Escolar Documentos
Profissional Documentos
Cultura Documentos
SUMMARY: It is well known that there is a lack of published systematic data and techniques suitable for the concep-
tual and preliminary design of high-speed ferries. On the contrary the interest and the applications of high-speed vessels
are increased, and so the development of available design process data and properly techniques for their designs are
strongly required.
In this regards, to avoid major recalculation at later design stages, it is fundamental to develop data and procedures,
which can be used for providing the starting point to estimate technical aspects such as main dimensions, masses
value, powering and material choice.
First of all the paper presents the analysis of a database, which includes high-speed ferries operating in the world; this
analysis permits to present a simplified procedure to individuate feasible and reliable initial calculation for the main di-
mensions, masses value and materials.
Successively, based on Steel and/or Aluminum materials and starting from a representative fast ferries, a parametric hull
mass estimation has been carried out using U.N.I.T.A.S. Rules of Bureau Veritas, Registro Italiano Navale and German-
isher Lloyd, both for monohulls and catamarans.
800
R = 0,812
equipment area, etc.).
From these data is possible to determine the following
600
parameters:
400 only pax PA = Passengers area;
200 car/pax DN = Number of decks;
KP2 = Percentage of passengers on main deck;
0 KP3 = PAMD/(LOA B)
0 1000 2000 3000 4000 where PAMD is the main deck area intended to passen-
gers.
Figure 1 pass number vs. displacement (t) for monohulls
PA can be obtained from:
Session A 160
LWL/BWL is a strategic parameter and the designer will D = T + 2.7 DN + 2 (5)
choose it according to stability consideration and his ex-
perience. Nevertheless, for fast ferries usually LWL/BWL
goes from 5.0 to 7.5.
To explain the proposed procedure, Table 1 shows step
Regarding B/BWL the data set is small for statistical
by step the proposed calculations.
analysis, taking also into account that some of the data
are referred to cars-passenger ferries, whose beam de-
pends strongly on garage geometry.
Figures 4 and 5 show the relation between the vessel Step Item Source
length and the parameters LOA/ LWL and B/BWL. S1 KP1 Initial hypothesis
S2 LOA/LWL Initial hypothesis
1.30 S3 BOA/BWL Initial hypothesis
LOA/LWL S4 LWL/BWL Initial hypothesis
1.25 S5 KP3 Initial hypothesis
1.20
S6 KP2 Initial hypothesis
S7 PA KP1 PN
1.15 S8 PAMD PA KP2
S9 LOA PAMD / K3
1.10 BOA
1.05
S10 KP1 4.210-4 (LOA BOA) + 0.91 (
LOA [m] 0.3or not)
1.00 New loops as long as KP1 is confirmed
20 40 60 80 100 120 140 Sn LOA/BO (LOA/LWL) (LWL/BWL) /
A (BOA/BWL)
Figure 4 Sn+1 BOA [(LOA BOA) / (LOA/BOA)] 0.5
Sn+2 LOA (LOA/BOA) BOA
1.30 Sn+3 LWL LOA / (LOA/LWL)
BOA/BWL Sn+4 BWL BOA / (BOA/BWL)
Sn+5 L/1/3 -2.210-4 LWL + 0.057 LWL + 5.5
1.25
Sn+6 [L/(L/1/3)]3
Sn+7
1.20 Sn+8 CB 0.158 Fn + 0.302
Sn+9 T / (LWL BWL CB)
Sn+10 D T + 2.7 DN + 2
1.15 Table 1.
LOA [m]
2.2 CARS-PASSENGERS FERRIES (MONOHULL)
1.10
20 40 60 80 100 120 140
As for the passengers ferries, also for this kind of vessels
Figure 5. a sample not enuogh sufficient for a correct statistics
In the first stage of design, the evaluation of CB and analysis has been analysed. The reference sample is
L/1/3 is a very hard task; their values have a great influ- compound from ferries whose significant dimensions
ence on and T. If there are not reliable indications, ac- are: LOA = 71142 m, CN = 57425 and V = 3043 kn.
cording to our experience and data, we suggest the fol- The whole sample is of 43 vessels but for every quantity
lowing relations for determining CB and L/1/3 has been considered reliable a smaller number of these
(10 13 vessels).
CB = 0.11 Fn + 0.34 (3)
Cars-passengers fast ferries main dimensions are strongly
influenced by cars number CN - cars lines numbers
L
1
= -2.2 10 4 L WL 2 + 0.057 L WL + 5.0 (4) CLN - and by how many floors are in the garage. This
3 parameter is quite dependent by the percentage of cars to
stow in the main level - KC2 - and of cars specific area
The equation (3) is valid for 0.5 < Fn < 0.9, while equa- (i.e. KC1 = Total Net Area of Garage / Cars Number).
tion (4) for 25 m < LWL < 125 m. Finally, for the first evaluation of the vessel characteris-
For the evaluation of the depth overall D (with super- tics, it is also necessary to assume the percentage of LWL
structures); we can use the following equation: BG destined to the reference garage KC3.
The range of KC2 is from 0.4 to 1, it depends from the
level numbers of the garage. For a layout on two levels,
included values between 0.5 and 0.65 are conformed.
Session A 161
The KC3 value depends on the main engines exhaust ar- the initial parameters. Nevertheless, at the later stage of
rangement and on the adopted layouts for vertical cargo design, after the first layout of the garage, with a reliable
shifts (cars ramps and passengers scale). The range of definition of lifting installations and exhaust arrange-
KC3 is from 0.70 to 0.85; it increases with the vessel di- ments, the same procedure, based on the updated dimen-
mensions. sions, become sensitively more reliable.
The KC1 value influences strongly the whole procedure.
If no indications are available, the following relation is
suggested: 14
D [m]
KC1 = 1.610-5 CN2 0.015 CN + 13.4 (6) 13
The following relation can be utilized to determine the 12
breadth of the garage BG:
11
BG = 2.2 CLN + 0.5 (7)
10
From the garage dimensions, the first evaluations of the
other form characteristics of the hull are estimable. 9
In particular to the BG are linked the reference breadth 8
B* (including the reinforced frame) and the BWL. Avail- D
able data suggest the following relations: 7 POLINOM
LOACLN
B*=1.06 BG (8) 6
400 600 800 1000 1200 1400
BWL=0.87 BOA (9)
Figure 7.
It is to be noted that usually the bulwarks of fast ferries
are very flared. Then it has been introduced B*, which is Table 2 shows all the steps of the procedure
an intermediate value between the maximum breadth B
and the BWL.
The correlation between the LWL and the LOA is shown in STEP ITEM SOURCE
Figure 5, whose data permit to obtain the following rela- 1.610-5 CN2 0.015CN +
S1 KC1
tion: 13.4
S2 CLN Initial hypothesis
LOA=210-3 LWL2 + 1.44 LWL7 (10) S3 KC2 Initial hypothesis
S4 KC3 Initial hypothesis
180
S5 NCPG NC KC2
160 LOA S6 APG KC1 NCPG
140 S7 LWLBG APG / KC3
120 S8 BG 2.2 NCL + 0.5
S9 BOA 1.06 BG
100
S10 BWL BOA / 1.15
80 S11 LWL (LWLBG) / BG
60 DATA S12 LOA 210-3 LWL2 7.0
40 S13 Fn V / (9.81 LWL)0.5
POLINOM. S14 CB 0.158 Fn + 0.302
20 LWL 2.210-4 LWL2 + 0.057 LWL +
0 S15 L/1/3
5.0
40 60 80 100 120 140 160 S16 LWL / L/1/3
Figure 6. S17 1.025
S18 T / (LWL BWL CB)
The relations to estimate CB and L/1/3 are also (3) and S19 D 4610-3LOACNL + 6.9
(4).
Table 2
Finally we suggest the following relation to obtain the
Depth D, calculated from the Baseline to the strength 3. STRUCTURAL WEIGHT ESTIMATION
deck. In Figure 6 data and the linear regression are Little changes of weight can have a significant effect on
shown. the performance of high-speed vessels.
Therefore, the need of appropriate methodologies to es-
D = 4610-3 LOA CNL + 6.9 (11) timate the hull and superstructures weights, is particu-
larly important.
The proposed procedure is usable for the first evaluation It must be distinguishing between two different
of the main dimensions and displacement. The critical methodologies; the first, that is approximate, can be
aspect is the coherence between the values attributed to easily used in the early stages for conceptual and
preliminary design. The other, rationally exactly, can be
Session A 162
The other, rationally exactly, can be utilized when the 1) it is supposes that the vessels (reference and derived)
basic design has been established and it could be devel- have similar hull weight distributions, then it is pos-
oped in more details. sible to write:
It appears evident from the weights list that, in the first
stage of the design process, poor information regarding W S w0 Loa
= (13)
the list of lightship weight are available, and so it must W 'S w'0 Loa '
be subject to approximate treatment more of others pa-
ramenters; starting from the lightship weight evaluation, where:
the hull and superstructures weights appear to be sub-
W S is the hull weight of the derived vessel;
jected to more uncertainty.
Watson and Gilfillan (1977), considered the weight of W 'S is the hull weight of the reference vessel;
the structures, which includes:
a) the weight of the basic hull to its depth amidships; w0 is the unitary hull weight of the midship section for
b) the weight of the superstructures (those full whose the derived vessel;
width is the extensions of the hull and that are above w'0 is the unitary hull weight of the midship section for
the basic depth amidships, e.g. a raised forecastle);
c) the weight of the deckhouses (those less than full the reference vessel.
width extensions on the hull and superstructure). Loa, Loa are respectively the length over all of the ref-
The subdivision is suitable to the commercial ships; be- erence and derived vessels.
cause ships are composed of stiffened panel, the area
2) it is supposed that the vessels (reference and derived)
Loax(B+D) has been taken as independent variable. Thus,
have similar superstructures weight distributions, then it
Watson and Gilfillan proposed a modeling approach us-
is also possible to write:
ing the Equipment Number E - with the hull, the super-
structure and deckhouse elements contributions. w s w su 0 l
= (14)
w's w'su 0 l '
3.1 THE WATSON AND GILFILLAN METHOD where:
EXTENTS TO HIGH-SPEED FERRY
ws is the superstructures weight of the derived vessel;
The more reliable methods to estimate the hull and
superstructures weight seem to be: w's is the superstructures weight of the reference vessel;
one of Karayannis at al. applicable to the monohulls wsu0 is the unitary weight of the superstructures mid-
and catamarans in aluminum alloy; ship section for the derived vessel;
one of Watson and Gilfillan (1977) applicable to the
conventional ships in steel. w'su 0 is the unitary weight of the superstructures mid-
Both the methods are based on the choice of a single ship section for the reference vessel.
variable that must necessarily reflect the influence of the
complete dimensions of the ship. l, l are respectively the maximum length of the reference
Such variable is equal to: and derived vessels superstructures (normally they are
( )
*
between 70 and 85 percent of the Loa).
E m = Loa B + T + 0.85 Loa ( D T ) - monohull - (12)
It is possible now to evaluate w0 by the following ex-
It is interesting to note that for the most only passengers pressions:
ferries, due to their integrated shape nature and same ma-
terials (aluminum alloy), the total weight of the hull and WT
superstructures may be considered as single weight. w0 = a A s + for only passengers vessels (15)
I
While for the most cars/passengers ferries, even if the
hull and superstructures can be integrated, due to their WTh
w0 = Ahs + for cars/passengers vessels (16)
different materials (the hull normally in high-steel ERS I
36- and the superstructures in aluminum alloy) a separa-
with:
tion of weights should be considered.
= steel density;
Based on the above considerations, it is possible to de- A hs = hull cross section structural area;
velop a procedure only for high speed vessels, which
permits to add data (to increase the suitability of the only a = aluminium alloy density;
known datum), considering the following assumptions: A s = total cross section structural area;
Session A 163
WT = frame weight (including hull and superstructures); using basic ship lengths between 30 and 70 meters
(range of Em between 300 and 1500 square meters),
I = frame interval respectively ordinary for transversal which are the most representative vessels of this cate-
structural type (reinforced for longitudinal structural gory.
type). Two regressions formulae can be proposed:
WTh = frame hull weight.
W s
= 2.056 *106 * Em ^ 2 + 0.10 * Em R
2
= 0.9897 (18)
The value of w su 0 is: (19)
W
2
s
= 0.053 * E m ^1.12 R = 0.9881
Session A 164
W S = 1.158 *10 * Em ^ 2 + 0.20 * Em R 2 = 0.9708 utilize the (20) and/or (21) relations, where the only pa-
5
rameter Em is required.
(20) In the case of catamaran only passengers numerical re-
sults for hull weight estimation are shown in Table 3
7
w s = 6.281*10 * Em ^ 2 + 0.012 * Em where Ecat is:
300 160.00
200 140.00
100 120.00
Em [m 2]
0 100.00
0 500 1 000 1 500 2 000 2 500 3 000 3 500 4 000
tonn
80.00
Figure 8
60.00
It is to be pointed out that at the early stage of the design,
the estimation of the hull weight typically involves the 40.00
use of models, which are developed from weight infor- (Ecat) (m^2)
20.00
mation for similar vessels. At this state, the designer does
0.00
not know the longitudinal distribution of the hull and su-
0.00 500.00 1 000.00 1 500.00 2 000.00
perstructures weights, but he had to estimate the main
dimensions of vessel (and then Em). The developed Figure 9
method is derived mainly from the utilization of the rela-
tions (13) and (15) (which are based on the hypothesis The results till now obtained regard the hull weight of
catamarans (hull and superstructures in aluminum) for
that reference and derived vessels have similar hull
only passengers high-speed vessels, and have been ob-
weight distributions). On the contrary the designer can
Session A 165
tained by the (13) and (15) equations, where midship di- Coppola, T.; Paciolla, A.; Quaranta, F. 2004. On the De-
mensioning has been obtained applying U.N.I.T.A.S. sign of Propulsion Systems and Hull Mass Estimates
Rules of Bureau Veritas, Registro Italiano Navale and for High-Speed Ferries. 4th International Congress on
Germanisher Lloyd. It is interesting also to note that the Maritime Technological Innovations and Research.
catamarans only passengers analysis has been carried out Barcelona, Spain.
using basic ship length between 20 and 60 meters (range Karayannis, C.T.; Molland, A.F. and Sarac Williams Y.
of Ecat between 150 and 1500 square meters), which are 1999. Design Data for High-Speed Vessels. FAST
99, Seattle, USA..
the most representative vessels of this category.
Molland, A.F. and Karayannis, T. 1997 Development of
The regression formulae can be proposed: a Concept Exploration and Assessment Mode for
6
W s = 2.05 *10 * Ecat ^ 2 + 0.087 * Ecat R 2 = 0.995 Advanced Fast Marine Vehicles. The Sixth Interna-
tional Marine Design Conference Design IMDC97
(23) Newcastle upon Tyne, UK.
W s = 0.039 * Ecat ^1.12 R 2 = 0.985 (24) Parsons, G.M. Ship Design and Construction I Volume
Chapter 11 Parametric Design Tomas Lamb editor.
Watson, D.G.M. and Gilfillan, A. W. 1977. Some Ship
4. CONCLUSIONS Design Methods. Trans. of The Royal Institution of
Naval Architects. Vol. 119.
Some principles for determining the displacement and a Bureau Veritas Rules for the Classification of High
set of main dimensions, which can be used as starting Speed Craft, 1997;
values for the design of high-speed fast ferries monohull,
have been presented.
Moreover, the exposed methodologies are suitable, with
more reliability, to evaluate the variational sensitivity of
the strategic dimensions in the design of vessels, whose
typology is quite known (i.e. are reliable and coherent
KP1, KC1, etc.).
An extension of the hull weight regression for only pas-
sengers and cars/passengers monohulls, based on a pa-
rametric procedure with vessels having similar hull
weight distribution, has been carried out.
The hull weight regression has been obtained also for
catamarans only passengers in aluminum, considering
the weight of hull and superstructures as single weight.
The methodologies and algorithms developed make
available further developments and refinements, includ-
ing the most representative cars/passengers high-speed
vessels catamarans, whose mains objectives will be aim
of successive researches.
ACKNOWLEDGEMENTS
REFERENCES
Session A 166