulbous bows. The reason for fitting moderate bulbs is due to the hydrodynamical interaction of hull and bulb and a suitable consideration of smooth-water performance and seakeeping qualities as well. For merchant ships which operate under different service conditions it is prudent to adopt the type and size of a bulbous bow to a wider range of speed and of draft alteration at the forward perpendicular, which results in a non-optimum smaller bulb at the design condition that would result from an optimization. Moreover, the bulb parameters should not be extrapolated as mentioned in my paper [8].

In comparison with merchant ships, the displacement of naval craft such as destroyers and frigates does not change considerably during operation. It is worth investigating the use of optimum bulbous bows of high-speed, fine-form ships. Since the design methodology does not provide the optimum bulb parameters, the wave resistance theory should be used in the optimization process as demonstrated by Takezawa [14] and Suzuki et al [15].

For an optimum bulbous bow, the greater the speed the greater the bulbous bow. Since at a Froude number of 0.85 the total optimum bulb volume amounts to 2 percent to 6 percent of the displacement volume--depending on block coefficient--the investigation of the seakeeping qualities becomes more important than the resistance reduction due to a bulb. In the case of an optimization, the cruising speed

8 Berlin Model Basin, Berlin, Germany.

Bow Design 51

should be chosen as the design speed and, in spite of the tempting high reduction of design power due to an optimum bulbous bow, the recommendations mentioned in Paper No. 3 of the 1985 Annual Meeting of SNAME [16] should be borne in mind also for naval craft. It is desirable to continue the seakeeping tests mentioned so that quantitative results are available and comparisons are possible.

One final question: Is the design methodology presented applicable to low-speed, full-form ships?

Additional references

14 Takezawa, S., "An Application of the Waveless Theory to the Design o{ a Destroyer Form," International Seminar on Theoretical Wave Resistance, The University of Michigan, Ann Arbor, Mich., 1963.

15 Suzuki, K., Higuchi, M., and Maruo, H., "Hull Form Design Based on Michell's Theory by Means of Nonlinear Optimization Technique," Bulletin of the Faculty of Engineering, Yokohama National University, Japan, Vol. 81, March 1982.

16 Blume, P. and Kracht, A. M., "Prediction of the Behavior and Propulsive Performance of Ships with Bulbous Bow in Waves," TRANS. SNAME, Vol. 98, 1985.

Michael B. Wilson, Member


This is an interesting paper, expecially in that it is aimed at illustrating how analytical results can be used to motivate and then augment the impact of a well-focused experimental program.

It is fascinating to see that a bow bulb can successfully reduce the resistance of a slender, combatant hull form, and that this can be accomplished at Froude numbers of practical interest. It is a bit disco-raging to note the large size of bulb

(big forward protrusion) that it takes to get the desired range. of total resistance improvement at the higher Froude numbers of 0.35 and up. Evidently long bulbs with center of volume well forward of the FP are required to produce meaningful destructive interference between the bulb and the ship hull wave systems, at least with the Nabla bulb configuration. As mentioned by the authors, it would be very interesting to determine the systematic performance of bulbs with O-type cross sections, compared with the results with the Nabla shapes. It is noted that the short O-type bulb had its best performance, and the best EHP reduction overall, at F , = 0.25 (V s ~ 17 knots). By analogy to the trends for Bulbs 4 and 6, it could be speculated that a longer O-type bulb would be better at a higher Froude number, possibly noticeably better than the Nable type. It should also be noted that the lower Froude number range tradeoff between the various bulbs is not correctly displayed by the calculated results.

I have cross-plotted the changes in EHP ratio against all the various bulb parameters. I have noted that the trends for the EHP ratio versus the protrusion length ratio CLPR and volume ratio CvpR show improvement at the largest values CLPR = 0.04 and Cven = 0.0047 (both curves slope downward). Do the authors conclude that an "'optimum" bulb may have an even larger value of Lpn?

A note of caution: Some care must be exercised if anything but the relative rankings of alternative hull shapes are to be chosen from the XYZFS results. This is because the variation of wave resistance is not uniformly accurate with respect to Froude number as predicted by the XYZFS analysis. A sam- ple comparison for a combatant hull form similar to FFG-7 is shown in Fig. 28 herewith for the predicted wave resistance coefficient Cw from XYZFS and experimental values of the wave pattern resistance CwIe) obtained from longitudinal wave cut measurements. This pertains to the realistic case of

i

3.O I

2.5

PanelM:d:d H~II~ S1 //

~ ~ 2.o i / C c I i ~ I U R E D

C A L t.~ 1.5

~ 1.o

Fig. 28

i , , ~ ' FRE j E TO I SINK AND TRIM

I -- 1 I I I I 01 02 03 04 05 05 0 7

FROUDE NUMBER, F n

Comparison between calculated wave resistance Cw and measured wave pattern resistance C~a for a high-speed displace-

ment hull, free to sink and trim

I I I I I TRIM

Hull S1 M o d e l 5 4 1 6 F I X E D

2 s

~ f z

~ 2o

~-) 1.s

~ CALCULATED Cw

~ 1.0

M E A S U R E D MEASwfURED

0.5

[ I K I I 0.1 0 2 0.3 0 4 0 5 0.e

FROUDE NUMBER, F n

Fig. 29 Comparison between calculated wave resistance and measured wave pattern resistance for the high-speed hull form, with fixed

trim


the hull free to sink and trim. One point of this comparison is that the region of good agreement is quite narrow, in the Froude number range of about Fn = 0.85 to 0.45. At lower Froude numbers, the discrepancy becomes larger. This vari- ability in accuracy could affect the reliability of even compar- ative trends in the low Froude number range.

It is noted that the XYZFS calculations for all the bulb shapes of this study were carried out with the same specified sinkage and trim which were measured or estimated for the FFG-7 hull fitted with the O-type bulb (Bulb No. 0). The sinkage and trim can be important to the accuracy of the calculations from XYZFS, and therefore to the interpretation of the results. Here is an example to show the order of magnitude of the influence. For the same hull form as in Fig. 28, Fig. 29 shows results of a comparison between the calculated wave resistance using XYZFS and the measured wave pattern resistance, at fixed trim (hull fixed at zero speed sinkage and trim). The agreement between the two is start- lingly good in the range F n = 0.85 to 0.5, but still relatively poor at lower Froude numbers. In the transition from Fig. 28 to Fig. 29, the trim was estimated from the hydrodynamically induced moment determined from a first-round XYZFS cal- culation with the panelling arranged with the waterline ini- tially at the zero speed location. The quantitive difference in the two comparisons shows how the trimmed hull orientation affects the accuracy of the final results for wave resistance.

Donald McCallum, Member


My congratulations to the authors for a fine piece of computational/experimental naval architecture. Some strong trends may be gleaned from the paper, aiding in the design of bulbous bows, but the paper also poses some questions, and highlights the need for further research in this area. NAV- SEA has conducted many experiments involving bulbous bows, particularly during auxiliary ship designs, and with good successes for the most part. The AO-177, AOE-6, LHD- 1, and CV-68 (Hull Expansion) are recent examples. We have been most encouraged by the work of Dr. Kracht of Germany [8,16[ (see Kracht discussion), and now this fine paper.

The use of the 15-deg stern wedge for the FFG-7 predictions and model tests is to be commended. This indeed is a feature which should be retrofitted into the FFG-7 ship design. It costs little to build and install, yet gains benefits in reduced fuel consumption and slightly increased top speed.

Using the XYZ Free Surface Program is vindicated as a "feasibility-level" design tool; however the comparison between computer predictions and model tests results is disap- pointing. Some of this difference can, I believe, be explained by the fact that the XYZFS program does not adequately address the viscous effects associated with large bulbs.

Other questions are as follows: • If Bulb 6 gave the best predicted power savings, why was

it not investigated for seakeeping? ° How did the model test results compare against the

Kracht (extrapolated) predictions for EHP savings? This would give a good check of the extrapolation of the Kracht data for high-speed ships.

• Figure 11 shows that Bulbs 0 and 4 are giving very good results in the 14-18 knot range. What would be required to "tweak" these designs to be optimized for 20 knots, which is more of an operational speed for the FFG-7 class?

• Was the FFG-7 model equipped with its keel-mounted sonar dome during the tests? If so, this could have played an

important part in pressure field cancelation. Also, a photo- graph of the FFG-7 model would have enhanced the paper, in addition to clarifying some of these points.

The paper helps to highlight some of the problems inherent in the tools used by ship designers. The Ship Motions Pro- gram (SMP) has limited capability to predict the pressure field created by the forward motion of the bulbous bow. The seakeeping model test data, as correctly interpreted by the authors, show that bulbous bows tend to produce more deck wetness, since a large bulb will tend to "dig-in'" to a wave due to the above-mentioned pressure field. This phenonenon was recently highlighted by full-scale experience on ships of the AO-177 class, which have large elliptical bulbs. Deck wetness problems were encountered, which had not been predicted by the earlier SMP-type runs; these problems necessitated an increase in scantlings around the fo'c'sle deck area. Dr. Kracht's recent paper [16] on the seakeeping considerations of bulbous how design states very clearly that bigger is not always better! A careful design methodology, utilizing XYZFS predictions, SMP predictions, resistance a n d seakeeping tests, as this paper emphasizes, is certainly the prudent course of action.

Jeffrey J. Hough, Member


As one who is involved in both the development and utilization of computer tools and hull form design and hydrodynamic performance, I am very pleased to see the work and results of the authors' efforts to utilize an advanced hydronumeric analysis tool. At NAVSEA we have in the past used the XYZFS computer tool for some work, including the evaluation of stern wedges for a recent combatant design and selection of the candidate wedges to be tested. In this work we found, much as the authors describe, that the relative ranking of the various candidate stern wedges was the same as predicted by XYZFS and model test data.

The method of utilizing XYZFS to rank candidate hull form modifications or appendage designs is a technique whose time has come and is sorely needed. With the com- pressed design schedules and limited time and funds for model testing, the use of advanced hydronumerie computer tools gives the hull form designer the ability to incorporate innova- tion into the hull at an early stage of design and not as an add- on afterthought. As part of the NAVSEA Hull Form Design System (HFDS) we plan to incorporate more hydronumeric analysis tools by building preprocessor programs to these computer codes that will use the surface geometry models which currently we are utilizing for hull form definition.

Now some specific comments on the paper. Even though the relative ranking of the various bulbous bows was correct, Figs. 28 through 26 in Appendix 8 show that the absolute value changes in EHP were quite different and the shapes of the curves were not even that close when compared with the model test data. This shows that the XYZFS program still has limited value in predicting absolute values for resistance. Would the authors please comment on what they feel is causing this limitation on the results and what plans are in the works to resolve them?

I thank the authors and the Society for giving me the opportunity to comment on this fine paper.

Authors' Closure

The authors wish to thank the discussers for their interest in this paper and for their thought-provoking comments.


SIDE VIEW

TOP VIEW

Fig. 30

BODY PLAN

W Inui S-201 hull form

In response to Mr. Karafiath, the ranking of the bulbs was based on their overall performance over a majority of the 18- 25 knot speed range. While the results for Bulb 8 were significantly different, the authors feel that the Bulb 8 model test results--not the XYZFS predictions--are at fault. In fact, the confidence we gained in the XYZFS predictions allowed us to forego the retesting of Bulb 8 because its inferior resistance characteristics had already been identified by the computer program.

Mr. Karafiath expressed concern that a potential improvement in powering might be missed if only the analytical prediction is used. We concur with him and reiterate our position that model testing is mandatory for those bulbs which performed best during the analytical computations. This tank testing would enable bulb designers to convert qualita- tive ratings into quantifiable results.

Finally, the authors wish to thank Mr. Karafiath for pointing out the most significant advantage of our bulb design methodology over the longitudinal wave cut approach--t ime savings!

The authors would like to thank Dr. Wilson for his interest in our paper. As far as we can recall, Mr. Cheng and his colleague, Mr. Miller, performed the XYZFS computations for Model 5416 about two years ago. At that time there were no experimental results for the hull. It is interesting to see the comparison between the XYZFS results and wave cut experiments which were conducted this year. We find the comparison encouraging. As we mention in our paper, we are more interested in the relative ranking than the absolute value of wave resistance. The results presented in Dr. Wilson's Fig. 29 indicate that a quantitative comparison is possible for this particular hull. Notice that the absolute value of the computed wave resistance is in excellent agreement with wave pattern resistance at higher Froude numbers (from 0.:35 to 0.5). This Froude number range corresponds to the higher operational speeds of high-speed ships. The fact that the agreement is so good at high operational speeds is very satisfying. Admittedly, the agreement is not so good at lower speeds. The most likely explanation is that the program's assumption of transom flow breakaway has been violated.

The authors are well aware of the importance of sinkage- and-trim predictions. Cheng and Dean [17] developed and implemented a new algorithm to compute the sinkage and trim. This algorithm improved the results of recent calculations, including those for the Model 5416 study. Further- more, we have found that using the experimental sinkage and trim of a similar hull can result in further improvement in wave resistance predictions. The computations reported in the present paper were performed using the experimental sinkage and trim obtained when the Maestrale O-type bulb was appended to the FFG-7 hull.

Mr. Lindenmuth asked two questions about the XYZFS program. It is interesting that both questions are related to current projects at DTNSRDC. For one of these projects, Mr. Cheng and his colleagues are calculating longitudinal wave cuts for comparison with the corresponding experimental measurements. The hull form that has been used as a test case is an Inui S-201. This is an idealized mathematical hull form modeled by a centerplane distribution of linear sources. Figure 30 herewith shows the side view, top view, and body plan of this hull form. Figure :31 presents a comparison of the longitudinal wave cut obtained from the XYZFS computations with experimental results published by Sharma for a Froude number of 0.319 [13]. In this experiment the wave probes were located away from the ship's centerplane at a y- value corresponding to 2y/Lop = 0.826. In Fig. 31 the di- mensionless wave elevation (100 g/Lee) is plotted against the longitudinal distance (2x/Lpp). The solid line represents the experimental wave cut, while the dashed line represents the XYZFS computations of the longitudinal wave cut. The wave length or the distance between consecutive peaks is correctly predicted. The phase relation seems in good agreement between computational and experimental results. However, because a linearized free-surface boundary condition was used in the XYZFS program, the wave amplitudes tend to be underpredicted. Figure 32 presents a similar comparison of longitudinal wave cuts for the same hull at a Froude number of 0.255. Again, the wave length is correctly predicted while the wave amplitude seems to be underpredicted. However, there is an unexplained phase shift between the XYZFS results and the experimental measurements for this Froude number. The cause of this discrepancy is being investigated.

A second project at DTNSRDC is tasked with addressing the question of adding propeller effects to ship wave problems. Mr. Cheng has just begun to work on this new task, which involves the addition of an actuator disk capability to XYZFS. The effectiveness of such a simplified propeller model should be better understood by next year.

The authors wish to thank Mr. Slager for bringing to the attention of the naval architecture community several important publications dealing with the design of bow bulbs.

In retrospect, the author's agree with Mr. Slager that "fails" is a bit strong. Nevertheless, we definitely feel that more research is needed in the area of bulb-to-hull fairing in order to provide guidance to the naval architect.

In response to Mr. Slager's request for information on the absolute differences in effective power due to bow bulb variations, the authors have developed Table 6.

The authors apologize to Dr. Kracht for their "non-optimum" use of the word "optimum." Perhaps a better way to describe our methodology is to simply define it as an attempt to enhance the ship's resistance performance by systematical- ly varying the bulb's parameters.

Based on our results, bow bulbs should be made as volumi- nous as possible. However, more research is needed in order to define the limits associated with this "bigger is better" philosophy. For example, recent research has shown that very large bulbous bows can have a detrimental effect on seakeeping [16]. Consequently, we encourage the design community to develop methods which will help to evaluate the seakeeping characteristics of various bulb forms.

The authors feel that the methodology presented in this paper should, in theory at least, be valid for low-speed, full- form ships. However, care must be taken that the flow code utilized for the resistance predictions is suitable for the hull form in question. To date, XYZFS has been used primarily for high-speed, fine-form ships of the destroyer/frigate vari-


1 75

I ~ l,',,i l i t 7! / \

:ls ~ I ' : I, ~o ~ Ii , I l l ' ~> 1 i .i l'

t I \ ! " ; I ~ ' Ii<,, ' \1 - ' 1

r :d

k . . f

- - SHARMA'S EXPERIMEN1

. . . . XYZFS

- 1 . 75

-~.5 # # # ,.5 STERN MIDSHIP BOW x= -1.0 x=0 x = l 0

2x LONGITUDINAL DISTANCE - -

Lpp

Fig. 31 Comparison of longitudinal wave cuts from XYZFS program and from experiments for Froude number 0.319

ety. Its applicability to other types of vessels is still being investigated.

We strongly agree with Mr. Hough and others that the time is ripe for naval architects to use a combined numerical and experimental approach for making resistance predictions. The stern wedge evaluation mentioned by Mr. Hough and

described by Cheng et al [18] at CADMO-86 is another example of this bilateral approach.

The absolute differences in the ship's total resistance coefficients between the XYZFS predictions and model tests were discussed in the present paper. Several explanations for these differences were advanced, including variations in hull geom-

1.75

- - SHARMA 'S EXPERIMENG XYZFS

8 " I I x

:1~ P I L I.~, \" I l l X\

F--. ' i

; -/ "" t l " " J

/ 7 "

J ~'. I •

Fig. 32

- 1 7

-,.5 f f f ,.5 STERN MIDSHIP BOW x= -1.0 x=0 x = l . 0

2x LONGITUDINAL DISTANCE - -

Lpp

Comparison of longitudinal wave cuts from XYZFS program and from experiments for Froude number 0.255


Table 6 FFG-7 EHP powering differences a

EHP stern EHP sternq wedge - wedge |

and bulb only J

Bulb Speed, knots No. 14 16 18 20 22 24 26 28

0 -69 -156 -134 -197 -174 -145 -198 -221 1 -51 -154 -181 -206 -174 -73 -83 -59 4 -9 -105 -176 -260 -240 -232 -312 -384 6 -3 -99 -176 -231 -300 -312 -406 -443

a EHP values derived by expanding modeltestresults.

etry (skeg) and discrepancies in the running sinkage and trim. Another factor is the questionable ability of numerical hydrodynamic codes to predict form drag and wave-breaking. The previously mentioned numerical wave cuts underpredict the observed wave amplitudes. Consequently, because of the linearized free-surface boundary condition incorporated in this method, the wave resistance will also be underpredicted. These difficulties represent a limitation that researchers in ship hydrodynamics must address in the future.

With regard to Mr. McCallum's comments, all of the bulbs that were built were tank-tested in irregular head seas. How- ever, only a few bulbs were investigated with the ship motions program since SMP was unable to clearly identify any difference in seakeeping performance between the various bow bulb configurations.

The experimentally derived EHP reduction at 20 knots for Bulb 1 (Kracht design) is about 5 percent. By comparison, the Kracht residual power reduction coefficient used in the design of Bulb 1 was approximately 3 percent. This would

seem to indicate that the actual performance of Bulb i ex- ceeded the Kracht design chart predictions. However, because the block coefficient was extrapolated beyond the range of these design charts, the authors feel that it is inappropriate to compare these two results.

"Tweaking" of a design is precisely what this methodology is designed to do. A designer could make small changes in bulb parameters, utilize XYZFS to determine the best of these design alternatives, and then model test the top performers in order to quantify their characteristics.

Neither the FFG-7 which was digitized for the XYZFS program nor the hull which was tank tested was appended with a keel-mounted sonar.

Based on the results presented in both this paper and by Blume and Kracht [16], it would appear that the bulbous bow design procedure for high-speed ships is evolving into a tradeoff between large-size bulbs with their favorable resistance characteristics and small bulbs, which tend to have better seakeeping qualities. In either case, there are several other important areas, such as those listed by Mr. Slager, which must be considered during the design of a bulbous bow. The authors certainly recognize that minimum resistance may not always be the most important objective for a particular application. Thus, as in nearly all other areas of hull design, the naval architect has to make some tough decisions in order to provide the best overall product to the customer.

Additional references 17 Cheng, B. H. and Dean, J. S., "The User's Manual for the XYZ

Free Surface Program," DTNSRDC Report 86/029, David W. Ta - lor Naval Ship Research and Development Center, Bethesda, M~., June 1986.

18 Cheng, B. H. Borda, G. G., Dean, J. S., and Fisher, S. C., "A Numerical/Experimental Technique for Wave Resistance Predic- tion" in Proceedings, International Conference on Computer Aided Design, Manufacture and Operation in the Marine and Offshore Industries, Washington, D.C., Sept. 1986.


Documents

Hoyle_J_W.Bulbous_Bow_Design_M.1986.TRANS.pdf