Journal of Naval Science and Engineering

2016, Vol. 12, No.2, pp.1-26

Gemi İnşa Mühendisliği/Marine Engineering

INVESTIGATION OF FLOW DISTRIBUTION AROUND A

SUBMARINE

Abdi Kukner*, Adem DURAN**, Tarkan ÇINAR***

* kukner@itu.edu.tr, Faculty of Naval Architecture and Ocean Engineering, Istanbul

Technical University
** admdrn@gmail.com, Faculty of Naval Architecture and Marine Engineering, Yildiz
Technical University
*** tarkancinar66@gmail.com, Faculty of Naval Architecture and Marine Engineering,
Yildiz Technical University

Date of Receive: 19.09.2016 Date of Acceptance: 01.11.2016

ABSTRACT
In this paper the flow field around a submarine has been investigated.
Pressure distribution and its impact on submarine hull form have been
studied. The accurate and efficient prediction of hydrodynamic pressure and
forces on a submarine has been achieved by investigating the flow related to
the interaction of the vertical flow shed from the sail and the cross-flow
boundary layer of the hull. Therefore this study aims to simulate the flow
field of a submarine by using finite volume method. Finite Volume Stress
Analysis Method and k-ω turbulence model have been used to simulate
turbulent flow past the submarine hull surface. A submarine hull with
overall length of 80 meters and diameter of 10 meters was chosen. It has aft
body length 11m and sail length 7m. The speed range of the submarine is 0
to 30 knots with 5 knots increments. Calculated pressure coefficients along
the submarine hull are discussed to show the effect of the sail lateral
position and the stern appendages. It is also discussed a Reynolds Averaged
Navier-Stokes (RANS) code application in the design of an "Advanced Sail"
for a submarine.
Keywords: Flow distribituon, submarine, pressure distribution, RANS.

1
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

ÖZ

Bu makalede, bir denizaltının gövdesi üzerine etkiyen akışkan basıncı ve

kuvvetlerinin daha iyi anlaşılması için denizaltının etrafındaki akış dağılımı
detaylı bir şekilde incelenmiştir. Denizaltıya etkiyen hidrodinamik basınç ve
kuvvetlerin doğru ve etkili bir şekilde tahmin edilebilmesi için denizaltının
yelkeni üzerindeki düşey akış ve teknenin sınır tabakasındaki enine akışının
etkileşimi ile ilgili olan akış incelenmiştir. Bu çalışmada Sonlu Hacim
Yöntemi (FVM) ve Hesaplamalı Akışkanlar Dinamiği (CFD) kullanılarak
denizaltı etrafındaki akış alanı simüle edilmiştir. Sonlu Hacim Stres Analiz
Yöntemi ve k-ω türbülans modeli kullanılarak denizaltının gövdesinin
yüzeyini takip eden türbülanslı akışın simülasyonu yapılmıştır. Boyu 80 m,
eni 10 m, kıç kuyruk uzunluğu 11 m, yelken uzunluğu 7 m olan bir denizaltı
modeli seçilmiş ve RhinoCerosTM programı kullanılarak çizimi yapılmıştır.
Denizaltının hızı 0’dan başlayarak 30 knota kadar 5 knot arttırılmak
suretiyle hesaplamalar yapılmıştır. Günümüzdeki güçlü bilgisayarların
getirdiği kolaylıklardan dolayı akışkan probleminin tam olarak Navier-
Stokes denklemiyle sayısal hassas çözümü geniş bir yelpaze alanı içerisinde
yapılabilmektedir. Denizaltı gövdesi boyunca hesaplanan basınç katsayıları
yelken ve kıç takıntılarının etkisini göstermek için ele alınmıştır. Ayrıca, bir
denizaltı için “Gelişmiş Yelken” tasarımında RANS (Reynolds Averaged
Navier-Stokes) kodunun uygulaması ele alınmış ve irdelenmiştir.

Introduction

A submarine is a vessel capable of independent operation underwater. It is

used as a surface naval weapons platform or as a tool of exploration and
recreation. Their stealth plays an important role in a modern naval force.
Therefore submarine is a warship with a streamlined hull design to operate
completely submerged in the sea for long periods, equiped with a periscope
and typically armed with torpedoes or missiles. Most large submarines have
a cylindrical body with hemispherical (and/or conical) ends and a vertical
structure, usually located amidships having navigation and other equipment
devices as well as periscopes. Sometimes known as the conning tower. This
vertical structure is called “sail” in U.S. Navy, “fin” in European Navies.

2
 Investigation Of Flow Distribution Around A Submarine

The propeller of submarine, vertical and horizontal control panels are

located at stern. As the thrust is generated, water pushes over the planes,
creating an upward or downward force that helps the sub gradually surface
or dive. The fins can be tilted to change the angle of attack at which it
climbs or dives.
The starting point of all scifientic studies is a literature survey to understand
the status quo of the investigated topic. It is important to understand the
reasons for the shape of submarines at different stages of their development
and why changes were made. To neglect full scientific studies would be a
serious mistake in the design of any future replacement submarine.
In submarine hydrodynamics, turbulence and vortex dynamics play an
important role. The classic picture of turbulence starts from a sequence of
bifurcations in a “smooth” flow, each of which introduces flow structures of
smaller and smaller scales. Designers had begun to change nose and tail
cone shape to improve the performance of submarine at operational speeds.
Other major sources of resistance may be improved. The establishment of
the detail performance of a submarine can be started by using computational
fluid dynamics to obtain pressure distribution and to calculate the drag
characteristics which will serve as the comparative foundation for any new
design. All features affecting the shape of submarine are discussed including
the boundary layer, laminar flow, transition, turbulence and separation and
how the flow should be as quiet and smooth as possible. At the beginning
the pressure distribution around submarine body without sail, and
appendages were investigated. The next step was; the sail, tailplanes and
foreplanes were added to obtain pressure distribution around the submarine
and to observe how effects and changes in flow distributions. Design looks
like a jigsaw puzzle where altering one piece requires alterations in all
surrounding features to make a workable complete design. It is clear that
scientific studies has to be a starting point for any future submarine design.
A review of relevant literature has been completed which covered priorities
in design and showed how enhancement of one feature interacts with other
features and may even result in an overall loss of performance despite the
perceived advantage of the enhanced feature. Hydrodynamic aspects were
then discussed starting with the shape and reasons what should be the beam-
to-depth ratio (B/D) to give minimum resistance as possible.

3
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

As it is well known, flow around submarines is exceedingly complicated,

even at simple flow conditions, and the need to reduce submarine signatures
from flow-induced noise put high demands on the computational model.
Most of the boundary layer on a submarine is predominantly turbulent
because of the high Reynolds (Re) number, which typically is encountered
in ship hydrodynamics. At the bow, the flow is usually laminar, but rapidly
undergoes transitions into a fully turbulent boundary layer, which often
makes it reasonable to assume a fully turbulent boundary layer along the
entire hull. The boundary layer is further affected by pressure gradients
(mainly around the bow and the stern) and the hull curvature, potentially
causing a vortex separation usually resulting in distortion of the propeller
inflow.

Prediction of Submarine Resistance

Whenever a body is placed in a flow, the body is subject to a force from the
surrounding fluid. In general, the force acting on a body is resolved into a
component D in the flow direction U and the component L in a direction
normal to U. The component D is called drag and L is called lift. The most
important difference between the resistance of a surface ship (or submarine
on the water surface) is that for a deeply submerged submarine will not have
wave resistance. Therefore the submerged submarine resistance will sum up
total skin friction and total submerged pressure. Skin friction drag acts
tangentially at the surface and is proportional to the wetted surface.
The total pressure has form resistance or form drag and induced resistance
or induced drag. The form drag is the viscous pressure resistance due to the
shape of the submarine. The induced drag is the resistance caused by lift.
This could be on appendages that are generating lift due to misalignment
with the flow, or to the hull, that may be generating lift due to symmetry.
The resistance of a submarine can be determined either by model testing, or
by Computational Fluid Dynamics (CFD). In this paper, CFD techniques
have been used to estimate the resistance of the deeply submerged
submarine. As the resistance of a deeply submerged submarine is dominated
by the frictional component, there are a number of difficulties with this, in
particular the choice of empirically based turbulence model. However, in
principle it is possible to use CFD to obtain results at full scale Reynolds

4
 Investigation Of Flow Distribution Around A Submarine

numbers, something which is not possible using model experiments [1].

There is one of the current complications with CFD is that there is no
standard method for predicting submarine resistance. This is largely because
both computing power, and CFD techniques, are developing rapidly. Thus,
great care needs to be taken when investigating the effect of the change in
resistance due to a change in hull shape.

Numerical Model

The use of computational tools to evaluate submarine flows have been

tremendously increased over the last decade since the capacity and speed of
computers were raised. Thus the applications of Computational Fluid
Dynamics (CFD) to the naval industry was guiding the design of
submarine. In view of these developments, CFD can offer a cost-effective
solution to many problems in underwater vehicle hull forms. However,
effective utilization of CFD for naval hydrodynamics depends on proper
selection of turbulence model, grid generation and boundary resolution. The
most common turbulence modeling approach of today is RANS, which is
based on Reynolds Averaged Navier-Stokes equations and is often adopted
in traditional Computational Fluid Dynamics (CFD). The first fully
appended submarine RANS calculation was done by Gorski et al [3] for the
submarine configuration, which was extensively measured to provide a
data-base to test CFD methods. The dependent variables are divided into a
mean part and a fluctuating component representing deviations from this
mean. The advantage of RANS is however that the approach is fast, and it is
available in most CFD codes. In particular, with the advent of parallel
computational capabilities, viscous RANS simulations have seen a larger
role in predicting these flow fields.
In this study, Reynolds Averaged Navier Stokes (RANS) equations and
continuity equation for mean velocity of the unsteady, incompressible fluid
have been used as governing equations in order to determine the mean
cartesian flow field, , and the mean pressure (P) of the water around the
hull. The well known SST (Shear Stress Transport) k-ω model have been
considered to simulate the turbulance flows.

5
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

where f i represents external forces. The influence on turbulence on the

mean flow is given in equation represents external forces. The influence on
turbulence on the mean flow is given in equation (2) by the Reynolds stress
tensor . There are many turbulence models to provide solutions to
the Reynolds stresses.

The k - ω model is well-suited for prediction in the vicinity of the wall,

while the k - ε model is for the remaining area near the boundary region.
The k- SST-model is using blending functions to be able to use the k-ω
model near the wall and the k-ε in the free stream and to get a smooth
transition between them. Therefore it is a hybrid between the k- ε and the k-
ω model. The SST k - ω model is known to be fairly effective for better
prediction of adverse pressure gradient and flow separation. This model has
been designed to promote turbulence in the congestion zone of fluid flow.

The SST k-ω turbulence model is a two-equation eddy-viscosity model

developed by Menter [4] to effectively blend the robust and accurate
formulation of the k-ω model in the near-wall region with the free-stream
independence of the k-ε model in the far field. To achieve this, the k- ε
model is converted into a k-ω formulation. Transport equations for the SST
k-ω model are given by:

In these equations, represents the generation of turbulence kinetic

energy due to mean velocity gradients, Gω represents the generation of ω ,

6
 Investigation Of Flow Distribution Around A Submarine

Гk and Гω represent the effective diffusivity of k and ω, respectively, Yk and

Yω represent the dissipation of k and ω due to turbulence, Dω represents the
cross-diffusion term, Sk and Sω are user-defined source terms.

The Model of Submarine Hull

A standard submarine hull model was used as a prototype for computations.
The bow of our submarine model has been chosen as ellipsoidal and the
stern has been chosen paraboloidal in shape with a portion of parallel mid-
body. Since CFD method was used for the computations, This method is a
very grid dependent technique. Therefore CFD method needs to be meshed
in proper ways to get reliable and converged results. The largest errors occur
where the largest gradients are. For this reason, the resolution should be
increased in such regions. Only a restricted amount of cells can be used due
to restrictions in computational power. Therefore it is beneficial to have a
denser grid where e.g. the curvature of the surface is high and having larger
cells closer to the middle of the surface. The discretization of the geometric
domain of the submarine has been divided into 950.000 hexahedral meshes
and every simulation has been iterated three hundured times. Since the CFD
calculations on the computer takes a lot of time and needs more memory. It
is 1/50 scale model rather than the actual size of the submarine has been
used for the computations. The hull model has an overall length L of 1.6 m
and maximum diameter D of 0.20 m. The sail is located in front of the hull
with a length of 0.24 m..
The profile of the submarine model hull is shown in Figure 1. Also shown is
the profile of nose cone, tail cone and sail shape.

Figure 1. CAD Model of the Submarine Hull

7
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 2. Submarine Control Surfaces

Numerical Computations
The numerical calculations were attempted by the following certain steps.
The first step was the bare submarine body which was taken as a
cylinderical shape for the flow calculations to observe how to change flow
distribution according to different B/D ratios. It was analysed according to
1, 2, 4 and 8 ratio values. The subsequent steps, the numerical computations
for hull with sail, with hull-sail and aft planes, and finally having all
necessary control surfaces components of submarine form have been carried
out seperately. All numerical computations were performed, in the
following figures, on the actual size of bare submarine body. For each case,
the results are shown in Figures 3 to 7 respectively.

8
 Investigation Of Flow Distribution Around A Submarine

Figure 3. The pressure, velocity and velocity vector distributions around a

cylindericsl submarine body (without sail, tails and other appendages) for
different B/D ratios.

In these figures, it can be seen that how the absolute pressure and velocity
distribution change around the submarine when the submarine body is
assumed to be fixed and flow is coming from front of it. The computed
values of absolute pressures and drag forces (resistance) for different B/D
ratios of bare submarine body is given in table 1. The values given in this
table are calculated at 25 knots of submarine speed.

9
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Table 1. The absolute pressures and forces (resistances) values for different
B/D ratios of bare submarine body
Drag Percent of
B/D Vmin(m/s) Vmax(m/s) Pmin(Pa) Pmax(mPa) force Changing
(N) in force
1 0 15.698 39795 1838 235.39 -
2 0 15.358 43192 1752 168.59 %29
4 0 14.673 54406 1747 129.43 %24
8 0 13.957 59520 1589 126.38 %3
The values given in table 1 states that, consequently increasing B/D ratio of
the bare submarine body will lead to reduced the drag force (resistance).
This shows that the resistance depends on the pressure distribution around
the body eventhough the minimum pressure values are increasing and the
maximum pressure values are decreasing acording to B/D ratios geting
higher.

Figure 4. The pressure, velocity and velocity vector distributions around the
bare submarine body (without sail, tails and other appendages).

10
 Investigation Of Flow Distribution Around A Submarine

Figure 5. The pressure, velocity and velocity vector distributions around the
submarine body with sail.

Figure 6. The pressure, velocity and velocity vector distributions around the
submarine body with sail and tails (aft planes).

11
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 7. The pressure, velocity and velocity vector distributions around the
submarine body with sail and tails ( Hull+Sail+Sail Planes+Aft Planes).

Table-2 A comparison of all results according to submarine’s components

As can be seen from Table 2 that the resistance increases because the sail
area creates additional surface to create more resistance. On the other hand,
the pressure values do not change dramatically as it was obtained without

12
 Investigation Of Flow Distribution Around A Submarine

having sail. The full body of the submarine gives less pressure value than
bare hull and Hull + Sail + Aft Planes form.

Flow Distribution Around Submarine According to Its Speed Variation

The absolute pressure and velocity distributions around the model
submarine have been computed for the different submarine velocities from 0
to 2.18 m/s with increments of 0.364 m/s (corresponds to 5 knots of
submarine speed). The results for each case are shown in figures 8 to 12.

Figure 8. The pressure, velocity and velocity vector distributions around the
submarine model at 0.364 m/s speed.

a) Pressure distribution b) Velocity distribution c) Velocity

vector distribution

13
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 9. The pressure, velocity and velocity vector distributions around the
submarine model at 0.728 m/s speed.

14
 Investigation Of Flow Distribution Around A Submarine

F
igure 10. The pressure, velocity and velocity vector distributions around the
submarine model at 1.0912 m/s speed.

Figure 11. The pressure, velocity and velocity vector distributions around
the submarine model at 1.455 m/s speed.

15
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 12. The pressure, velocity and velocity vector distributions around
the submarine model at 1.818 m/s speed.

Figure 13. The pressure, velocity and velocity vector distributions and Wall
Shear Stress around the submarine model at 2.1824 m/s speed.

16
 Investigation Of Flow Distribution Around A Submarine

The computed values of drag and lift forces (resistance) for different speed
of submarine model is given in table 3.

Table 3. The absolute pressures and drag and lift forces (resistances) values
for different speed of model submarine.
* The speed values are given in the table for the model submarine. They correspond to 5 to 30 knots
of actual submarine speeds with 5 knots increments.

The values of maximum pressure, drag force and lift force acting on the
submarine model are given interms of Reynolds number in Figure 14-16,
respectively.

Figure 14. Reynolds Number vs Pressure (Max) (Pa)

17
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 15. Reynolds Number vs Drag Force (N)

Figure 16. Reynolds Number vs Lift Force (N)

Investigation of the Sail Position According to Flow Distribution

This study was subsequently expanded to investigate the effect of sail
position on design of submarine hydrodynamic. It is well known that the

18
 Investigation Of Flow Distribution Around A Submarine

details of the position and shape of the submarine sail will depend on the
number of masts, type of power source, type of periscope as well as effects
on steering and dynamic stability. Indeed it provides a bridge platform for
conning the submarine on the surface and a supporting structure for about
number of masts. It may also support the forward control fins. In the past,
the location of the sail has been dictated by through-hull penetration masts
like periscopes which could only be located in certain positions. This should
not apply in the future because of improved designs of such systems to
provide non hull penetrating masts. Choosing the correct position and height
is important. If too tall it affects the centre of mass and may cause a greater
snap roll [8]. Any non penetrating mast needs to be properly supported
Arentzen and Mandel [6] report that the drag of these large appendages may
be between 15-30 % of the bare hull drag.
In this study, six sail positions were examined, the first position has been
taken from the nose point by L/(6.9) m for determination of the flow and
absulate pressure distributions around it (see table 4). Then the sail position
has been changed to backward by taking equal increment from its position
at each step for the computation of the flow and absulate pressures
distributions. On the other hand, the computations were carried out for three
different sail cross sections such as NACA0012, NACA0018 and
NACA0024. The velocity and absolute pressure distribution values obtained
from CFD computations depending on changing the position of the
submarine sailing are shown in figure 19 and 20, respectively where the
sailing cross section has been taken as NACA0018. Before it can be
considered the design of the submarine sail, it is important to review the
basic phyics of the flow around foil sections. For example it is assumed that
the foil has constant section, and is long enough; in this case, the flow
around all sections of the sail foil is the same, and this is describe as 2D
flow. Studying 2D flow can give many insights about the effect of the
section shape on the performance.

19
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 17. The first and last sail positions distance from the front of the
submarine

Table 4. NACA Profiles to be used in model sail

Thickness Airfoil Lenght Thickness / Airfoil Lenght

NACA0012 8,4 mm 70,0 mm 0,12
NACA0018 12,6 mm 70,0 mm 0,18
NACA0024 16,8 mm 70,0 mm 0,24

Figure 18. Sail dimension for three NACA Profiles

20
 Investigation Of Flow Distribution Around A Submarine

Figure 19. Velocity distribution with respect to the sail positions.

21
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

Figure 20. Absolute pressure distribution with respect to changing of sail

position

From the position of the sail of the submarine with the flow lines were
calculated by taking the values of absolute pressure in certain places.

Figure 21. Absulate pressure distribution around the submarine in 3-D for
different position of the sail having NACA0018 cross section

Conculusion and Recommendations

The increasing capacity and speed of computers raised the use of

computational fluid dynamics (CFD) to the maritime industry. In the last
decades, many developments have been observed in different areas of
incompressible flow modeling including grid generation techniques,
solution algorithms and turbulence modeling, and computer hardware

22
 Investigation Of Flow Distribution Around A Submarine

capabilities. One important conclusion is that CFD gives the quite accurate
predictions, but requires many CPU-times. It can offer a cost-effective
solution to many problems in underwater vehicle hull forms. However,
effective utilization of CFD for naval hydrodynamics depends on proper
selection of turbulence model, grid generation and boundary resolution. The
most common turbulence modeling approach of today is RANS (with the
SST k-ω turbulence model), which is based on a statistical treatment of the
fluctuations about an average flow; it is expected that RANS will be the
preferred, and fully sufficient, engineering tool for most design aspects. The
advantage of RANS is however that the approach is fast (since only the
mean flow is sought), and it is available in most CFD codes. This method
can accurately predict the velocity field and absolute pressure distribution
around a submarine and its resistance components. It also gives the
possibility to visualize problem areas, such as separation zones. All CFD
calculations were performed at model-scale Reynolds numbers of ~107. The
study easily can be extended to full-scale Reynolds numbers.
The ratio of beam to diameter (same as length to diameter) bears a strong
effect on the total resistance. The more wetted surface the greater the skin
friction. This can be seen from the computational results of the submarine
model used in this study (see table 1), the resistance of the bare submarine
body decreases with increasing B/D ( and L/D) ratio. This states that bare
submarine’s body resistace depens on the pressure distribution around the
body eventhough the minumum pressure is increasing and the maximum
pressure is decreasing. Therefore if the displaced volume of the submarine
is contained in a long thin shape, then the skin friction is greater than for a
shorter, beamier shape of the same volume which has less wetted surface. It
is proposed that a new shape be considered of beamer shape or shorter
length and greater diameter which will reduce the total drag force closer to
the ideal.

In case of full submarine body including the sail and appendages, when the
speed increases the resistance of the submarine increases as expected but the
minimum pressure falling and maximum pressure increases opposite to the
bare body case. This states that sail and appendages play an important role
in submarine design. Besides, the mesh blocks in the vicinity of the sail

23
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

affected by geometry and mesh topology changes. The mesh away from the
sail remained unchanged, leading to more consistent CFD results.

Apart from the hull shape, important items like the sail and control surfaces
need to be optimised for position, size and shape to maximise operational
effectiveness and minimise resistance. The details of the position and shape
of the sail will depend on the number of items beeing built inside of sail.
These details should be considered after the testing of the model of the bare
hull. As a tentative first move the sail is drawn moved forward by
approximately L/7 m from the front of the submarine in order to maintain
the lateral stability and counter-balance the loss in lateral area aft. Indeed,
the sail position, shape and size might be well provided according to the
required volume for advanced future payloads. The sail of the submarine
can now be discussed as it plays a major role in producing drag and hence
its design is critical. Research has proven that a sail may contribute up to
30% of total submarine resistance.

As is known, it is one of the major problems in submarine noises. Flow

noise is primarily caused by turbulence, and the general shape of the hull is
less of a cause of turbulence than poor detailing. Many class of submarines
have had the "old style" sail for many years, with only comparatively minor
attempts at streamlining. Their sails have sharp corners to produce noise. It
would be obvious that the "rounded" "streamlined" sail would produce less
noise. Flow separation is the big cause of unavoidable turbulence, and
unfortunately there isn't much you can do about that beyond a certain point.
One question will rise in our mind from a hydrodynamic point of view,
which is better? The hull shape leads to flow noise which is caused by flow
disturbance thus the hull shape effects the submarine speed.

To validate the CFD code on similar sail shapes and positions calculations
will be compared with experimentally obtained data at the same from in a
wind tunnel or in a water channel. This data comparison includes flow
visualization, axial velocity and surface pressures. The agreement will
demonstrate that RANS codes can be used to provide the significant
hydrodynamics associated with these sail shapes and positions. To improve
the design several modifications can be done on sail position are evaluated

24
 Investigation Of Flow Distribution Around A Submarine

using the RANS code. Based on the predicted secondary flow downstream
of the sail as well as the drag a new design is chosen, without having to
build and test the inferior shapes, reducing time and cost for the program.

25
 Abdi Kukner, Adem DURAN, Tarkan ÇINAR

References

1) “Submarine Hydrodynamics” by Martin Renilson, SpringerBriefs in

Applied Sciences and Technology, New York Dordrecht London, 2015.
2) P.N.Joubert, “Some Aspects of Submarine Design Part 1.
Hydrodynamics” by DSTO Platforms Sciences Laboratory, 506 Lorimer
St, Fishermans Bend, Victoria 3207 Australia. 2004.
3) Gorski, J. J., Coleman, R. M., and Haussling, H. J., "Computation of
Incompressible Flow Around the DARPA SUBOFF Bodies," David
Taylor Research Center Report, DTRC - 90/016, 1990.
4) Menter, F. R., “Two-Equation Eddy-Viscosity Turbulence Models for
Engineering Applications”, AIAA Journal, Vol. 32, No. 8, pp.1598-1605,
August, 1994.
5) N. Vlahopoulos and C. G. Hart., “A Multidisciplinary Design
Optimization Approach to Relating, Affordability and Performance in a
Conceptual Submarine Design”, Journal of Ship Production and Design,
Vol. 26, No. 4, November 2010, pp. 273–289.
6) Arentzen, E. S. and Mandel, P. E., “Naval Architectural Aspects of
Submarine Design”,
Trans. SNAME, pp. 622-692, 1960.
7) D. Ang, L. Chen and J. Tu., “Unsteady RANS Simulation of High
Reynolds Number Trailing Edge Flow”, 15th Australasian Fluid
Mechanics Conference, The University of Sydney, Sydney, Australia,
13-17 December 2004.
8) P. N. Joubert, “Some Aspects of Submarine Design Part 2. Shape of a
Submarine 2026”, Defence Science and Technology Organisation
DSTO-TR-1920, Australia, 2006,
9) Hervey, J. B., “Submarines”, Brassey’s, London, 1994.

26