UDC 629.5.015.2:629.5.021.

12
N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC FORM FACTOR DETERMINATION OF THE FULL...

Nastia DEGIULI1
Neven HADŽIĆ1 Form Factor Determination
Marta PEDIŠIĆ BUČA2
Gordana SEMIJALAC2 of the Full, Large Breadth and
Shallow Draught Ship Series
Original scientific paper
By using the traditional approach, the total resistance can be broken down into Froude and
Reynolds number dependent components. These are then scaled according to their respective
scaling laws. The common practical method of separating the resistance components is by the
use of a form factor to estimate the total viscous component. This paper presents the form fac-
tor determination of the full ships with large breadth and shallow draught for full load and ballast
conditions using Prohaska's method. Form factors have been determined from the results of
resistance tests carried out in Brodarski institute with ship models of the Jelsa series. Since full
ships with block coefficients CB > 0.8 are concerned, higher powers of the Froude number, n =
4 ÷ 6, have been used in the expression for the wave resistance. Obtained values for the total
resistance have been compared with measured ones. It has been shown that all three powers
give satisfactory agreement of results.
Keywords: form factor, Prohaska's method, total resistance, viscous resistance, wave resis-
tance.

Određivanje faktora forme serije punih brodova velike širine i malog gaza
Authors’ addresses:
1
University of Zagreb, Faculty of Izvorni znanstveni rad
Mechanical Engineering and Naval Primjenom tradicionalnog pristupa ukupni otpor se može razložiti na komponente ovisne o
Architecture, Ivana Lučića 5, 10000 Froudeovom i Reynoldsovom broju, koje se prenose s modela na brod prema odgovarajućim
Zagreb, Croatia zakonima sličnosti. Uobičajena praktična metoda razdvajanja komponenti otpora je primjenom
2
Brodarski institute, Avenija faktora forme kako bi se procijenila komponenta ukupnog viskoznog otpora. U radu je prikazano
Većeslava Holjevca 20, Zagreb, određivanje faktora forme punih brodova velike širine i malog gaza za puno opterećenje i balast
Croatia metodom Prohaske. Faktori forme određeni su na temelju rezultata modelskih ispitivanja serije
Jelsa provedenih u Brodarskom institutu. Budući da se radi o punim brodovima s koeficijentima
Received (Primljeno): 2007-05-23 punoće CB > 0,8, upotrebljene su više potencije Froudeovog broja, n = 4 ÷ 6, u izrazu za otpor
Accepted (Prihvaćeno): 2007-07-27 valova. Dobivene vrijednosti ukupnog otpora uspoređene su sa izmjerenim vrijednostima i poka-
Open for discussion (Otvoreno za zano je da sve tri potencije daju zadovoljavajuće slaganje rezultata.
raspravu): 2008-12-24 Ključne riječi: faktor forme, otpor valova, Prohaskina metoda, ukupni otpor, viskozni otpor.

Nomenclature RT total resistance, N / ukupni otpor

RV viscous resistance, N / viskozni otpor
a coefficient / koeficijent RW wave resistance, N / otpor valova
BWL breadth on waterline, m / širina na vodnoj liniji Rn Reynolds number / Reynoldsov broj
CF frictional resistance coefficient / koeficijent otpora T draught mean, m / srednji gaz broda
trenja S wetted surface of bare hull, m / oplakana površina golog
CT total resistance coefficient / koeficijent ukupnog otpora trupa
CV viscous resistance coefficient / koeficijent viskoznog xCB longitudinal distance of centre of buoyancy abaft ami-
otpora dship, m / uzdužni položaj težišta istisnine od sredine
CW wave resistance coefficient / koeficijent otpora valova broda
Fn Froude number / Froudeov broj — displacement volume, m3 / volumen istisnine
g acceleration of gravity, m/s2 / gravitacija v ship speed, m/s / brzina broda
k form factor / faktor forme r water density, kg/m3 / gustoća vode
LPP length between perpendiculars, m / duljina broda između CB block coefficient (on the basis of LPP) / koeficijent punoće
okomica (na temelju LPP)
LWL length on waterline, m / duljina broda na vodnoj liniji CBWL block coefficient / koeficijent punoće
n power of Froude number / potencija Froudeovog broja CWP load waterline coefficient / koeficijent punoće vodne
RF frictional resistance, N / otpor trenja linije

380 58(2007)4, 380-388

FORM FACTOR DETERMINATION OF THE FULL... N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC

CP prismatic coefficient / prizmatički koeficijent plate and the form factor accounts for the three-dimensional nature
CX maximum transverse section area coefficient / koeficijent of the ship hull. This includes the effect of the hull shape on the
punoće najvećeg poprečnog presjeka boundary layer growth and also on the viscous pressure resistance
BWL/T breadth on waterline-draught ratio / omjer širine na component. The ITTC-57 correlation line is an empirical fit and
vodnoj liniji i gaza some form effect is included. Thus it can be written as:
LWL/BWL length on waterline-breadth on waterline ratio / omjer
duljine na vodnoj liniji i širine na vodnoj liniji RT (Rn, Fn) = RV (Rn) + RW (Fn) = (1+k) RF (Rn) + RW (Fn) (1)
M length-displacement ratio / omjer duljine i istisnine
The form factor can be determined by different methods:
S wetted surface coefficient / koeficijent oplakane direct measurement of viscous and/or wave pattern resistance,
površine Prohaska’s method from slow speed data, Huges’ method based
on geosim data and full scale thrust measurements.
1 Introduction
2 Basic Form and the Jelsa Series Geometry
In recent years, the development of full form ships has taken
a specific direction because of the demand for an increase in ship The Jelsa systematic series is a series of full ships with large
economics and limitations of main dimensions. These limitations breadth and shallow draught, consisting of thirteen models and
are conditioned by the depth of ports and waterways, and length one control model. This series has arisen from the need to esti-
of cargo quays. The economics is expressed by the ship-owner’s mate the resistance of full ships with large breadth-draught ratio
requirement for the transport of maximum possible cargo quantity because approximate methods for resistance determination for
with minimum fuel consumption. This, together with the pre- such form parameters do not give satisfactory results. The basic
viously mentioned limitations, leads to the conclusion that ship form, the M-938 ship model, Figure 1, is formed by slight mo-
breadth has the maximum possibility of enlargement. This has dification of the main dimensions of the ship type Argosy, which
resulted in the development of full ships with something smaller was constructed in the 3. maj shipyard in Rijeka.
LWL/BWL ratio, but with considerably increased BWL/T ratio which
can now achieve a value of up to 5.0. Block coefficient CBWL has
remained within the same limits (CBWL = 0.8 ÷ 0.85).
Time limitation of the project drafting conditioned the calcula-
tion of the ship resistance and the required power by approximate
methods based on multiannual analysis of the model test results
and trials. However, approximate methods of resistance determi-
nation applied till now have not given exact results for the before
mentioned form parameters. Therefore, the need for testing a new Figure 1 Body plan, fore and aft contour of the basic form, the
systematic model series of full, large breadth and shallow draught M-938 ship model
ships has arisen. The Jelsa series has been developed within the Slika 1 Nacrt rebara, pramčana i krmena kontura osnovne
scope of the research program dealing with the development of forme, model M-938
new hulls and methods for the optimization of hydrodynamic
performances of actual and perspective ship types. The research Previous model tests of full ship type Argosy have shown that
was performed in Brodarski institute in Zagreb, [1], in cooperation the combination of VHC-bulb bow and butterfly stern with bulb
with shipyards 3. maj, Brodosplit and Uljanik. is the best possible solution of the hull form for ships of these
As the resistance of the full-scale ship can not be measured characteristics. Consequently, the same form of the bow and
directly, it is determined from model tests. The measured calm stern is adopted for all models of the Jelsa series. VHC-bulb bow
water resistance is usually decomposed into various components, (V-vertical, H-horizontal, C-cylinder), Figure 2, [3], developed
[2], although all these components interact mutually and most of at Brodarski institute in Zagreb, is suitable for ships with fuller
them can not be measured individually. waterlines for the purpose of transition moderation between bow
The calm water resistance of a ship is a result of shear and taper and bow shoulder. Also, such bow lowers the height of the
normal stresses acting on the wetted surface of a ship. The shear bow wave, which results in a decrease in the required power for
stress is due to the viscosity of the fluid, while the normal stress wave resistance overcome. Butterfly stern with bulb, Figure 3,
can be divided into two components: wave making and a viscous [3], has also been developed in Brodarski institute in Zagreb for
pressure component. The first one is due to the generation of free ships with limited draught, i.e. with large breadth-draught ratio.
surface gravity waves and the second one is caused by a pressure
deficit at the stern due to the presence of the boundary layer. Figure 2 VHC – bulb bow
The standard procedure is to break down the total resistance Slika 2 VHC – bulb forma pramca
into viscous and wave resistance components. The wave resistance
can be decomposed into wave pattern and wave breaking resistance
components, where the latter is present for ships with high block
coefficient. The viscous resistance includes the resistance due to
shear stress (friction resistance) and the viscous pressure resistance
(form resistance). Viscous resistance is usually estimated by using
“ITTC-57 model-ship correlation line” for CF and appropriate
form factor. CF is an approximation for the skin friction of a flat

58(2007)4, 380-388 381

N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC FORM FACTOR DETERMINATION OF THE FULL...

Figure 3 Butterfly stern with bulb

Slika 3 Leptir krma s bulbom

Figure 4 Geometry of the Jelsa series ship models

Slika 4 Geometrija modela serije Jelsa

The stern form has been faired along buttocks because the flow forms, this stern form enables a propeller arrangement of a bigger
around the form takes place approximately along these curves. diameter with higher efficiency. The stern taper on the sides is
In that way the streamline curvature decreases, so this stern form made as smooth as possible, which decreases the viscous pressure
results in a more uniform wake distribution (circular form of the resistance component. The butterfly form of the stern withstands
equal wake curves) and a higher water flow velocity in the upper bigger displacement and enables sharper bow, resulting in the
part of the propeller disc. Although the average nominal wake shift of the centre of buoyancy towards stern. The geometry of
across the propeller disc is considerably lower than for classic the Jelsa series ship models is shown in Figure 4, [4].

382 58(2007)4, 380-388

FORM FACTOR DETERMINATION OF THE FULL... N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC

2.1 Overview of the Jelsa series for full load 2.2 Overview of the Jelsa series for ballast
condition condition
In the analysis of Prohaska’s method suitability it is necessary Ship models of the series have different ratios LWL / BWL, BWL/T,
to distinguish different ship load conditions, i.e. full load and and a constant block coefficient and displacement volume. For
ballast conditions. ballast condition, the following values apply:
Ship models of the series have different ratios LWL / BWL, BWL
/ T, and a constant block coefficient and displacement volume. LWL
For full load condition, the following values apply: = 3.56 ÷ 6.89
BWL
LWL BWL
= 3.75 ÷ 7.25 = 4.58 ÷ 7.74
BWL T
BWL C BWL = 0.787
= 3.0 ÷ 5.0
T
C BWL = 0.835 A Scheme of the Jelsa series for ballast condition is shown in
Figure 6. Main characteristics of ship models for ballast condition
A scheme of the Jelsa series for full load condition is shown are given in Table 2.
in Figure 5. Main characteristics of ship models for full load
condition are given in Table 1.

Figure 5 Scheme of the Jelsa series for full load condition Figure 6 Scheme of the Jelsa series for ballast condition
Slika 5 Shema serije Jelsa za puno opterećenje Slika 6 Shema serije Jelsa za balast

Table 1 Main characteristics of ship models, full load condition

Tablica 1 Glavne značajke modela, puno opterećenje

MODEL M-938 M-939 M-940 M-941 M-942 M-943 M-944 M-945 M-946 M-1015 M-1122 M-1123 M-1124 M-1125
— (m3) 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283 1.283
xCB (%LPP) 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862 1.862
LWL/BWL 5.500 3.750 7.250 5.500 5.500 3.750 3.750 7.250 7.250 4.625 4.625 6.375 6.375 4.625
BWL/T 4.000 4.000 4.000 3.000 5.000 3.000 5.000 3.000 5.000 4.000 4.500 4.500 3.500 3.500
LPP/LWL 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980 0.980
M 5.191 4.021 6.240 4.716 5.591 3.654 4.332 5.670 6.723 4.625 4.810 5.959 5.480 4.424
S 6.564 5.782 7.194 6.039 7.107 5.322 6.260 6.617 7.791 6.198 6.452 7.179 6.614 5.945
CBWL 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835 0.835
CB 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832 0.832
CP 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836 0.836
CX 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995 0.995
CWP 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898 0.898

58(2007)4, 380-388 383

N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC FORM FACTOR DETERMINATION OF THE FULL...

Table 2 Main characteristics of ship models, ballast condition

Tablica 2 Glavne značajke modela, balast

MODEL M-938 M-939 M-940 M-941 M-942 M-943 M-944 M-945 M-946 M-1015 M-1122 M-1123 M-1124 M-1125
— (m3) 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809 0.809
xCB (%LPP) 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870 0.870
LWL/BWL 5.230 3.560 6.890 5.230 5.230 3.560 3.560 6.890 6.890 6.400 4.400 6.060 6.060 4.400
BWL/T 6.110 6.110 6.110 4.580 7.740 4.580 7.640 4.580 7.740 6.110 6.870 6.870 5.340 5.340
LPP/LWL 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031 1.031
M 6.052 4.690 7.278 5.500 6.554 4.261 5.052 6.612 7.840 5.394 5.609 6.950 6.390 5.158
S 7.522 6.617 8.253 6.733 8.360 5.919 7.307 7.385 9.110 7.100 7.466 8.319 7.490 6.720
CBWL 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787 0.787
CB 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.758 0.750 0.758
CP 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769 0.769
CX 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992 0.992
CWP 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899 0.899

The Jelsa series has arisen from the basic form M-938 where CV is the viscous resistance coefficient and CF is the
with the development of thirteen models with the same block friction resistance coefficient in two-dimensional flow.
coefficient and same displacement volume for the same loading If no separation is present, the total resistance coefficient
condition. The development of the series is based on three es- can be written as:
sential equations:
CT = (1 + k ) ⋅ C F + CW (4)
∇ = C BWL ⋅ LWL ⋅ BWL ⋅ T = const
with the wave resistance coefficient CW assumed in the following
LWL / BWL = const form:
BWL / T = con
nst
CW = a ⋅ Fn n (5)
Model tests of the Jelsa series were carried out within the
where a is a coefficient and Fn is the Froude number.
project “Investigation of optimal forms of wide ships with shallow
Power n varies, depending on the block coefficient CB,
draught”, [1]. Ship models were made of paraffin wax in the scale
between 4 and 6.
1:45. Trip wire of 1mm in diameter placed at the station No. 19
Dividing the expression (4) for the total resistance coefficient
was used as a turbulence stimulator. The surface of all models
by the friction resistance coefficient results in the following
was technically smooth. Displacement volumes of all models
equation:
for full load and ballast conditions were — = 1.283 m3 and — =
0.809 m3, respectively. CT C Fn n
The following tests were carried out: resistance tests, self = (1 + k ) + W = (1 + k ) + a ⋅ (6)
propulsion tests, open water tests, measurements of wake field CF CF CF
and streamline tests with paint.
During tests the models were towed by a horizontal force at If the wave resistance component in a low speed region is
different speeds. The point of application of the towing force was assumed to be a function of Fnn, the straight line plot of CT /CF
in the centre of buoyancy. Ship models had all degrees of freedom versus Fnn/CF will intersect the ordinate (Fn = 0) at (1 + k),
except the deflection from the line in the horizontal plane. enabling the form factor to be determined. For drawing such
straight line it is necessary to measure about ten points in the
Froude number range Fn = 0.1 ÷ 0.2(0.22), because in that range
3 Prohaska's method the wave resistance is negligible. The uncertainty of measuring
Prohaska’s method, [5], is based on the Hughes’ method, [6], resistance at very low speed is relatively large, which means that
but the two methods differ in the definition of the form factor. it is difficult to determine the “run-in-point” exactly. Friction
Prohaska defines the three-dimensional form factor as follows: resistance coefficient is calculated according to the “ITTC-57
model-ship correlation line”:
CV − C F CV
k= = −1 (2) 0.075
CF CF CF = (7)
(log Rn − 2)2
CV
= 1+ k (3) For full hull forms, CB > 0.8, the points may plot on concave
CF curves indicating that either (1 + k) or a, or both, are speed de-

384 58(2007)4, 380-388

FORM FACTOR DETERMINATION OF THE FULL... N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC

pendent. It is more appropriate for full ships to use a power of

Fn between 4 and 6 instead of 4, [2].
A regression method is used to determine the magnitude of
form factors. When experimental information is represented by a
regression, there are two kinds of uncertainties. One is uncertainty
due to the uncertainty in the original experimental results; the
other is introduced if the wrong regression model is used. ITTC
recommended Prohaska’s method for experimental evaluation
of the form factor, [7].
Viscous resistance is then calculated according to the fol-
lowing equation:

RV =
1 2
2
( (
ρv S C F 1 + k )) (8) Figure 9 Relative error comparison for the M-944 ship model, full
load condition
Slika 9 Usporedba relativne pogreške za model M-944, puno
4 Results opterećenje

4.1 Full load condition

The results for full load condition, [8], are shown in Figures 7-
13 and Table 3. Only the results for the central model and models
with extreme values of LWL / BWL and BWL / T are presented in this
paper. Other experimental results are available on request.
Relative error between measured and calculated values of the
total resistance coefficient can be determined by:
Figure 10 Relative error comparison for the M-945 ship model,
CT ,measured − CT ,Pr ohaska
( )
Relative error % =
CT ,PPr ohaska
⋅100 (9)
full load condition
Slika 10 Usporedba relativne pogreške za model M-945, puno
opterećenje

Figure 11 Relative error comparison for the M-946 ship model,

full load condition
Figure 7 Relative error comparison for the M-938 ship model, full Slika 11 Usporedba relativne pogreške za model M-946, puno
load condition opterećenje
Slika 7 Usporedba relativne pogreške za model M-938, puno
opterećenje Figure 12 Form factor as a function of the LWL / BWL ratio, full load
condition
Figure 8 Relative error comparison for the M-943 ship model, full Slika 12 Faktor forme u ovisnosti o omjeru LWL / BWL, puno
load condition opterećenje
Slika 8 Usporedba relativne pogreške za model M-943, puno
opterećenje

58(2007)4, 380-388 385

N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC FORM FACTOR DETERMINATION OF THE FULL...

Table 3 Form factor and maximum relative error for full load condition
Tablica 3 Faktor forme i maksimalna relativna pogreška za puno opterećenje

Fn4 Fn5 Fn6

Maximum Maximum Maximum
Model LWL / BWL BWL /T Form factor, k Form factor, k Form factor, k
relative error, % relative error, % relative error, %
M-938 5.500 4.000 0.232 2.341 0.238 2.198 0.241 2.081
M-939 3.750 4.000 0.379 2.088 0.409 2.878 0.429 3.754
M-940 7.250 4.000 0.032 2.275 0.055 2.615 0.071 3.144
M-941 5.500 3.000 0.241 4.743 0.256 3.776 0.267 3.435
M-942 5.500 5.000 0.166 1.607 0.178 1.701 0.187 1.964
M-943 3.750 3.000 0.833 6.228 0.868 5.998 0.893 6.059
M-944 3.750 5.000 0.358 4.036 0.381 3.852 0.397 3.557
M-945 7.250 3.000 0.154 2.174 0.162 2.224 0.166 2.256
M-946 7.250 5.000 0.074 2.192 0.092 1.871 0.103 1.616
M-1015 4.625 4.000 0.27 3.746 0.285 3.435 0.295 3.192
M-1122 4.625 4.500 0.239 2.764 0.255 2.992 0.266 3.149
M-1123 6.375 4.500 0.146 6.077 0.161 5.871 0.172 5.713
M-1124 6.375 3.500 0.161 3.581 0.17 3.525 0.177 3.444
M-1125 4.625 3.500 0.417 5.532 0.419 5.463 0.421 5.433

Figure 13 Form factor as a function of the BWL / T ratio, full load

condition
Slika 13 Faktor forme u ovisnosti o omjeru BWL / T, puno opte- Figure 15 Relative error comparison for the M-943 ship model,
rećenje ballast condition
Slika 15 Usporedba relativne pogreške za model M-943,
balast
4.2 Ballast condition
The results for ballast condition, [8], are shown in Figures 14-
20 and Table 4. Only the results for the central model and models
with extreme values of LWL / BWL and BWL / T are presented in this
paper. Other experimental results are available on request.
Figure 16 Relative error comparison for the M-944 ship model,
ballast condition
Figure 14 Relative error comparison for the M-938 ship model, Slika 16 Usporedba relativne pogreške za model M-944,
ballast condition balast
Slika 14 Usporedba relativne pogreške za model M-938,
balast

386 58(2007)4, 380-388

FORM FACTOR DETERMINATION OF THE FULL... N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC

Figure 17 Relative error comparison for the M-945 ship model, Figure 18 Relative error comparison for the M-946 ship model,
ballast condition ballast condition
Slika 17 Usporedba relativne pogreške za model M-945, Slika 18 Usporedba relativne pogreške za model M-946,
balast balast

Each curve in Figures 7-11 and Figures 14-18 is identified by the standard deviation.

Table 4 Form factor and maximum relative error for ballast condition
Tablica 4 Faktor forme i maksimalna relativna pogreška za balast

Fn4 Fn5 Fn6

Maximum Maximum Form factor, Maximum

Model LWL / BWL BWL /T Form factor, k Form factor, k
relative error, % relative error, % k relative error, %

M-938 5.230 6.110 0.219 2.059 0.237 2.499 0.25 2.941

M-939 3.560 6.110 0.437 5.213 0.461 4.735 0.477 4.754
M-940 6.890 6.110 0.084 3.173 0.115 3.203 0.136 3.119
M-941 5.230 4.580 0.25 2.373 0.273 1.808 0.29 2.388
M-942 5.230 7.740 0.169 1.694 0.194 1.903 0.21 2.41
M-943 3.560 4.580 0.61 4.583 0.647 4.024 0.673 4.264
M-944 3.560 7.740 0.336 3.028 0.363 3.435 0.381 3.688
M-945 6.890 4.580 0.123 1.578 0.145 1.92 0.16 2.362
M-946 6.890 7.740 0.127 2.385 0.145 2.558 0.158 2.773
M-1015 4.400 6.110 0.288 2.173 0.308 2.049 0.322 2.347
M-1122 4.400 6.870 0.22 0.867 0.242 1.106 0.258 1.455
M-1123 6.060 6.870 0.116 4.439 0.133 4.002 0.145 3.575
M-1124 6.060 5.345 0.113 2.543 0.127 3.028 0.131 2.929
M-1125 4.400 5.345 0.304 3.161 0.324 2.592 0.337 2.528

Figure 20 Form factor as a function of the BWL / T ratio, ballast

condition
Slika 20 Faktor forme u ovisnosti o omjeru BWL / T, balast
Figure 19 Form factor as a function of the LWL / BWL ratio, ballast
condition
Slika 19 Faktor forme u ovisnosti o omjeru LWL / BWL, balast

58(2007)4, 380-388 387

N. DEGIULI, N. HADŽIĆ, M. PEDIŠIĆ-BUČA, G. SEMIJALAC FORM FACTOR DETERMINATION OF THE FULL...

5 Conclusions for the full load condition the maximum relative error is among
the lowest.
Based on the results of resistance tests performed with four- Obtained values point out to the fact that total resistance
teen models of the Jelsa series, (full, large breadth and shallow measurements in this Froude number range represent a particular
draught ships), the form factors for full load and ballast conditions problem and impose strict requirements on the equipment for
have been determined and presented in the paper. The tests were total resistance measurement.
carried out in Brodarski institute, Zagreb. For the form factor Results of the analysis have shown that all three powers give
determination, Prohaska’s method has been applied. Since full fairly good agreement with measured values since maximum
ships are concerned, greater powers of the Froude number in the relative error amounts to 6.23% for full load condition, and to
wave resistance expression have been used. 5.21% for ballast condition. The average value of the maximum
In the process of the form factor determination using linear relative error for all models amounts to 3.49% for the full load
regression analysis, the values of the determination coefficient condition and to 2.84% for the ballast condition.
for full load condition are within the range 0.6978<R2<0.9819, Since the form factor is used for the calculation of the viscous
except for the M-1125 ship model for which the range is resistance (see equation 8), it can be said that the relative error of
0.1508<R2<0.1927. Form factor values are within the range
the viscous resistance calculation, due to the relative error of form
0.055<k<0.893.
factor value for the tested Jelsa series, has the maximum value
In the process of the form factor determination using linear
of 2.83% for the full load condition and of 1.59% for the ballast
regression analysis, the values of the determination coefficient for
condition. That means that Prohaska’s method is quite appropriate
ballast condition are within the range 0.8904<R2<0.9927. Form
for the determination of the form factor of the tested ship series.
factor values are within the range 0.084<k<0.673.
As far as the dependency of the form factor value on variable The best results for most models at full load condition are
characteristics of the Jelsa series is concerned, the following can obtained by applying power n = 6 in the wave resistance ex-
be concluded: pression, and by applying power n = 4 at ballast condition. It
- as the LWL / BWL ratio gets larger, the value of the form factor is important to emphasize that the difference of the maximum
k is smaller for the both tested conditions, i.e. full load and relative errors of the results obtained with the powers 4, 5 and 6
ballast conditions, is not large. No connection between a particular model and the
- dependency on the BWL / T ratio is not significant. power of the Froude number has been noted.
Such results are in accordance with expectations because the In the future work the values of the form factors should be
form factor value represents a portion of the form resistance in compared with the same values obtained by applying Hughes’ and
the total viscous resistance, which is lower for finer hulls. Shirose’s methods, [9], and the method proposed by Zotti [10].
Maximum value of k for ballast and full load conditions
occurs when the values of both ratios, LWL / BWL and BWL / T, are References
minimum in the tested range. The minimum value of k for the
[1] “Istraživanje optimalnih formi širokih brodova malog gaza”,
full load condition occurs when the values of the ratios LWL / BWL
Interni izvještaj, Brodarski institut, Zagreb, 1990.
and BWL / T are maximum, and for the ballast condition when
[2] HARVALD, Sv.Aa.: “Resistance and Propulsion of Ships”, John
LWL / BWL is maximum, and BWL / T is in the middle of the tested Wiley and Sons, New York, United States of America, 1983.
range. Values of the form factors for the ballast condition are [3] Gamulin, A.: “Otpor broda”, skripta, Brodarski institut, Zagreb,
smaller and more equable than for the full load condition. That 2000.
is also according to expectations because CB has a smaller value [4] ŠKOLNEKOVIĆ, I.: “Puni brodovi velike širine i malog gaza
for the ballast condition. s koeficijentom istisnine CBWL = 0,835, serija “Jelsa” - analiza”,
But, similar conclusions cannot be drawn about the dependen- Diplomski rad, Fakultet strojarstva i brodogradnje Sveučilišta
cy of the relative error on the variable characteristics of the Jelsa u Zagrebu, Zagreb, 1997.
series. Dependency of the relative error on LWL / BWL and BWL / T [5] PROHASKA, C.W.: “A Simple Method for Evaluation of Form
ratios is not significantly pronounced. For the ballast condition it Factor and the Low Speed Wave Resistance”, Proc. 11th ITTC,
can be noted that the influence of the LWL / BWL ratio is present, but 1996, p.65-66.
the influence of BWL / T is not significant at all. Ship models with [6] HUGHES, G.: “Friction and Form Resistance in Turbulent
the highest and lowest values of of the determination coefficients Flow and a Proposed Formulation for Use in Model and Ship
have, for example, the same BWL / T ratio. Ship models with the Correlation”, Trans. RINA, VOL. 96, 1954, p.314-376.
lowest values of LWL / BWL have lower values of the determination [7] Final Report and Recommendations to the 24th ITTC, The
coefficients compared to other models. On the other hand, at the Resistance Committee, Proceedings of the 24th ITTC-Volume
full load condition the dependency on the LWL / BWL ratio is not I, United Kingdom, 2005.
evident. When the values of the BWL / T ratio are equal to 3 and [8] HADŽIĆ, N.: “Određivanje faktora forme serije punih brodova
3.5 for the full load condition (which are the lowest values for velike širine i malog gaza”, Završni projekt, Fakultet strojarstva
i brodogradnje Sveučilišta u Zagrebu, Zagreb, 2007, p.123.
the tested range), the values of the determination coefficients are
[9] SHIROSE, Y., HIRONO, S.: “A method to determine form
mainly lower than for the other values of this ratio.
factors of ship models from resistance tests”, IHI Review, vol.15,
The connection of a particular ship model and relative errors n.3, 1982, p.9-13.
for the full load and ballast conditions, respectively, is also not [10] ZOTTI, I.: “The Form Factor Calculation: A Critical Analysis and
present. For example: for the M-1123 ship model maximum a New Proposal for Full Form Hulls”, IX CONGRESS IMAM
relative errors are among the greatest for both conditions, on the 2000, Ischia, Naples, April 2000, Proceedings Vol. I, p.32-39.
other hand, for the M-939 ship model the maximum relative error [11] ROSEMAN, D.P.: “The MarAd Systematic Series of Full Form
for the ballast condition is the greatest for the tested models, and Ship Models”, SNAME, 1987, p.384.

388 58(2007)4, 380-388