See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/257456232

Hydrodynamic hull form optimization using parametric models

Article  in  Journal of Marine Science and Technology · March 2012

DOI: 10.1007/s00773-011-0148-8

CITATIONS READS

46 723

3 authors, including:

Soonhung Han Yeon-Seung Lee

Korea Advanced Institute of Science and Technology Hongik University
213 PUBLICATIONS   1,998 CITATIONS    11 PUBLICATIONS   111 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hausdorff Distance Computation View project

Hello, I am a professor at naval architecture and oceam engineering at Hongik university of KOREA. View project

All content following this page was uploaded by Soonhung Han on 10 November 2015.

The user has requested enhancement of the downloaded file.

J Mar Sci Technol (2012) 17:1–17
DOI 10.1007/s00773-011-0148-8

REVIEW ARTICLE

Hydrodynamic hull form optimization using parametric models

Soonhung Han • Yeon-Seung Lee • Young Bok Choi

Received: 1 April 2011 / Accepted: 28 October 2011 / Published online: 1 February 2012
 The Author(s) 2012. This article is published with open access at Springerlink.com

Abstract Hydrodynamic optimizations of ship hull forms Keywords Fairness optimized B-Spline form parameter
have been carried out employing parametric curves gen- curve  Constrained transformation function  Full
erated by fairness-optimized B-Spline form parameter parametric hull form design  Hydrodynamic optimization 
curves, labeled as F-Spline. Two functionalities of the Fuel economic ship design  Green ship technologies
parametric geometry models are used in the present study:
a constrained transformation function to account for hull List of symbols
form variations and a geometric entity used in full para- AP Aftward perpendicular
metric hull form design. The present F-Spline based opti- AM Midship section area
mization procedure is applied to two distinct hydrodynamic B Maximum beam
hull form optimizations: the global shape optimization of CB = r/(LPP BT) Block coefficient
an ultra-large container ship and the forebody hull form for CF Frictional resistance
the hydrodynamic optimization of an LPG carrier. coefficient
Improvements of ship performance achieved by the pro- CM = AM/(BT) Sectional area coefficient at
posed F-Spline procedure are demonstrated through midships
numerical experiments and through correlations with CP=r/(LPP AM) = CB/CM Prismatic coefficient
experimental data. The ultra-large containership was built CR Residual resistance
and delivered to the ship owner. The present study vali- coefficient
dates the effectiveness of the proposed hydrodynamic CTS Total resistance coefficient in
optimization procedure, ushering in process automation full scale
and performance improvement in practical ship design CFITTC Frictional resistance
practices. coefficient by ITTC
CW Wave resistance coefficient
from pressure integration
S. Han  Y.-S. Lee (&) FN Froude number
Department of Ocean System Engineering, KAIST, FP Forward perpendicular
291 Daehak-ro, Yuseong-gu, Daejeon 305-701, LPP Length between
Republic of Korea
perpendiculars
e-mail: [email protected]
LE Length of SAC entrance
S. Han
LM Length of parallel middle
e-mail: [email protected]
body
Y. B. Choi LR Length of run
Hydrodynamics R&D Group, Ship and Ocean R&D Institute, PD Delivered power
Daewoo Shipbuilding and Marine Engineering Co., Ltd.
RT Total resistance
(DSME), 85, Da-dong, Jung-gu, Seoul 100-180,
Republic of Korea RTM Total resistance in model
e-mail: [email protected] scale

123
2 J Mar Sci Technol (2012) 17:1–17

RTS Total resistance in full scale Implementation of these technologies as tools for ship
RWP Wave resistance from wave design requires a sufficient understanding of each tech-
cut analysis nique, the use of practical design experience and method-
RA Correlation allowance ologies in the optimization process, and the appropriate
RAA Wind resistance evaluations of the optimization results to verify improved
RFITTC Frictional resistance by ITTC ship operation performance. With the development of each
S Wetted surface area technology, CFD simulation-based ship design (SBD) has
T Design draft continuously improved and become more appropriate for
TFP Design draft at forward practical use [3–9, 16, 17, 26, 27, 30–32, 34]. The current
perpendicular CFD simulations make it possible to reliably predict and
TAP Design draft at aftward validate ship resistance, propulsion performance and sea-
perpendicular keeping performance [4, 13, 29, 35]. Although ship design
r Volume of displacement is generally a multi-objective optimization problem,
objective function evaluations based on a RANS solver
Abbreviations require effectiveness and efficiency. Effective global opti-
CAD Computer aided design mization algorithms such as variable fidelity models, filled
CAE Computer aided engineering function-based algorithms, and a particle swarm optimi-
CFD Computational fluid dynamics zation have been introduced for SBD [3, 21, 24, 32].
DoE Design-of-experiments In the field of ship geometry design, increasingly
DWL Design waterline sophisticated design methodologies are continuously sug-
EFD Experimental fluid dynamics gested and validated [6, 9, 15–20, 22, 23, 26, 28, 32]. An
FOB Flat of bottom SBD project entitled ‘FANTASTIC’, whose goal was to
FOS Flat of side improve the functional design of ship hull shapes, was
F-Spline Fairness-optimized B-Spline form parameter curve implemented over a period of 3 years by 14 European
ITTC International towing tank conference partners [16]. As a result, three approaches to shape
LCB Longitudinal center of buoyancy modeling emerged: ship parametric modeling using
PMB Parallel middle body Friendship-modeler software, a template approach involv-
RANS Reynolds-averaged Navier–Stokes equations ing the use of NAPA [16], and shape transformation
SAC Sectional area curve functions via GMS/Facet [16]. Major progress in the
SBD Simulation based design optimization process has been obtained through the use of
parametric modeling, which provides the most flexible ship
design variations. Finally, geometric modeling for SBD
1 Introduction addresses the problem of shape generation and transfor-
mation as necessary to generate design variations with the
Green ship technologies are gaining in importance in required flexibility and functionality. These efforts not only
diverse areas of ship design. Because CO2 emission levels optimize ship hull performance but also ensure the effec-
are directly related to fossil fuel consumption, shipbuilding tiveness of the automated design process.
industries are focusing on developing new design concepts The simplest and, therefore, most widely used method of
and technologies towards fuel economic ship design generating hull shape parameters is the fitting of discrete
including the use of new propulsion devices and renewable ship offset data with parametric representations such as
energy sources. Hydrodynamic optimal shape designs are polynomials, cubics, and Bezier and B-Spline curves or
one component of the overall fuel economic design, surfaces. The discrete data obtained, including hull offset
although the percentage reduction of CO2 emission can be data and control points, are directly used as design vari-
achieved 2–3% at most. However, given the prevailing ship ables to modify the shape. This approach provides flexi-
construction process, it is an indispensable step, because bility in controlling each control point but results in less
hull form design is a starting point of the new shipbuilding functionality because the high degree of freedom in control
process and it influences resistance and propulsion per- points variations can cause deviations from the desired
formance. In addition to its hydrodynamic effects, hull shape [10, 26, 33].
form design influences costs, the construction process and In a more sophisticated method, the ship shape is par-
safety considerations in shipyards. titioned into several sections, defined within several boxes
There are three core technologies used for hydrody- and controlled by using a small number of control points as
namic hull form optimization: geometric modeling, design variables. An effective grouping of control points
hydrodynamic analysis, and optimization technologies. can yield sufficiently large and realistic changes in the hull

123
J Mar Sci Technol (2012) 17:1–17 3

form [14, 16, 31]. If the hull shape is generated using form 2.1 Fairness-optimized parametric curve
parametric data and a fairing process, the shape can be with constraints [2, 6, 19, 20]
generated using a highly sophisticated fully parametric
model. This fully parametric model generates the entire The general formulation of the fairness-optimized para-
hull surface from a limited parameter set based on higher- metric curve with constraints follows below:
level geometrical properties [1, 6, 18–20, 28]. Let r(t) be a vector valued free form curve parameter-
A set of transformation functions can be defined by ized by t
types of data including analytical curves, surface data, rðtÞ ¼ ðxðtÞ; yðtÞ; zðtÞÞ ð1Þ
and discrete data. They can then be used in any type of
geometry, offset data, or functional model. Although the Minimizing the mth order fairness criterion yields Lm,
formulation of the transformation function is a critical which is given by
issue in geometric variation, it is mainly dependent on Z1
geometric modeling and the optimization problem [2, 12, Lm ¼ ðDm rðtÞÞ2 ; where Dm ¼ dm =dtm ð2Þ
26, 27].
0
From a practical viewpoint, the geometric modeling
process for ships is expected to be simple and similar to the and is subject to the following constraints:
design approaches with which ship designers are already • Distance constraints: The Euclidean distance between
familiar. The principal methods of ship design include the the given data points Pi and the resulting r(t), taken at
distortion transformation of a given hull form and the the associated parameter knot ti, weighted by wi, and
ab initio design of a new hull form using form parameters. squared, must be no greater than the error tolerance e.
In the present study, distortion transformation and form In normalized form for n ? 1 data points,
parameter design based on the F-Spline technique were
investigated [6, 7]. Two hull form optimization approaches X
n

were used with practical ship designs. The applicability and A¼ fwi ðrðti Þ  Pi Þg2 e  0; e  0 ð3Þ
i¼0
subsequent improvements in hydrodynamic performance of
these techniques are discussed in this article. • End constraints: For the first and last points on the
The commercial software SHIPFLOW [13] was used as curve, tangent vectors Qi and curvature vectors Ki,
a CFD solver, and FS-Framework [1, 2, 6–9] was used as a i = 0, or n, may be given as follows:
geometric modeler and optimizer. The resistance and pro-
Type I end conditions:
pulsion performance tests were performed at the Hamburg
ship Model Basin (HSVA, Germany) and Maritime and E1 ¼ D1 rðti Þ  Qi ¼ 0 ð4Þ
Ocean Engineering Research Institute (MOERI, Korea),
Type II end conditions:
respectively.
E2 ¼ D2 rðti Þ  K i ¼ 0 ð5Þ

2 Parametric hull form design

• Area constraints: The actual area under a curve, S, shall
The process of curve generation begins with the complete match a given area value, S0:
mathematical definition of the curve from a set of given
data elements of the curve. Since the data elements usu- F ¼ S  S0 ¼ 0 ð6Þ
ally describe only a subset of the properties of the curve • Other constraints: Many other types of constraints can
shape, the results of the curve generation process may be imposed in equality or inequality form.
vary. These results depend on the choice of mathematical
curve representations and the properties of the generation This constrained optimization problem is formulated to
process [18–20]. In this study, fairness-optimized para- minimize the unconstrained objective functional I:
metric curves with constraints were used. These curves min I ¼ Lm þ kA0 þ l1 E1 þ l2 E2 þ vF; A0 ¼ A þ d 2
were represented by B-Spline curves and generated using
ð7Þ
the fairness optimization process with sets of form
2
parameters as constraints. These fairness-optimized para- where k, li, and m are the Lagrange multipliers and d is a
metric curves have been used as unconstrained transfor- slack variable. If this is an interpolation problem, d2 equals
mation functions for distortion transformation and as zero. The degree of the fairness criterion affects the char-
geometric entities to generate new hull form for the form acteristics of the shape. Table 1 shows the form parameters
parameter design. of a planar curve. Figure 1 presents an example of an

123
4 J Mar Sci Technol (2012) 17:1–17

Table 1 Form parameters

C1 C2 (circle) C3
describing a planar curve
i=0 i=1 i=0 i=1 i=0 i=1

Position Xi, Yi (m) 1, 0 0, 1 1, 0 0, 1 1, 0 0, 1

Tangent angle Qi () 90 180 90 180 90 180
Curvature Ki 4.80 4.80 1 1 -0.624 -0.624
Area S (m2) p/4/1.25 p/4 1.25 9 p/4
Centroid XC, YC (m) 0.37, 0.37 0.42, 0.42 0.49, 0.49
Fairness E2 8.72 6.09 15.14

Y The unconstrained transformation function is a mathe-

matical formulation used to define the degree of transfor-
1.0 mation of the given geometry without constraints. It can be
C3 defined in any type of function and applied to any part of
1. the geometry for a complex and flexible transformation.
C2(Circle) This function is mostly used in hydrodynamic hull form
C1 optimization because it is easy to implement. However, an
additional iteration is required to ensure that the design
requirements are met and the target geometry is achieved.
This function is, therefore, typically used for small local
variations [16, 26, 27, 30].
The constrained transformation function is a mathe-
matical formulation used to define the degree of transfor-
1.0 X mation of the given geometry with design constraints. This
function should be derived so as to satisfy design
Fig. 1 Fairness-optimized parametric curve with different area requirements before it is applied to hull form variation. The
constraints essential prerequisite to define the mathematical relation-
ship between the degree of local transformation and the
F-Spline with three different area constraints. The positions design requirements makes the application complicated.
and tangent angles are given as input form parameters. Therefore, the functions employed so far generally feature
They are the same as those of C2, the unit circle. The curve a simple form and fewer degrees of freedom [6, 12].
is generated from the point on the x-axis in a counter- The Lackenby transformation [12] is a classical distortion
clockwise direction. The properties of the generated based on a constrained transformation function expressed by
F-Spline are compared by means of the curvature, the a first or second degree polynomial. These polynomial-based
centroid and their fairness. transformation functions are analytically formulated and
applied to SAC variation to derive a new required SAC.
2.2 Distortion transformation based on fairness- Because it yields robust and effective hull form distortions,
optimized B-Spline form parameter curves this process is often used in the preliminary ship design
(F-Spline) with constraints phase. However, the Lackenby transformation has limited
flexibility because of its simple polynomial-based transfor-
Distortion is the transformation of a given hull form via mation function. The function is expressed by a few form
transformation functions to obtain new hull form data. A parameters, such as CP, LCB, LR and LE.
transformation function is a mathematical formulation that In the present study, F-Spline was used as a constrained
defines the degree of transformation of a given geometry transformation function for the SAC distortion transfor-
necessary to achieve a certain new shape. In ship design, it mation [1, 2, 6–9]. Two F-Splines were used to define the
is conventionally used to transform basic ship curves. The transformation functions for the run and entrance parts of
flexibility and variability of the distortion transformation SAC, respectively. The form parameters describing the
depend on the geometrical properties and applicability of F-Spline include the degree of changes in the position,
the transformation function. There are two types of trans- angle, and volume as well as centroid (LCB), as shown in
formation functions: unconstrained and constrained. Fig. 2. In this case, the curvature need not be considered.

123
J Mar Sci Technol (2012) 17:1–17 5

Fig. 2 SAC distortion based on

F-Spline transformation
function

The B-Spline curves that satisfy the required form considered. This constrained optimization problem can be
parameter constraints can be optimized for the second reformulated as an unconstrained problem using the
degree fairness criterion. The generation process for the Lagrange multiplier method.
transformation function for the SAC is as follows [2]: X
F ¼ E2 þ ki hi ¼ 0 ð11Þ
• Define an open uniform third degree B-Spline curve i¼1; m
with n ? 1 vertex points.
The problem is formulated to minimize the function F
  X
n and the unknowns are form parameters and Lagrange
x ðtÞ
r ðtÞ ¼ ¼ Bi  Nik ðtÞ ð8Þ multipliers. Thus, the minimum conditions are
y ðtÞ i¼0
oF oF
• Minimize the second order fairness criterion: ¼ 0; ¼ 0; i ¼ 1; . . .; m ð12Þ
oFPi oki
ZtE ( 2 2  2 2 )
d x d y This yields a nonlinear system of equations for the
E2 ¼ þ dt ð9Þ
dt2 dt2 unknown, vertex points and Lagrange multipliers that can
tB
be solved numerically.
which is subject to m-equality constraints: The ability to control the angle parameter of each part of
the SAC using the proposed approach (10) is the critical
hi ¼ FPiinitial  FPirequired ¼ 0; i ¼ 1; . . .; m ð10Þ
functionality that distinguishes it from the Lackenby
where FPi represents the form parameter constraints listed transformation. The differential form parameters expressed
in Table 2 and m is the number of constraints to be as Da in Table 2 have a significant influence on resistance

123
6 J Mar Sci Technol (2012) 17:1–17

Table 2 Form parameter

Form parameter constraints FPi
constraints for SAC
transformation function Run part Entrance part
Beginning tB = X0R End Beginning End
tE = X1R tB = X0E, tE = X1E

Position DX0R DX1R DX0E DX1E

Tangent angle Da0R Da1R Da0E Da1E
Area DCP
Centroid of area DXLCB

performance, transforming the entrance angle and stern Table 3 Basic curves describing a hull form
profile shape. Figure 2 shows an example of a SAC dis-
Curve Symbol
tortion based on F-Spline transformation function
Position Design waterline DWL
2.3 Form parameter design based on F-Spline Flat of side curve FOS
with constraints Center plane curve CPC
Flat of bottom curve FOB
An arbitrary ship hull form can be fully parametrically Deck DEC
expressed by suitably combining the form parameters Tangent angle Tangent angle at beginning and end TAB, TAE
constituting the basic curves and hull section frames. When Curvature Curvature at beginning and end CAB, CAE
the parametric design of the longitudinal basic curves is Area Sectional area curve SAC
generated using fairness-optimized parametric curves, Centroid of area Vertical moments of sectional area VMS
transversal basic curves can be generated. Similarly, if the Lateral moments of sectional area LMS
transversal sections are derived by parametric modeling, a
set of surfaces that interpolate or closely approximate the
design sections is generated. Thus, efficient and effective The parametric modeling was based on the given shape
form variations based on the parametrically generated hull by importing the IGES file. Based on the IGES surface, a
form surfaces are possible [6, 18–20]. However, since there dense offset distribution for the forebody and aftbody was
are many form parameters that make up each form generated as offset data. The form parameter value
parameter basic curve, and since there are possible corre- describing each basic curve of the initial hull form was
lations among the form parameters, it is not straightforward extracted from the given initial hull form to begin the
to determine the parametric input data sets. The increasing parametric modeling. The number of form parameters is
flexibility of the basic curve designs makes the design strongly dependent on the shape variation strategy. To fully
process increasingly complex. exploit the F-Spline optimization and develop a completely
Basic curves can be parameterized using the F-Splines. new type of forebody hull form, approximately 80 form
Since the curves are geometrically optimized for fairness to parameters were used for the parametric modeling. The
meet the least number of curve requirements, the use of relationships among the parameters were investigated in
F-Spline reduces the complexity arising from the parameter depth and sharing points were identified. A well-developed
input sets, while increasing flexibility and improved shape structure of parameter dependencies can help to avoid
quality [1, 2, 6, 7]. unnecessary increases in parameter inputs and degenera-
In this study, the forebody hull form of an LPGC was tion in hull form variation.
fully parameterized for the shape variation, whereas the
aftbody hull form was kept unchanged during the optimi-
zation. The parametric model of the forebody hull form 3 SAC optimization of an ultra-large containership
includes global parameters such as ship dimensions and based on a constrained F-Spline transformation
basic curves indicating the longitudinal distribution of the function
geometric parameters such as position, tangent, curvature,
and area. Table 3 shows basic curves to define the essential The size of containerships has been continuously increas-
main shape characteristics of the ship. Figure 3 shows the ing. With the expansion plan for the new Panamax, the
fully parametric forebody model based on several basic largest size of operating containerships will reach around
curves and the form parameters defining those basic curves 14,000 TEU. Containerships of approximately 18,000 TEU
[9]. are also currently under development. The block coefficient

123
J Mar Sci Technol (2012) 17:1–17 7

Fig. 3 Fully parametric forebody model and form parameters defining the basic curves

CB reaches 0.7, and the design speed range decreases to resistance performance. Second, a SAC optimization
approximately 0.2 in FN or even lower. The design concept problem was formulated based on the second baseline SAC
for ultra-large containerships tends to make them larger, design, and the feasible form parameter range was deter-
fuller and slower. They can facilitate improvements in ship mined using the sensitivity analysis. A formal optimization
fuel economy and help to create ‘‘greener’’ ships. was carried out using a search method. The performance of
Another design trend to improve the fuel economy is the the optimal ship was tested and the results discussed. This
multi-objective shape optimization based on the life cycle analysis had two main objectives:
operating conditions. The speed performance in the ballast
(a) To verify whether the design variations yielded by the
draft condition has become increasingly important. How-
F-Spline transformation functions were geometrically
ever, it has still been a standard process for containership
flexible and functional and could be conveniently
design in ship yards that the hull form design is conducted
integrated into the classical design practice.
only for the performance in the design draft condition, and
(b) To verify whether the design variations yielded by the
then the speed performances are generally simulated and
F-Spline transformation functions were acceptable in
tested at two drafts, design and ballast conditions. This
terms of resistance performance.
section describes a procedure used to optimize the hydro-
dynamic hull form for an ultra-large containership. Optimal
hull form has been developed only for the performance in 3.1 Design configurations
the design draft condition.
A SAC provides an effective and simple description of The design configurations of the object ship are shown in
global geometric properties. At the same time, it is closely Table 4. A baseline design that satisfies the main require-
related to the resistance and propulsion performance of a ments was developed. It had a relative V-type forebody
ship. From this point of view, the ship hull form distortion hull and a U-type aftbody hull.
approach based on SAC transformation is one of the most
effective global design methods for the preliminary design 3.2 CFD simulation
stage. The optimization procedure was performed using the
SAC transformation with F-Spline as the transformation The wave profiles and wave making resistance associated
function. First, a sensitivity analysis was performed to with alternative ship hull designs were simulated using
determine the effect of each form parameter of F-Spline on SHIPFLOW XPAN by FLOWTECH. SHIPFLOW pro-
the geometric properties of SAC transformations and hence vides a non-linear potential flow code that uses Rankine
on ship variations and wave resistance performance. For panels to discretize the hull and free surface. The nonlinear
example, the SAC angle parameters influence the entrance boundary condition was iteratively solved. Trim and sink
angle and incident angle of the design waterline, the full- were allowed during the simulation.
ness around the parallel middle body and the transom. The number of panels on the hull and free surface
These shapes have a considerable influence on wave depends on the fullness and complexity of the hull

123
8 J Mar Sci Technol (2012) 17:1–17

Table 4 Principal particulars of compare wave resistance levels only during the optimiza-
Design
an ultra-large containership tion process. However, the wetted surface area is generally
condition
monitored during the entire simulation-based optimization
FN 0.20
process.
CB 0.70
LPP (m) 360.0
3.3.1 The effect of the parallel middle body (PMB)
B (m) 51.0
on the wave resistance
T (m) 14.0
The effect of PMB length and position on wave resistance
performance was investigated during this study. The length
geometry and ship speeds mainly. A detailed verification of the PMB, i.e., LM, in the baseline design is defined as the
process including grid sensitivity analysis and error esti- distance between the starting position for PMB, i.e., X0R
mation for the numerical predictions were undertaken for a and the end position for PMB, i.e., X0E. LM is expressed as
wide range of ship series and ship speeds using SHIP- a percentage of LPP, whereas X0R and X0E are expressed as
FLOW. In this thesis, sensitivity analyses for panel con- the station number. The parameters LM, X0R and X0E were
vergence, computational range convergence and iterative systematically varied by DX0R and DX0E. They were
convergence were performed [15]. The numerical errors applied to each position X0R and X0E in the baseline design
that propagated in the wave resistance coefficient were and varied by intervals of 2.5% LPP (0.5 station; see
estimated. The panel convergence analysis determined hull Fig. 10) respectively. The variation is shown in Table 5:
panel variation in terms of size, aspect ratio and distribu- Case 1 represents the baseline design. LM is varied from
tion as well as free surface panel variation in terms of size 7.5% LPP to 15.0% LPP by intervals of 2.5% LPP. An
per wavelength, computation region on the free surface, increase in LM leads to a sharper entrance angle and a
aspect ratio, panel distribution near the hull. The results thinner stern volume because of the constant volume con-
indicated that there exist systematic error trends between straint. Figure 4 shows the SAC transformations as a
simulations and experiment as a function of ship geometry function of LM from Case 1 to Case 4.
and ship speeds. Using a potential code based on the Figure 5 shows the percentage reduction in CW in
SHIPFLOW analysis, a simulation-based design can be relation to that of the baseline design. The CW decreases
created using a comparison of the performance of alter- as LM increases. The CW reduction rate reaches a maxi-
native ship designs of the numerical configuration is mum of 8% with a 15% LPP increase in LM. LM has a
carefully determined. The degree of numerical error based greater influence on the decrease in CW than the variation
on the potential code is somewhat small and randomly in X0R. Case 4, in which the PMB is positioned extremely
distributed in comparison with the effect of geometric ship in the forebody direction, exhibits a diminishing rate of
variation. decrease in CW even with the increases in LM. Case 11 is
The hull surface was discretized using around 2,000 positioned extremely in the aftbody direction. Considering
panels. On the longitudinal plane, a resolution of 25 panels that resistance performance is based on potential flow, we
per fundamental wavelength was utilized. The computation limit X0E from Station 8.0 and extend X0R to Station 11.0.
region of the free surface extended from x/L = -0.5 to 2.0 Case 9, with an LM of 12.5% LPP from Station 8.0 to
upwards and downwards, respectively, and from y/L = 0.0 Station 10.5, exhibits the most favorable resistance
to 0.7 from side to side. To reduce oscillation, the aspect performance.
ratio of free surface panel was maintained at levels larger
than 1.5. 3.3.2 The effect of the differential form parameters
on wave resistance
3.3 Sensitivity analysis of F-Spline form parameters
A systematic investigation of the differential form param-
A systematic investigation of the SAC transformation was eters Da0E and Da1E for the entrance part and Da0R, Da1R
performed with respect to the form parameters for the for the run part of SAC was carried out. Four differential
F-Spline transformation functions for the wave resistance form parameters were systematically varied based on six
performance. Wave resistance is not representative of cases using intervals of 15 (D). Therefore, the hull form
overall ship performance. However, because the sensitivity variations based on the SAC transformation and wave
analysis including further formal optimization was based resistance calculations were determined for all 24 cases.
on the generalized Lackenby method, the shape topology The positive change in the angle is expressed as D, 2D, 3D
and characteristics of the ship were not significantly at AP and FP, Da1R and Da1E make the SAC sharper at
changed during the process. Thus, it was possible to each curve end and for a parallel middle angle, Da0R and

123
J Mar Sci Technol (2012) 17:1–17 9

Da0E increase the fullness near the area. The design 3.4 SAC optimization
waterline shapes as a function of the entrance angle Da1E
are shown in Fig. 6. The geometrical properties of SAC variation using F-Spline
As shown in Fig. 7, there is a linear functional rela- form parameters and the design trends for minimum wave
tionship between the differential parameters of the SAC resistance were investigated in this study. In the next step,
and the wave resistance performance. The wave resistance based on the previous sensitivity analysis, the SAC optimi-
performance improved with increased angle variation. The zation problem was formulated by setting a feasible range of
effect of the entrance angle Da1E on wave resistance form parameters as constraints. The first design variation was
performance is noteworthy. An improvement of approxi- used as the second baseline design for the formulated SAC
mately 15% in wave resistance performance was found optimization process. This problem was solved using the
for that angle as compared with the baseline design. The tangent search method as an optimization technique.
fullness of the PMB based on the changes of Da0R and Objective function: Minimize Cw
Da0E is related to wave resistance performance. The
Design variables:DXC ; DX0E ; Da0E ; Da1E ; DX0R ; Da0R
degree of steepness of the SAC was quantitatively
determined based on the combination of the differential Constraints:
form parameters. We selected the first design variation  1  DXC ð%LPP Þ  1
based on the main parameter sensitivity analysis by  6:175  DX0E ðmÞ  6:175
adding the best PMB configuration and the differential  6:175  DX0R ðmÞ  6:175
form parameter configuration. The ship design was
 20  Da0E ðdegreeÞ  30
derived by independently applying Case 9 for PMB var-
iation and the 2D degree of change in the entrance angle  10  Da1E ðdegreeÞ  10
variation, Da1E (2D) to the baseline design. The degree of  10  Da0R ðdegreeÞ  10
improvement of the first design variation in the wave
An optimal design was achieved after 37 iterations within
pattern is evaluated with respect to the baseline design in
the given convergence criterion. The wave resistance
Fig. 8.
Through the sensitivity analysis, the hull form with the
fuller shape around the middle body and the sharper
Black: Case 1
shape around FP and AP, namely the hat-shaped SAC, Red: Case 2
was found to be preferable. Although this trend complies Blue: Case 3
with the slowly operating full ship design concept, this Green: Case 4

optimal SAC shape is valid for the ultra-large container-

ship with the present design conditions listed in Tables 4
and 6.
Fig. 4 SAC variations as a function of LM

The Effect of Parallel Middle Body on Wave Resistance

Table 5 Variation cases for PMB 0
Parameters PMB DCW
(%) X0R=ST.9.0
Case DX0R DX0E Start End Length X0R=ST.8.5
-2 X0R=ST.8.0
X0R - X0E ? DX0E LM (%
X0R=ST.7.5
DX0R (station) LPP)
% reduction in CW

(station)
-4
1 0 0 9.0 10.0 5.0
2 0.5 10.5 7.5 -2.8
3 1.0 11.0 10.0 -4.1 -6
4 1.5 11.5 12.5 -4.5
5 0.5 0 8.5 10.0 7.5 -2.6
6 0.5 10.5 10.0 -5.3 -8

7 1.0 11.0 12.5 -6.5

8 1.0 0 8.0 10.0 10.0 -5.2
-10
9 0.5 10.5 12.5 -6.8 5 7.5 10 12.5 15 17.5

10 1.0 11.0 15.0 -8.2 LM(%/L PP )

11 1.5 0 7.5 10.0 12.5 -6.2
Fig. 5 Percent reduction rate of the CW

123
10 J Mar Sci Technol (2012) 17:1–17

The Effect of Angle Parameter on Wave Resistance

20

-1Δ , -2 Δ , -3Δ Δα1R

15
Δα0R
Δα0E
10 Δα1E

% reduction in C W
1 Δ , 2Δ, 3Δ 5

-5
Fig. 6 Design waterline shape as a function of the entrance angle
-10
performance was slightly improved but remained similar to
that of the hull form in the second baseline design, with a -15
smaller entrance angle and small bulb. The PMB of the
optimal design increase slightly. In the low speed full -20
-4 -3 -2 -1 0 1 2 3 4
containership design range, the effect of the LCB position Angle Variation Δ
was relatively small within a certain variation range. A
reduction in the entrance angle of the SAC resulted in a very Fig. 7 CW with respect to the tangential angle variation of SAC
narrow waterline angle at the FP. The entrance angle of the (entrance and run)
optimum SAC was mainly influenced by the incident angle of
the design waterline when a bulb shape was given. 0.01
Figure 9 shows the wave contours of the optimal design
(top) and the baseline design (bottom) at a design speed of
Wave Elevation / L

24 knots. Figure 10 shows an experimental setup for a

resistance test using the optimal design at the HSVA towing
tank. The wave patterns seen in the numerical simulations
0.00
and model tests were found to be favorable and very stable,
respectively. An additional comparison is shown in Fig. 11
that compares the experimental total resistance of the
derived ship with that of ships tested at the HSVA. The total
resistance of the derived ship was extrapolated from the Initial Hull Form
Optimized Hull Form
model test results according to the power prediction meth- -0.01
-0.5 0.0 0.5 1.0 1.5
ods of the HSVA. The other ships compared in Fig. 11 are X/L
from HSVA database; they have similar characteristics for
CB, L/B and B/T and are scaled to the similar displacement. Fig. 8 Comparison of the wave profiles of the baseline design and
The derived ship shows excellent resistance performance at first design variation
the design speed of 24 knots.
During the optimization process, the major part of the
Table 6 Form parameters used
improvement was achieved during the first step via sys- for F-Spline variation
F-Spline form
tematic parameter variation. This trend depended on the set parameter
up of the optimization problem. In this investigation, the Entrance DX0E, DX1E (station)
first step was to conduct a Pareto search for optimal designs Da0E, Da1E ()
based on a small number of systematic analyses in a wide Run DX0R, DX1R
range of design spaces. Generally, this step is formulated (station)
by optimization strategy such as DoE. The selected designs Da0R, Da1R ()
were also used as a starting point for deterministic opti-
mization in the second step. The final optimum design was
associated with a relatively small Pareto range. shifting was possible in each section; local variations in the
In the present study, an ultra-large containership hull section frame and variations in the bulb shape were not
design was developed using an SAC optimization based on an possible. Nevertheless, the simplicity and efficiency of this
F-Spline transformation function. This hull form showed process are sufficiently valuable to support the use of this
good resistance performance although limited longitudinal technique in the preliminary hull form design stage.

123
J Mar Sci Technol (2012) 17:1–17 11

Optimized Hull Form

Initial Hull Form

Fig. 9 Wave contour of two hull forms (optimal design baseline design) at 24.0 knots

4 Full parametric hull form optimization of an LPG

carrier
20
18 16 14 12 10
An LPG carrier is a tanker ship that is designed to transport
liquefied petroleum gas. The hydrodynamic hull form
design of LPG carriers has generally been considered to
involve particular requirements because of the operational
peculiarities and geometric limitations of these ships. One
such peculiarity is that an LPG carrier operates evenly in Fig. 10 Model test of the optimized hull form at 24.0 knots
the three draft conditions, i.e., the design, ballast, and
hydrodynamic properties. Table 7 shows the principal
scantling conditions. Therefore, the speed performance
characteristics of the target ship.
under all draft conditions should be considered at the same
time. Similarly, one geometric limitation of the design of 4.1 Problem statement for forebody hull form
these ships is that the LPG tank must be positioned in the optimization
most forward part of the ship. This constraint makes it
difficult to design a hydrodynamically superior waterline The forebody hull form optimization was formulated to
shape and results in increase of wave resistance. minimize the full scale total resistance in the ballast draft
In this section, a hydrodynamic optimization problem is condition while maintaining the total resistance in the design
presented for an existing LPG carrier. The baseline design condition and satisfying the geometric constraints. The goal
required a performance improvement in the ballast draft of the optimization was to reduce delivery power (DHP) by
condition as the model already showed good hydrodynamic 3% in the ballast draft condition while maintaining perfor-
performance in the design draft condition. The aim was to mance in the design draft condition. The goal was for the new
develop a new hull form for an LPG carrier with superior hull form to be compliant with breadth constraints at four
performance in the design and ballast draft conditions. This frames with nine points for the LPG tank installation. The
design problem was complicated not only because it was total displacement change was not to exceed ±3%.
necessary to set up a constrained multi-objective optimi-
• Minimize
zation problem but also because of the volume of shape
variations involved in investigating a new geometric con- The total resistance, RT, in the ballast draft condition
cept. The goal of the optimization is to improve the speed
• Subject to:
performance in the ballast draft condition without
decreasing the speed performance in the design draft con- Inequality constraint: RT in the design draft condition
dition. The forebody hull form was fully parametrically Equality constraint: Keeping breadth at four frames
modeled to analyze diverse hull shapes with different Range constraint: Displacement change within ± 3%

123
12 J Mar Sci Technol (2012) 17:1–17

Fig. 11 Comparison of the total

resistance of optimized hull
form with the HSVA database

• Objective function Table 7 Principal particulars for the LPG carrier

The total resistance globally reflects the ship perfor- Design condition Ballast condition
mance as a function of the hull form. The total resistance Vs (knots) 16.2 16.5
has been utilized as a typical criterion for evaluating FN 0.191 0.194
hydrodynamic ship performance. In this study, the total CB 0.75 0.70
resistance at the full scale was compared for the perfor- T (m) 11.0 6.0/8.0
mance evaluation because the scale effect for the model
scale ship is not systematic when the performance in two
different draft conditions or that of completely different DRT denotes an additional resistance component, and RAA
ship shapes is considered. The total resistance at full scale represents the wind resistance. The additional resistance
was formulated as components RA, DRT, and RAA were estimated at each draft
RT ¼ RWP þ ð1 þ kÞRFITTC þ RA þ DRT þ RAA ð13Þ conditions based on the experience with similar ship hulls.
These components were applied to each design variant as a
where RWP is the wave pattern resistance derived from a function of wetted and exposed surface area respectively.
wave cut analysis of the wave pattern computed by For the final deterministic optimizations, a mixed
SHIPFLOW, and RFITTC is the frictional resistance objective function was introduced that represented the
according to the ITTC friction line, weighted sum of the total resistance ratios in the ballast and
0:075 design draft conditions.
CFITTC ¼ ð14Þ The optimization functional was defined as
ðlog10 RN  2Þ2
2 1
The form factor k of the baseline hull form was F ¼ RTratioðBÞ þ RTratioðDÞ þ HPPenalty þ rPenalty
3 3 ð15Þ
estimated using the correlation between that by Holtrop RTratio ¼ RTðNewVariantÞ =RTðBaselineÞ
method [11] and measurement tested by MOERI in the
design and ballast draft conditions. It yielded In Eq. (15), the total resistance ratio RTratio is the ratio of
(1 ? k) = 1.229 for ballast and (1 ? k) = 1.209 for the total resistance of the new variant hull to the total
design draft of the baseline hull form. During the resistance of the baseline design. RTratio(B) indicates the
optimization process, the form factor was recomputed for total resistance ratio in the ballast draft condition and
each design variant based on the empirical form factor [11] RTratio(D) indicates the total resistance ratio in the design
that is a function of LPP, B, T, LR, CB, CP and LCB, which draft condition. The total resistance ratios in both
yielded small adjustments. RA is a correlation allowance, conditions were considered in the objective function by

123
J Mar Sci Technol (2012) 17:1–17 13

introducing the corresponding weights. The additional

terms HPPenalty and rPenalty are penalty functions which
were formulated to account for hard-point and volume
constraints. These functions are expressed in Eqs. (16) and
(17). The mixed objective function F was introduced to
solve the unconstrained deterministic optimization problem
by converting the explicit constraints to implicit
corresponding exterior penalty functions.
• Constraints
LPG tank hard-points: There were nine hard-point
constraints that were considered for the LPG tanks. In
Fig. 12, their locations within the forebody of the baseline Fig. 12 Hard-point constraints for LPG tank installation
design are shown. The hard-point requirements were the
leading constraints and, hence, the restricting constraints in
the optimization. To address the hard-point constraints in with dr representing the change in the displacement vol-
the deterministic optimization, the constraints were for- ume and mr as a user specified positive scaling factor. As a
mulated as in Eq. (16) and replaced by penalty terms in the result, the penalty became active only if the decrease in the
objective function. The hard-point penalties HPi penalty displacement volume exceeds the margin of 750 m3. No
were formulated in terms of normalized exterior penalty penalty was used for the upper displacement bound.
functions in a least square sense as
!
X 9 4.2 CFD simulation
HPPenalty ¼ mHP HPzeroOffset þ HPiPenalty
i¼1 For the performance assessment the potential code
with ð16Þ XPAN SHIPFLOW [13] was used to compute non-linear
8 2 9
< dHPi = ship wave systems with trim and sink. The SHIPFLOW
HPiPenalty ¼ Beam=2  100 if dHPi \0 configuration is similar to those described for ultra-large
: ; containerships in Sect. 3.2. The hull surface was dis-
0 otherwise
cretized using 1,940 panels. The free surface domain
dHPi denotes the shortest 1-norm distance from the extended one ship length aft of the stern and to the side
frame to its respective hard-point i, and mHP is a user and, included 4,025 panels in total. The geometric inputs
defined positive scaling factor. A penalty becomes active to SHIPFLOW were based on an offset representation of
only if the respective frame lies inside its hard-point the ship hull surface. Figure 13 shows grouped (differ-
definition. To amplify the penalty in the infeasible domain ently colored) SHIPFLOW offsets for the baseline LPGC
close to the constraint bound, a constant zero offset hull.
HPzeroOffset was added to the penalties if and only if one or
more hard-point constraints actually became active: i.e., 4.3 Optimization procedure
dHPi \ 0 (for i = 1,…, 9). Finally, all hard-point penalties
were added together and added to the mixed objective. Two step optimizations were performed. In the first step, a
Displacement volume: The change in the displacement systematic design space exploration was conducted based
volume was required to remain within a margin of on the Sobol algorithm using the design of experiments
±750 m3. Similar to the hard-point constraints in the final (DoE) procedure [7, 25]. The primary purpose of analyzing
deterministic optimization runs, the displacement con- the design space was to detect reasonable starting points for
straint was introduced and then fully replaced by a penalty deterministic optimization. Throughout the optimization
term that was added to the mixed objective function. The process, both the ballast and the design draft were moni-
penalty was formulated as a rate of increase in the dis- tored for the targeted speeds, respectively. Through the
placement volume as compared with that of the baseline design space analysis, 2,470 designs were generated in
design with an exterior penalty factor as follows: total. The 946 feasible designs that fulfilled the necessary
(  2
drþ750m3 constraints are shown in Fig. 14. This figure shows the
rPenalty ¼ m r r  100 if dr [ 750 m3
Baseline designs that were successfully analyzed by the Sobol
0 otherwise algorithm in the generalized design space, which, for N free
ð17Þ variables, maps the multidimensional design space from RN

123
14 J Mar Sci Technol (2012) 17:1–17

into the domain of the generalized design variables X and A comparison the total resistance achieved via the CFD
Y: i.e., R2 with simulation with that indicated in the EFD results yields
vffiffiffiffiffiffiffiffiffiffiffiffiffi vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi some discrepancies. The improvement in the resistance
u N=2 u X
uX u N performance of the optimal ship is much greater in the EFD
Xnorm ¼ t fi ; Ynorm ¼ t
2 fi2 ð18Þ
i¼1
case than in the CFD case. This still holds given the
i¼N=2þ1
experimental and computational uncertainties involved.
fi denotes the value of the free variable i mapped into the The new bulb may result in favorable wave-viscous inter-
normalized domain of its lower and upper bound [0..1] [8]. action not only by reducing the breaking waves but also by
The feasible designs were further analyzed by means of influencing the stern wave and boundary. The scale effect
CFD. Based on the design space exploration, deterministic extrapolated from model to ship, in which the form factor
optimizations from different start designs were performed. must be used, should be carefully considered. In any case,
Figure 15 shows the range of improvement in the total
resistance in two draft conditions at the same time. The 2.5

designs in the bounded square are Pareto optimal with

minimum total resistance in both the ballast and design
draft conditions. These studies indicate that the bulb shape 2.0
and the incident angle of the waterlines influence the wave
resistance performance in the ballast draft condition. The

Y-Norm
hull form variations are focused on this area and the second
stage of deterministic optimization was restarted. Figure 16 1.5

shows part of convergence history for the objective func-

tions in the ballast draft condition.
1.0
4.4 Optimum hull form and performance

The final optimum hull form is shown in Fig. 17. The

forebody profile and bulb shape are remarkably different 0.5
0.5 1.0 1.5 2.0 2.5
from those of the conventional LPG ship. They are similar
X-Norm
to those of a low speed container ship. The volume dis-
tribution around the ballast draft and bottom area was Fig. 14 Generalized design space X over Y
considerably reduced, whereas the volume near the design
101.0
draft increased. All geometrical constraints were satisfied.
The wave contour in the ballast draft condition is shown
R T (16.2kn,TDes ) / R T Baseline Des

100.8
in Fig. 18 for the optimal ship. A decrease of 1.75% in the
total resistance in the ballast draft condition and a decrease
100.6
of 0.98% in the design draft condition were achieved using
SHIPFLOW simulation. The wetted surface area S was
100.4
substantially reduced in both conditions. The selected
design features a smaller bulbous bow and a finer entrance
100.2
angle that is expected to decrease the wave breaking in the
ballast draft condition at around 16.5 knots. The resistance Baseline
and propulsion tests were performed in the towing tank at 100.0 FSobol001_253

MOERI, Korea. The performance of the two hull forms is

compared for two draft conditions in Table 8. In the ballast 99.8
99.0 99.5 100.0 100.5 101.0 101.5 102.0 102.5
draft condition, the optimum hull form shows a 5.7% R T (16.5kn,T Ball ) / R T Baseline Ball
improvement in the total resistance RTS and a 7.8%
improvement in delivery power. Fig. 15 Definition of the feasible designs

Fig. 13 SHIPFLOW offset of

the LPGC baseline hull

123
J Mar Sci Technol (2012) 17:1–17 15

this type of forebody hull form is favorable for propulsion functions and for form parameter design as fully parametric
and resistance performance. models. This design process can be considered as a relevant
design method in the preliminary design stage for the fol-
lowing reasons:
5 Conclusions
1. Since F-Spline is automatically generated via the
optimization process for fairness to satisfy the given
This study involved hydrodynamic hull form optimizations
requirements, a fair curve is always generated even
based on parametric models generated by F-Spline. Two
with a few subsets of form parameters available to
functionalities of F-Spline in the context of ship design
satisfy the given requirements. This curve can, there-
were validated through this research. They were used to
fore, be easily used in the complicated form generation
evaluate shape distortion as constrained transformation
and form transformation processes.
2. Hull form distortion based on the constrained trans-
100.0
formation function F-Spline is simple, flexible, and
well suited to the optimization process. In particular,
R T(16.5kn, TBall) / R T Baseline Ball

the implementation of differential form parameters for

99.5 the SAC transformation remarkably improved the
availability of the hull form distortion by transforming
the entrance angle and stern profile shape, which are
99.0 significant factors in resistance performance. An ultra-
large container ship optimized using this process
showed excellent resistance performance at the design
speed of 24 knots.
98.5
3. The fully parametric ship hull design offers the
flexibility to generate new types of hull forms. This
holds even if it is still difficult to model and control the
98.0 ship hull because of its complex structure. This design
0 50 100 150 200
Number of Variants approach was used in forebody hull form optimization
for an LPG carrier. The optimal ship with a completely
Fig. 16 Convergence history plot of the objective function in the
ballast condition
different bulb shape was successfully validated,

Fig. 17 Comparison of the

optimum forebody hull form
(blue, small bulb) and initial
hull form (red, large bulb)
(color figure online)

123
16 J Mar Sci Technol (2012) 17:1–17

Optimized Hull Form

Initial Hull Form

Fig. 18 Comparison of wave contours

Table 8 Improvements in performance characteristics Open Access This article is distributed under the terms of the
Creative Commons Attribution Noncommercial License which per-
Design condition (% diff.) Ballast condition (% diff.) mits any noncommercial use, distribution, and reproduction in any
medium, provided the original author(s) and source are credited.
S -1.06 -1.20
r -0.36 -0.99
RTS 0.74/-0.98 (CFD) -5.70/-1.75 (CFD)
RTM -0.99 -4.22
References
PD -0.50 -7.79
CR 5.38 -10.08 1. Abt C, Harries S (2007) A new approach to integration of CAD
CTS 1.81 -4.54 and CFD for naval architects. In: Sixth international conference
on computer applications and information technology in the
% diff. = (optimized-initial)/initial 9 100 maritime industries (COMPIT), Cortona
2. Birk L, Harries S (eds) (2003) OPTIMISTIC—optimization in
showing a 5.7% improvement in total resistance and a marine design, Mensch & Buch Verlag, Berlin, ISBN 3-89820-
514-2
7.8% improvement in delivery power. 3. Campana EF, Liuzzi G, Lucidi S, Peri D, Piccialli V, Pinto A
4. The flexibility and functionality of geometric modeling (2009) New global optimization methods for ship design prob-
constitutes an important technology in hydrodynamic lems. Optim Eng 10:533–555
optimization, not only because it is directly related to 4. Campana EF, Peri D, Tahara Y et al (2006) Shape optimization in
ship hydrodynamics using computational fluid dynamics. Comput
the performance of the optimal hull form but also Methods Appl Mech Eng 196:634–651
because it is related to the effectiveness and automa- 5. Campana EF, Peri D, Tahara Y (2004) Comparison and valida-
tion of the optimization process. tion of CFD based local optimization methods for surface com-
batant bow. In: 25th symposium on naval hydrodynamics,
St. John’s, Newfoundland and Labrador, Canada, August 8–13
Acknowledgments This research was supported by World Class 6. Harries S (1998) Parametric design and hydrodynamic optimi-
University (WCU) program through the National Research Founda- zation of ship hull forms. Ph.D. Thesis, Institut für Schiffs-und
tion of Korea funded by the Ministry of Education, Science and Meerestechnik, Technische Universität Berlin, Germany; Mensch
Technology (R31-2008-000-10045-0). The author would like to thank & Buch Verlag, Berlin
Prof. K.C. Park of University of Colorado at Boulder for his sincere 7. Harries S, Valdenazzi F, Abt C, Viviani U (2001) Investigation
advice and encouragement and Dr. Stefan Harries of Friendship on optimization strategies for the hydrodynamic design of fast
Systems GmbH for his long and constant friendship. ferries, FAST’01, Southampton

123
J Mar Sci Technol (2012) 17:1–17 17

8. Harries S, Abt C, Hochkirch K (2004) Constraint management for 23. Pérez F, Suárez JA, Clemente JA, Souto A (2007) Geometric
marine design applications. In: International symposium on modeling of bulbous bows with the use of non-uniform rational
practical design of ships and other floating structures (PRADS B-Spline surfaces. J Mar Sci Technol 12:83–94
2004), Lübeck-Travemünde, Germany, September 2004 24. Peri D, Campana EF (2005) High-fidelity models and multi
9. Harries S, Friendship Systems (2007) Hydrodynamic optimiza- objective global optimization algorithms in simulation-based
tion of the DSME 60K LPG Carrier, FS-Report 175-02-01 design. J Ship Res 49:159–175
10. Hinatsu M (2004) Fourier NUBS method to express ship hull 25. Press WH, Flannery BP, Teukolsky SA, Vetterling WT (1988)
form. J Mar Sci Technol 9:43–49 Numerical recipes in C. Cambridge University Press, New York
11. Holtrop J (1984) A statistical re-analysis of resistance and pro- 26. Saha GK, Suzuki K, Kai H (2004) Hydrodynamic optimization of
pulsion data. Int Shipbuild Progress 31(363):272–276 ship hull forms in shallow water. J Mar Sci Technol 9:51–62
12. Lackenby H (1950) On the systematic geometrical variation of 27. Saha GK, Suzuki K, Kai H (2005) Hydrodynamic optimization of
ship forms. Trans INA 92:289–316 a catamaran hull with large bow and stern bulbs installed on the
13. Larsson L (1997) SHIPFLOW user’s manual and theoretical center plane of the catamaran. J Mar Sci Technol 10:32–40
manual, FLOWTECH Int. AB, Gothenburg 28. Sariöz E (2006) An optimization approach for fairing of ship hull
14. Lee YS, Kim SY, Kang KT (1995) Hull form generation by using forms. Ocean Eng 33:2105–2118
TSK fuzzy model. Fuzzy logic and its application, information 29. Stern F, Wilson RV, Coleman HW (2001) Comprehensive
science and intelligence systems approach to verification and validation of CFD simulations—part
15. Lee YS (2003) Trends validation of CFD predictions for ship 1: methodology and procedures. J Fluid Eng 123:793–802
design purpose. Ph.D. Thesis, Institut für Schiffs-und Meeres- 30. Tahara Y, Stern F, Himeno Y (2004) CFD-based optimization of
technik, Technische Universität Berlin, Germany a surface combatant. J Ship Res 28:159–175
16. Maisonneuve JJ, Harries S, Marzi J, Raven HC, Viviani U, Piippo 31. Tahara Y, Wilson RV, Carrica PM, Stern F (2006) RANS sim-
H (2003) Toward optimal design of ship hull shapes. In: 8th ulation of a container ship using a single-phase level-set method
international marine design conference, IMDC03, Athens with overset grids and the prognosis for extension to a self-pro-
17. Min KS, Lee YS, Han BW, Kang SH (2004) Hull form optimi- pulsion simulator. J Mar Sci Technol 11:209–228
zation for fine higher-speed ships. In: 25th symposium on naval 32. Tahara Y, Peri D, Campana EF, Stern F (2008) Computational
hydrodynamics, St. John’s, Newfoundland and Labrador, Canada, fluid dynamics-based multiobjective optimization of a surface
August 8–13 combatant using a global optimization method. J Mar Sci Technol
18. Nowacki H, Creutz G, Munchmeyer FC (1977) Ship lines crea- 13:95–116
tion by computer—objectives, methods, and results, symposium 33. Valdenazzi F, Harries S, Viviani U, Abt C (2002) Seakeeping
on computer-aided hull surface definition. Annapolis, MD optimisation of fast vessels by means of parametric modeling,
19. Nowacki H, Boor MIG, Oleksiewicz B (1995) Computational HSMV02, Naples
geometry for ships. World Scientific, Singapore 34. Valorani M, Peri D, Campana EF (2003) Sensitivity analysis
20. Nowacki H, Kaklis PD (eds) (1998) Creating FAIR and SHAPE- methods to design optimal ship hulls. Optim Eng 4, 337–364
preserving curves and surfaces. B.G. Teubner, Stuttgart 35. Wilson RV, Stern F, Coleman HW (2001) Comprehensive
21. Papanikolaou A, Zaraphonitis G, Boulougouris E, Langbecker U, approach to verification and validation of CFD simulations—part
Matho S, Sames P (2010) Multi-objective optimization of oil 2: RANS simulation of a cargo/container ship. J Fluid Eng
tanker design. J Mar Sci Technol 15:359–373 123:803–810
22. Pérez-Arribas F, Suárez-Suárez JA, Fernández-Jambrina L (2006)
Automatic surface modeling of a ship hull. Comput Aided Des
38:584–594

123

View publication stats