You are on page 1of 9

856

2014,26(6):856-864
DOI: 10.1016/S1001-6058(14)60094-9

Hydrodynamic optimization of a triswach*


YANG Chi, HUANG Fuxin
School of Physics, Astronomy and Computational Sciences, George Mason University, Fairfax, VA 22030, USA,
E-mail: cyang@gmu.edu
KIM Hyunyul
NAVATEK, South Kingstown, RI 02879, USA
(Received November 29, 2014, Revised December 14, 2014)
Abstract: A new methodology for hydrodynamic optimization of a TriSWACH is developed, which considers not only the positions
of the side hulls but also the shape of the side hulls. In order to account for the strong near-field interference effects between
closely-spaced multihulls, an integrated hydrodynamic computational tool that consists of a potential-flow based simple CFD tool
and an Euler/RANS/Navier-Stokes based advanced CFD tool has been further developed and integrated into a practical multiobjective hydrodynamic optimization tool. The other components of this hydrodynamic optimization tool consist of a hull shape
representation and modification module and an optimization module. This enhanced multi-objective hydrodynamic optimization tool
has been applied to the hydrodynamic design optimization of the TriSWACH for reduced drag by optimizing the side hulls only. A
new methodology is developed to optimize side hull forms so that the TriSWACH has a minimal drag for a wide speed range and for
various side hull positions. Two sets of the side hulls are developed and used for the design of two optimal TriSWACH models.
Model tests are carried out for two optimal TriSWACH models at Webb Institute for validations. Substantial drag reductions have
been obtained for a wide range of speed.
Key words: hydrodynamics optimization, drag reduction, CFD, trimaran, surface modification

Introduction
The TriSWACH is a novel Trimaran that has a
Small Waterplane Area Center Hull (TriSWACH) and
two small side hulls. There are several advantages of
using TriSWACH, such as inherently good seakeeping
performance at small displacement, good intact stability, large usable deck area compared to monohulls,
and small installed power compared to the Small
Waterplane Area Twin Hulls (SWATHs)[1]. However,
flow characteristics of the TriSWACH differ substantially from monohull ships. Thus, it is very important
to understand the flow features around hull forms, the
influence of the side hull configurations, and the impacts of the hull forms on its performance when designing and optimizing a TriSWACH.
Researchers and engineers from different institutes conducted a number of investigations on general
trimarans and the TriSWACH due to their promising
* Biography: YANG Chi, Female, Ph. D., Professor

features and complex flow characteristics[2-6]. The hydrodynamic properties, such as resistance, sinkage,
trim and seakeeping of the TriSWACH with a variety
of side hull configurations, were evaluated by numerical simulations and experimental measurements. Studies show that side hull locations have a significant
impact on the flow interference effects between hulls,
thus affect the hydrodynamic performance of the
TriSWACH. There has been little research that considers both side hull forms and side hull locations in
TriSWACH hydrodynamic design optimization due to
its difficulties and complexities.
In the present study, a new methodology for hydrodynamic optimization of a TriSWACH is developed, which considers not only the locations of the side
hulls but also side hull forms. An integrated hydrodynamic computational tool has been developed for evaluating steady free-surface flow about a multihull ship
(TriSWACH in the present study), for studying the strong near-field interference effects between closelyspaced multihulls, and for optimizing the hull forms
by minimizing the total resistance. The present integrated hydrodynamic computational tool consists of a

857

Fig.1 A TriSWACH with 9 side hull positions (A-I) (m)

potential-flow based simple CFD tool (computer code


SSF) and an Euler/RANS/Navier-Stokes based advanced CFD tool (computer code FEFLO). Both CFD
tools have been integrated into the hydrodynamic module of a practical multi-objective hydrodynamic optimization tool. The other components of this hydrodynamic optimization tool consist of a hull shape representation and modification module and an optimization module. The simple CFD tool in the hydrodynamic module is used to evaluate the wave drag and
total drag in the optimization process and the advanced CFD tool is used for the validation of the optimal
hull forms obtained. The hull shape representation and
modification module is based on a combined local and
global hull form modification approach. The multi-objective genetic algorithm (MOGA) is implemented in
the optimization module to produce optimal solutions
that minimize total drag at given design speeds.
The multi-objective hydrodynamic optimization
tool developed in this study has been applied to the
hydrodynamic optimization of the TriSWACH for reduced drag by optimizing the side hulls only. A new
methodology is developed to optimize side hull forms
and locations so that the TriSWACH has a minimal
resistance for a wide speed range and for various side
hull positions. Two sets of side hulls are developed
and used for the design of two new optimal
TriSWACH models. Model tests are carried out for
two optimal TriSWACH models at Webb Institute for
validations. Substantial drag reductions have been obtained for a wide range of speed.

toward the ship bow. The Froude number FN and the


Reynolds number Re are defined as
U

FN =

Re =

UL

(2)

where U is the ship speed, g is the gravitational


acceleration, and is the kinematic viscosity of the
water.
The total drag of a ship hull is approximated by
the summation of wave drag and the friction drag. The
wave drag is calculated by the Neumann-Michell (NM)
theory. A detailed description of the formulation and
the numerical solution procedure for NM theory can
be found in authors' previous work[7-10].
The total drag coefficient is defined as
CT =

RT
0.5 U 2 S wet

= C W + CF

(3)

where is the water density, S wet is the wetted surface area, R T is the total drag, and C W is the wave
drag coefficient evaluated by non-dimensionalizing
the wave drag RW using a conventional way as

CW =
1. Drag evaluation
Consider the steady flow about a multi-hull ship
advancing in calm water. Non-dimensional coordinates x ( x, y, z ) X / L are defined in terms of a reference length L , typically taken as the ship length.
The z axis is vertical and points upward, and the
mean free surface is taken as the plane z = 0 . The x
axis is chosen along the path of the ship and points

(1)

gL

RW
0.5 U 2 S wet

(4)

and CF is the frictional drag coefficient evaluated


using ITTC 1957 Model-Ship Correlation Line as follows

CF =

0.075
(lg10 Re 2)2

(5)

858

The computer code (SSF) based on the NM theory and the ITTC 1957 Model-Ship Correlation Line
has been used to evaluate the wave drag and the total
drag for various monohull forms and catamarans. Numerical results predicted by the NM theory are in fairly good agreement with experimental measurements[7-18].

Fig.2(b) Comparison of the drag and drag coefficient of the


TriSWAH with the side hull at position F

Fig.2(a) Comparison of the drag and drag coefficient of the


TriSWAH with the side hull at position C

For the case of evaluating friction drag of the


TriSWACH, it should be noted that the side hulls are

shorter than the center hull, and thus operate at lower


Reynolds number. The friction drag coefficient of the
side hull needs to be evaluated in terms of the
Reynolds number defined by the side hull length. The
friction drag coefficient of the TriSWACH can then
be evaluated as follows
CFtrimaran = CFcenterhull

Scenter hull
2S
+ CFsidehull side hull
S trimaran
S trimaran

(6)

859

where CFtrimaran

is the friction coefficient of the

TriSWACH, CFcenter hull is the friction coefficient of the


center hull, CFsidehull is the friction coefficient of the
side hull, Scenter hull is the wetted surface area of the
center hull, Sside hull is the wetted surface area of the
side hull, and S trimaran is the summation of the wetted
surface area of the center hull and two side hulls.

Fig.2(c) Comparison of the drag and drag coefficient of the


TriSWAH with the side hull at position I

Before the optimization study is carried out, the


validation study of the simple CFD tool (computer
code SSF) is first performed for the initial hull form
considered in this study, i.e., the TriSWACH model
with various side hull configurations (A-I) shown in
Fig.1. Figures 2(a)-2(c) show the comparisons of experimental measurements and numerical predictions
of the wave drag coefficient, total drag coefficient,
wave drag and total drag at three different longitudinal
side hull positions C, F and I depicted in Fig.1. Specifically, all numerical results shown in Figs.2(a)-2(c)
are obtained using simple CFD tool (computer code
SSF), where the wave drag is calculated by the
Neumann-Michell (NM) theory using the near-field
(pressure integration) approach and the friction drag
coefficient is evaluated for the model scale using
Eq.(6).
It can be observed from Figs.2(a)-2(c) that the
wave drag coefficient, wave drag, total drag coefficient and total drag evaluated by the simple CFD tool
(computer code SSF) are consistent with the experimental measurements. Figures 2(a)-2(c) also suggest
that this simple CFD tool can predict the wave drag
and total drag for the TriSWACH with reasonable accuracy. It should also be noted that the computing
time for the TriSWACH that is discretized by approximately 7 500 panels is about 16 seconds per Froude
number on a PC (3.2 GHz Intel P4, 2G Bytes RAM,
Linux OS, Intel compiler). Thus, present simple CFD
tool is a highly efficient and effective tool. Therefore,
this NM theory based simple CFD tool is well suited
for the hydrodynamic optimization of the TriSWACH
model.
2. A methodology for TriSWACH optimization
The hydrodynamic optimization of a TriSWACH
is more complex and difficult than that of a monohull
ship. These difficulties are due to the fact that both
hull form and side hull location need to be considered
in the optimization. In addition, design speed and offdesign speed may require different side hull location
as well. It should also be noted that the side hull locations may be controlled by operating conditions as
well. In the present study, a new methodology is developed to optimize side hull forms so that the
TriSWACH has a minimal drag for a wide range of
speed and for various side hull positions shown in
Fig.1.
As the purpose of this TriSWACH optimization
study is to keep center hull unchanged and design
completely different side hull forms to achieve a drag
reduction for a wide speed range and for all 9 hull positions shown in Fig.1, a specific design optimization
procedure is developed. Figure 3 shows the flow chart
of this proposed design optimization procedure.

860

Fig.3 Flow chart of the TriSWACH optimization procedure

In order to design a completely different side hull


form with a few given concepts of shapes derived
from the flow around the TriSWACH simulated using
advanced CFD tool (computer code FEFLO), the first
step is to create a prototype side hull from scratch in
Rhino in a non-dimensional scale. The side hull is
then scaled in length and depth in order to adjust the
displacement to meet the design requirement.
In the second step, the prototype side hull will be
used to build a new TriSWACH, and a selection process is defined according to the drag performance of
the new TriSWACH. Specifically, the drag of the
TriSWACH that is constructed by the center hull and
the new prototype side hulls at 9 positions defined in
Figure 1 is evaluated, respectively, using the simple
CFD tool SSF. If the average drag of the new
TriSWACH at 9 positions does not show any reduction, this side hull model is disregarded and the first
step needs to be repeated to generate a new side hull
model in Rhino. Otherwise, the model is kept and
used in the next step for further shape optimization.
In the third step, the good prototype side hull generated from the step 2 is used to form new
TriSWACH models and a searching for the best side
hull position with respect to the center hull is performed. Specifically, a new side hull is combined with
the center hull to form a series of new TriSWACH
models with specified spacing between them. Three
design speeds that corresponding to FN = 0.25, 0.35
and 0.45 are used, respectively, to evaluate the wave
drag for the new TriSWACH models. The average
wave drag for three given design speeds can then be
obtained for each new TriSWACH model, and the
computed wave drag can be used to generate a two
dimensional wave drag contour plot for the given
range of the longitudinal and horizontal spacing between the center hull and the side hulls. The best side
hull position can then be obtained, which means that
the new TriSWACH model built with the original center hull and the new side hulls placed at this best side

hull position has a minimum average wave drag for


FN = 0.25, 0.35 and 0.45 in the current study.
In the fourth step, a multi-objective hydrodynamic optimization procedure is applied to the new
TriSWACH model generated in the third step to optimize the shape of the side hulls at three design speeds,
FN = 0.25, 0.35 and 0.45, for reduced total drag. Finally, the best position of the obtained optimal side hull
form for the TriSWACH is rechecked.
3. Design procedure, results and analysis
In the optimization process of the current
TriSWACH, more than fifty prototype side hull models are created. Two best prototype side hulls (called
M43 and M44) are kept for the further shape optimization. Before the shape optimization of the prototype
side hull is carried out, the best position of the side
hull with respect to the original center hull is determined using the procedure described in the previous
section. Table 1 shows the final best positions of the
original side hulls and two sets of the best prototype
side hulls, respectively.
Table 1 Optimal position of side hulls

X (m)

Y (m)

Initial (SD)

1.003

0.216

Model 43 (M43(BEST))

1.010

0.231

Model 44 (M44(BEST))

1.022

0.202

It can be seen that the best position for the original side hull is the location F shown in Fig.1, and
the best positions of two sets of the best prototype side
hulls are close to the location F. Numerical tests show
that the variation of the best side hull positions is relatively small even though the shape of the side hull
has a relatively big change.

861

Fig.4 Comparison of wave drag coefficient (a), wave drag (b),


total drag coefficient (c) and total drag (d) between the
original TriSWACH (Thick black line: average value for
9 side hull positions. Vertical gray line: minimum and
maximum values for 9 side hull positions) and the optimal TriSWACH (Thin black line: optimal side hull
M43 at its best position)

Fig.5 Comparison of wave drag coefficient (a), wave drag (b),


total drag coefficient (c) and total drag (d) between the
original TriSWACH (Thick black line: average value for
9 side hull positions. Vertical gray line: minimum and
maximum values for 9 side hull positions) and the optimal TriSWACH (Thin black line: optimal side hull
M44 at its best position)

The shape optimization of the TriSWACH that


was formed with the original center hull and a chosen
set of prototype side hulls (M43 or M44) placed at
their best position is performed at three design speeds

( FN = 0.25, 0.35 and 0.45) for reduced total drag. During the optimization, only the shape of the side hull is
modified using the radius basis function (RBF) interpolation method. The details of the hull form modi-

862

Fig.6(a) 3-D view of the original TriSWACH model

Fig.6(b) 3-D View of the optimal TriSWACH model, M43 (BEST)

Fig.6(c) 3-D View of the optimal TriSWACH model, M44 (BEST)

fication with RBF method can be found in authors


previous work[14,15]. In the present study, six design
variables are adopted to control the modification of
the shape of the whole side hull in transverse direction.
Multi-objective genetic algorithm is used for searching the optimal solution sets. The population size is
set as twenty one and one hundred generations of simulations are conducted. At the end of the simulation,
the approximated Pareto front solution sets are found
and the final optimal solution is determined by the designer. Two optimal TriSWACH models are obtained,
and they are denoted as M43 (BEST) and M44
(BEST), respectively. After two optimal side hull
forms are obtained, the optimal side hull position for
the new TriSWACH model is rechecked.
In order to analyze the performance of the optimal TriSWACH models, M43 (BEST) and M44
(BEST), in the entire speed range, the comparison of
the wave drag between original TriSWACH and the
optimal TriSWACH models is carried out. Specifically, the wave drag of the original TriSWACH is evaluated at 9 side hull positions for the given speed
range ( FN = 0.1- 0.55) . For a given speed, an average wave drag, a minimum wave drag and a maximum
wave drag are obtained for 9 side hull positions depicted in Fig.1. The average wave drag and wave drag
coefficient, and the average total drag and total drag
coefficient for the entire speed range are plotted with
thick black solid lines in Figs.4 and 5, respectively,

and the minimum and maximum values of the wave


drag wave drag coefficient, total drag and total drag
coefficient at each speed are plotted with vertical gray
lines in Figures 4 and 5, respectively. In order to
compare with the results of the original TriSWACH,
the wave drag, wave drag coefficient, total drag and
total drag coefficient of two optimal TriSWACH
models, M43 (BEST) and M44 (BEST), are evaluated
for the entire speed range as well, and plotted with
thin black lines in Figs.4 and 5, respectively. The 3-D
view of the original and optimal TriSWACH models
is shown in Figs.6(a)-6(c).
It can be observed from Figs.4 and 5 that both
optimal TriSWACH models (with the original center
hull and the M43 and M44 side hulls placed at their
best positions) have a significant drag reduction in a
wide speed range. In addition, it is interesting to notice that both optimal side hulls M43 and M44 are
asymmetry. The main difference between model M43
and M44 is that the model M44 has a transom stern.
In order to validate the optimal TriSWACH models obtained in this study, model tests are carried out
for the original and two optimal TriSWACH models
at Webb Institute[19]. Figure 7 shows the comparison
of the residual drag coefficient (a), residual drag (b),
total drag coefficient (c) and total drag (d) of the original TriSWACH model and the two optimal
TriSWACH models (side hull M43 and M44), where
the line with circles, line with triangles and line with

863

by the two optimal models at a wide range of speed,


especially at high speed. Therefore, the hydrodynamic
optimization tool developed in this study and the
proposed design optimization procedure are very
effective in generating optimal TriSWACH.
4. Conclusion
A new methodology for the hydrodynamic design optimization of the TriSWACH is developed in the
present study, which considers not only the side hull
positions, but also the shape of the side hulls. Two
sets of side hulls are developed and used for the design of two new optimal TriSWACH models. Model
tests are carried out for the original and optimal
TriSWACH models at Webb Institute to validate the
hydrodynamic performance of the optimal
TriSWACH models. Experimental measurements and
numerical results are consistent, which show that two
optimal TriSWACH models obtained using the present hydrodynamic optimization tool and proposed methodology can attain substantial drag reductions for a
wide range of speed in comparison with the original
TriSWACH model. Therefore, the present computational tool can be used effectively for the future design
optimization of the TriSWACH, or trimarans, in general.
Acknowledgments
This work was sponsored by the Office of Naval
Research. Ms. Kelly Cooper was the technical monitor.
The authors would like to thank Ms. Kelly Cooper for
her support and encouragement. The authors would
also like to thank the Center for Innovation in Ship
Design at the NSWCCD for the technical guidance,
design insight, support and encouragement. The validation model test was carried out at Webb Institute
and sponsored by the ONR. The authors would like to
thank Prof. Richard Royce of Webb Institute and his
students for conducting model test and providing the
experimental data for the validation.
References
Fig.7 Comparison of experimental measurements of drag (or
drag coefficient) of the original TriSWACH model (original side hull: line with circles) and two optimal
TriSWACH models (optimal M43 side hull: line with
triangles, optimal M44 side hull: line with quads) at their
best positions

quads represent the experimental measurements of the


drag (or drag coefficient) of the TriSWACH models
with the original side hull, M43 side hull and M44
side hull at their best positions, respectively. It can be
seen that large amount of drag reduction is achieved

[1]
[2]

[3]
[4]

SUZUKI K. TriSWACH ASW Corvette[R]. NSWCCDCISD 2011/016, Bethesda, MD, USA: Naval Surface
Warfare Center Carderock Division, 2011, 25.
YANG C., NOBLESSE F. and LOHNER R. Practical
CFD applications to design of a wave cancellation multihull ship[C]. Proceeding of the 23rd Symposium on
Naval Hydrodynamics. Val de Ruil, France, 2000,
206-222.
YANG C., SOTO O. and LOHNER R. et al. Hydrodynamic optimization of a trimaran[J]. Ship Technology
Research, 2002. 49(2): 70-92.
HARRINGTON A., WELLS J. D. TriSWACH small
model testing[R]. NSWCCD-CISD-2011/009, Bethesda,

864

[5]
[6]

[7]

[8]

[9]
[10]

[11]

[12]

MD, USA: Naval Surface Warfare Center Carderock


Division, 2009, 31.
KLAG J., MACMAHON I. Calm water resistance of a
novel trimaran[C]. Senior Thesis, Glen Cove, New
York, USA: Webb Institute, 2011, 188.
WILSON M. B., HSU C. C. and JENKINS D. S. Experiments and predictions of the resistance characteristics
of a wave cancellation multihull ship concept[C]. Proceeding of the 23rd American Towing Tank Conference. New Orleans, USA, 1993, 103-112.
YANG C., KIM H. Y. and NOBLESSE F. A practical
method for evaluating steady flow about a ship[C]. Proceedings of the FAST2007. Shanghai, China, 2007,
118-126.
GUEDES SOARES C., KOLEV P. Maritime industry,
ocean engineering and coastal resources. Vol. 1 Maritime transportation[M]. Abingdon, UK: Taylor and
Francis, 129-136.
NOBLESSE F., HUANG F. and YANG C. The
Neumann-Michell theory of ship waves[J]. Journal of
Engineering Mathematics, 2013, 79(1): 51-71.
HUANG F., YANG C. and NOBLESSE F. Numerical
implementation and validation of the Neumann-Michell
theory of ship waves[J]. European Journal of Mechanics B/Fluids, 2013, 42: 47-68.
KIM H. Y., YANG C. and LOHNER R. et al. A practical hydrodynamic optimization tool for the design of a
monohull ship[C]. Proceedings of the 18th ISOPE.
Vancouver, Canada, 2008, 98-107.
YANG C., KIM H. Y. and LOHNER R. et al. Practical
hydrodynamic optimization of ship hull forms[C]. Proceedings of the GCMS08. Edinburgh, UK, 2008, 435444.

[13]

[14]

[15]
[16]

[17]

[18]

[19]

KIM H. Y., YANG C. and KIM H. et al. Application of


a practical multi-objective optimization tool to hydrodynamic design of a surface combatant ship[C]. Proceedings of the GCMS09. Istanbul, Turkey, 2009, 56-65.
KIM H. Y., YANG C. and NOBLESSE F. hull form optimization for reduced resistance and improved seakeeping via practical designed-oriented CFD tools[C]. Proceedings of the GCMS10. Ottawa, Canada, 2010, 375385.
KIM H. Y., YANG C. Hydrodynamic optimization of
multihull ships[C]. Proceedings of the FAST2011.
Honolulu, Hawaii, USA, 2011.
KIM H. Y., JEONG S. and YANG C. et al. Hull form
design exploration based on response surface method[C]. Proceedings of the 21st ISOPE. Maui, USA,
2011, 816-825.
KIM H. Y., YANG C. Design optimization of bulbous
bow and stern end bulb for reduced drag[C]. Proceedings of the 23rd ISOPE. Anchorage, USA, 2013, 765772.
HUANG F. KIM H. Y. and YANG C. A new method
for ship bulbous bow generation and modification[C].
Proceedings of the 24th ISOPE. Busan, Korea, 2014,
823-830.
LAROSE M. D. and SMITH M. V. Validation of CFD
optimized demihulls for TriSWACH[C]. Senior Thesis,
Glen Cove, New York, USA: Webb Institute, 2013.

You might also like