A.

Yücel Odabaşı Colloquium Series

st
1 International Meeting - Propeller Noise & Vibration
6th – 7th November 2014, Istanbul, Turkey

Underwater Radiated Noise Prediction for a Submarine Propeller in

Different Flow Conditions

M. Cansın Özden1*, Ahmet Y. Gürkan1, Yasemin Arıkan Özden2, Talat G. Canyurt1, Emin Korkut1
1
Istanbul Technical University, Faculty of Naval Architecture and Ocean Engineering, 34469 Maslak-Istanbul, Turkey
2
Faculty of Naval Architecture and Maritime, Yıldız Technical University, 34349 Beşiktaş-Istanbul, Turkey

Abstract: International Maritime Organisation (IMO) and other bodies have been trying to set-up regulations to
reduce/limit noise levels at sea which influence marine environment particularly marine mammals and fish varieties.
Ships with low noise characteristics will be a must in the near future for almost all ship types, however for special ships,
such as naval surface vessels, fisheries, submarines, etc. this has already been an issue for their missions. Propeller is
one of the main sources of underwater noise generated by ships and it is important to predict and control the underwater
noise characteristics of propellers. Within this respect, the main objective of this study is to calculate numerically the
propeller radiated noise. Therefore, propeller noise is investigated numerically for the INSEAN E1619 submarine
propeller in open water and behind a generic DARPA suboff submarine at non-cavitating conditions due to their deeply
submerged operations. Flow around the propeller is solved with a commercial CFD software using Reynolds Averaged
Navier-Stokes (URANS), while hydro-acoustic analysis is performed using a model based on Ffowcs Williams-
Hawking equation. This paper reports on preliminary results of the study. The paper includes the details of the bodies
and study, and discusses further improvement of the methodology.
Keywords: Propeller noise, RANS, non-cavitating, DARPA Suboff submarine, INSEAN E1619 submarine propeller, Ffowcs
Williams–Hawkings (FWH) model.

1 INTRODUCTION computing power and numerical practice, this method

Ship underwater radiated noise has recently been became available also for hydro-acoustic predictions. Seol
concerned by international bodies, such as International et. al. (2002) investigated the non-cavitating propeller
Maritime Organisation (IMO), classification societies as noise employing Boundary Element Methods (BEM) for
well as by conservation groups and governmental the calculation of flow around propeller in time-domain
departments due to negative effects of shipping on marine and used FWH method to predict the far-field acoustics.
environment. Increasing shipping activities also increases Seol et. al. (2005) extended their work to cavitating noise
ambient noise levels at seas around the world and this has stage. They predicted cavity extent by the sheet cavity
a potential impact on marine wildlife, particularly on volume model and used the sheet cavity volume data and
marine mammals and fish varieties. time dependent pressure as the input for the FWH
Ships with low noise characteristics has always been an equation to predict far-field acoustics (Seol et. al. 2005).
important issue for some ship types, naval vessels, Salvatore and Ianniello (2003) published the preliminary
fisheries, research vessels and also for submarines due to results for cavitating propeller noise predictions. A
their missions and operations. Propeller is one of the most hydrodynamic model for transient sheet cavitation on
dominant sources of noise on ships. Propeller noise propellers in non–uniform inviscid flow was coupled with
becomes the only traceable signal with sonars particularly a hydroacoustic model based on the Ffowcs Williams–
for naval surface ships and submarines since all other Hawkings equation. They split the noise signature into
sources can be eliminated by appropriate insulation thickness and loading term contributions. They
methods. Nevertheless, it determines the detectability, demonstrated that noise predictions by the FWH equation
operability and even survivability of the ship. Therefore, were in satisfactorily agreement with those obtained by
noise predictions for propellers in cavitating and non- using the Bernoulli equation model (Salvatore and
cavitating states have become a momentous subject of Ianniello 2003).
naval architecture for a long time. Barbarino and Casalino (2012) studied and validated
Empirical, semi-empirical methods and Bernoulli-based noise predictions for a NACA-0012 airfoil. Then they
methods have been investigated by many researchers applied the same method to compute the broadband noise
(Testa 200). However, generation of a method by aero- spectrum of an aircraft. Gao et al. (2012) simulated
acousticians Ffowcs Williams–Hawkings (FWH) for numerically the unsteady viscous flow around AUV with
calculation of noise of an arbitrary body moving in a fluid propellers by using the Reynolds-averaged Navier-Stokes
can be considered a mile stone in acoustic predictions (RANS) equations, shear-stress transport (SST) k-
(Ffowcs et. Al. 1969). With the development in model and pressure with splitting of operators (PISO)

* Corresponding author e-mail: [email protected] 113

algorithm based on sliding mesh. The hydrodynamic dependent pressure data is used as the input for the FWH
parameters of AUV with propellers such as resistance, equation to predict far-field acoustics.
pressure and velocity reflected well the real ambient flow 2.1 Numerical Methods and Flow Solver
field of AUV with propellers. Then, the semi-implicit For the numerical calculations ANSYS 13 Fluent is used
method for pressure-linked equations (SIMPLE) to satisfy the following governing equation for continuity
algorithm is used to compute the steady viscous flow field Alin et. al. (2010);
of AUV hull and propellers, respectively. The 𝜕𝜌 𝜕
computational results agree well with the experimental + (𝜌𝑣𝑖 ) = 0 (1)
𝜕𝑡 𝜕𝑥𝑖
data, which shows that the numerical method has good where 𝑥𝑖 and 𝒗𝒊 are the tensor form of axial coordinates
accuracy in the prediction of hydrodynamic performance and velocities, respectively. Then the momentum equation
(Gao et al. 2012). On the other hand, after the 22nd ITTC becomes;
Workshop on Propeller RANS/Panel Methods, a number
of studies have been published (ITTC 2008). In 2004, 𝜕(𝜌𝑣𝑖 ) 𝜕(𝜌𝑢𝑖 𝑢𝑗 )
Kawamura et. al. (2004) comparatively analysed different + = (2)
𝜕𝑡 𝜕𝑥𝑗
turbulence models for the prediction of open water 𝜕𝑝 𝜕 𝜕𝑢𝑖 𝜕𝑢𝑗 2 𝜕𝑢𝑙 𝜕 ̅̅̅̅̅
performance for a conventional propeller. Later Li − + [𝜇 ( + − 𝛿𝑖𝑗 )] + (−𝜌𝑢 𝑖 𝑢𝑗 )
′ ′
𝜕𝑥𝑖 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝑥𝑖 3 𝜕𝑥𝑙 𝜕𝑥𝑗
published his results of estimating open water ̅̅̅̅̅̅
′ ′
where 𝛿𝑖𝑗 is Kronecker Delta and −𝜌𝑢 𝑖 𝑢𝑗 are the
characteristics of a highly skewed model propeller
unknown Reynolds stresses.
employing k-ω turbulence model and validation study
with experimental data (Li 2006). The detailed literature For the turbulence modelling, SST 𝑘 − 𝜔 turbulence
review on the prediction of open water performance of model is employed due to its good performance on wall
propellers can be found in 26th ITTC (2011). bounded boundary layer flows (Li 2006).
In the above context a study has been carried out to FLUENT employs cell-centered finite volume method.
investigate the prediction of propeller noise for RANS formulation is used with absolute velocity
submarines and underwater vehicles. The main objective selection. Transient solution is performed with second
of the study is to obtain accurate propeller noise order implicit pressure based solver. Velocity and
prediction and to use this information to control noise on pressure coupled via SIMPLE algorithm. Green Gauss
submarines. This paper presents the results of the study Node Based is used for gradient and PRESTO for
following the previous research project (Ozden et. al. pressure discretizations. For Momentum, Turbulent
2012 & Korkut et. al. 2013). Firstly a validation study has Kinetic Energy and Specific Dissipation Rate
been carried out for the prediction of DARPA Suboff calculations, QUICK scheme is selected.
generic submarine model’s hydrodynamic drag force and
wake properties in comparison with the experimental
2.2 Noise Predictions
results given in Liu and Huang (1998) and Chase (2012),
Ffowcs Williams – Hawkings Method
respectively. Calculations were continued with the
validation of open water hydrodynamic characteristics of Ffowcs Williams Hawkings (FWH) equation is an
E1619 generic submarine propeller in comparison with inhomogeneous wave equation derived from the
the experimental results published by Di Felici et al. continuity and Navier-Stokes equations (ANSYS 2009).
(2009). Simulations were performed for DARPA Suboff
fitted with E1619 propeller under self-propelled 1 𝜕 2 𝑝′ 𝜕2 𝜕
𝑎02 𝜕𝑡 2
− ∇2 𝑝′ = 𝜕𝑥𝑖 𝜕𝑥𝑗
{𝑇𝑖𝑗 𝐻(𝑓)} − 𝜕𝑡 {[𝑃𝑖𝑗 𝑛𝑗 +
condition. Results were compared with the ones carried
𝜕
out by Chase and Carrica (2013). Acoustic calculations 𝜌𝑢𝑖 (𝑢𝑛 − 𝑣𝑛 )]𝛿(𝑓)} + 𝜕𝑡 {[𝜌0 𝑣𝑛 + 𝜌(𝑢𝑛 − 𝑣𝑛 )]𝛿(𝑓)}
were performed for E1619 propeller in open water
(3)
condition, inlet induced DARPA Suboff wake condition
where
and self-propulsion condition.
𝑢𝑖 : flow velocity in 𝑥𝑖 direction
The method used for the noise prediction is given in
Methodology Section. Results of the study are included. 𝑢𝑛 : flow velocity normal to the surface (𝑓 = 0)
Finally some conclusions withdrawn from the study are 𝑣𝑖 : surface velocity component in 𝑥𝑖 direction
also given.
𝑣𝑛 : surface velocity component normal to the
surface
2 METHODOLOGY 𝛿(𝑓): Dirac delta function
Flow around a propeller is solved using a RANS solver
𝐻(𝑓): Heaviside function
with the SST k- turbulence model. Then, transient
solution is performed with second order implicit pressure P’, is the far-field sound pressure (𝑝′ = 𝑝 − 𝑝0 ). 𝑓 = 0 is
based solver. Velocity and pressure coupled via SIMPLE a mathematical surface used to facilitate the application of
algorithm Numerical Methods and Flow Solver. Time the generalised function theory and the free-space Green
function to the unbounded space exterior flow problem

114
( 𝑓 > 0 ) to reach the solution. This surface (𝑓 = The expressions in the equations denoted by a subscript
0) represents a source (emission) surface and can be used are elements of vectors and unit vectors. For example 𝑟⃗̂
as a suface overlapping with the body (impermeable) or and 𝑛⃗⃗ denote the unit vectors in radiation and wall-normal
as a permeable surface far from the body. 𝑛𝑖 is the unit directions in 𝐿𝑟 = 𝐿⃗⃗ ∙ 𝑟⃗̂ = 𝐿𝑖 𝑟𝑖 and 𝑈𝑛 = 𝑈⃗⃗ ∙ 𝑛⃗⃗ = 𝑈𝑖 𝑛𝑖 .
normal vector indicating the exterior region (𝑓 = 0). 𝑎0 is The dot over a variable denotes source-time
the sound velocity in the far field and 𝑇𝑖𝑗 is the Lighthill differentiation of that variable (ANSYS 2009).
tensor defined as below (ANSYS 2009);
𝑇𝑖𝑗 = 𝜌𝑢𝑖 𝑢𝑗 + 𝑃𝑖𝑗 − 𝑎02 (𝜌 − 𝜌0 )𝛿𝑖𝑗 (4) 3 VERIFICATION AND VALIDATION
The methodology and the software used have been tested
𝑃𝑖𝑗 is the compressive stress tensor. For a Stokesian fluid for a fisheries research vessel propeller (Korkut et. al.
it is defined as follows; 2013). Atlar et.al (2001) presented the results of
𝜕𝑢 𝜕𝑢 2 𝜕𝑢 cavitation tunnel tests carried out with model propeller of
𝑃𝑖𝑗 = 𝑝𝛿𝑖𝑗 − 𝜇 [𝜕𝑥 𝑖 + 𝜕𝑥𝑗 + 3 𝜕𝑥𝑘 𝛿𝑖𝑗 ] (5) a Sumitomo Heavy Industries Fisheries Research Vessel
𝑗 𝑖 𝑘
(FRV) and those of noise measurements with its full-scale
The free-stream quantities are denoted by the subscript 0. propeller to validate the low-noise performance of this
propeller. The tests involve the simulation of a target
The solution of Eq. (4) is obtained by the use of the free-
wake using a wake screen and the determination of the
space Green function (𝛿(𝑔)/4𝜋𝑟). The complete solution
noise behind the simulated wake.
includes the surface and volume integrals. While the
surface integrals include the effects of monopole, dipole
and partially quadropole effects the volume integrals
include only the quadropole sources except the regions of
the source surface. In cases where the flow is in low
subsonic region the value of the volume integral value
diminishes and the source surface encloses the source
region. Thus the volume integrals are not included to the
calculations and the equations below are obtained;

𝑝′ (𝑥⃗, 𝑡) = 𝑝𝑇′ (𝑥⃗, 𝑡) + 𝑝𝐿′ (𝑥⃗, 𝑡) (6)

where;
Figure 1 Fisheries Research Vessel propeller model
4𝜋𝑝𝑇′ (𝑥⃗, 𝑡) =
𝜌0 (𝑈𝑛̇ +𝑈𝑛̇ ) 𝜌0 𝑈𝑛 {𝑟𝑀̇𝑟 +𝑎0 (𝑀𝑟 −𝑀2 )}
∫𝑓=0 [ 𝑟(1−𝑀 2 ] 𝑑𝑆 + ∫𝑓=0 [
𝑟 2 (1−𝑀𝑟 )3
] 𝑑𝑆
𝑟) A 4-bladed fisheries research vessel propeller was
(7) investigated where cavitation tunnel hydrodynamic
4𝜋𝑝𝐿′ (𝑥⃗, 𝑡) = performance and hydro-acoustics tests were conducted in
1 𝐿̇𝑟 𝐿𝑟 −𝐿𝑀
∫ [
𝑎0 𝑓=0 𝑟(1−𝑀𝑟 )2
] 𝑑𝑆 + ∫𝑓=0 [𝑟 2 (1−𝑀 2 ] 𝑑𝑆 + Emerson Cavitation Tunnel in University of Newcastle
𝑟)
1 𝐿 {𝑟𝑀 ̇ (𝑀
+𝑎0 𝑟 −𝑀 2 )} upon Tyne, UK. During the tests, the target wake velocity
∫ [ 𝑟 𝑟𝑟2 (1−𝑀
𝑎0 𝑓=0 3 ] 𝑑𝑆 ratios (Va/Vs) were simulated using different size of
𝑟)
(8) meshes attached to a square frame and the wake velocities
where; were measured using a pitot static tube comb and
𝜌 “scanivalve” assembly. Main particulars of the propeller
𝑈𝑖 = 𝑣𝑖 + 𝜌 (𝑢𝑖 − 𝑣𝑖 ) (9) is given in Table 1
0
𝐿𝑖 = 𝑃𝑖𝑗 𝑛̂𝑗 + 𝜌𝑢𝑖 (𝑢𝑛 − 𝑣𝑛 ) (10) Table 1 Main particulars of propeller
When the integration surface is overlapping with a closed
Number of Blades, Z 4
wall (impenetrable wall) the expressions 𝑝𝑇′ (𝑥⃗, 𝑡) and
𝑝𝐿′ (𝑥⃗, 𝑡) from Equations (7) and (8) are referred as the Propeller Diameter, D 0.30 m
thickness noise term and the loading noise term (ANSYS
Pitch Ratio at 0.7R, P/D 0.8464
2009). The thickness noise term express the noise
generated by the displacement of the flow and the loading Expanded Blade Area Ratio, 0.55
noise term express the noise generated by the thrust AE/A0
generated by the rotation of the blade (Test 2008). The Boss Ratio, rh/R 0.276
terms in brackets in Equations (7) and (8) express that the
Rake 0 Degrees
kernels of the integrals are solved for the retarded time
steps (𝜏) expressed as in Equation (11) where t is time Skew 40 Degrees
and r is the observer distance; Direction of rotation Right handed

𝑟 To take into account the influence of the wake

𝜏=𝑡− (11) characteristics, velocity profile of the wake was
𝑎0

115
implemented to velocity inlet of the calculation domain to Concerning the noise prediction there is a good agreement
simulate the non-uniformity. Figure 2 shows the measured between predicted results with the experiments for the
wake and adaptation of the velocity inlet. frequency range of 200 Hz to 5000Hz. For lower and
Both cavitation tunnel tests and CFD calculations were higher ranges RANS over predicted the flow noise.
conducted for the 0.3m diameter propeller turning with
978 RPM at an advanced speed of 3m/s and the density of
the water in the tunnel was 1002 kg/m3.

1.5 NON-DIMENSIONAL AXIAL VELOCITY DISTRIBUTION

350

10
0
340

20
33

V/Vs

30
1
0
32

0.600 1.000

40
31
0

0.900
0.7

0 50
00

30 0.800
0.6

00
0 60
0.800
0.5
0
00

0.700
50

0.5
0.

290 0.900
70 0.600
0.7

0.500
00

280 0.
80
r/R

80

0.900
0 Figure 3 Comparison of CFD results with the experiment at
0 270 90
J=0.61
260 100
110
-0.5 250
0 12
24 13
0 4 GEOMETRIES OF BODIES
0
23 0
4.1 DARPA Suboff
14
0
22

-1
15

0
0

The DARPA Suboff AFF8 is a generic submarine model

21

160
200

170

0
190
180

geometry with a length of 4.36m sum of a forebody of

-1.5 1.02m, a midbody of 2.23m and an aftbody of 1.11m. It
-1 0 1 2
r/R has cylindrical cross-section with a maximum diameter of
0.508m. AFF8 has a sail which is located at the top dead
center of the hull starting at x=0.92m from the bow and
ends at x=1.29m. It has cross shaped rudder design where
rudders and hydroplanes are located at x=4m. The hull
and appendage arrangement of DARPA Suboff AFF8 is
given in Figure 4 and main particulars are given in Table
3. Views of the calculation domain and DARPA suboff
are shown in Figure 5 and Figure 6, respectively.

Figure 4 Geometry of DARPA Suboff AFF8

Figure 2 Measured wake (top) and adapted wake (bottom)
Calculation was performed for the J value 0.61. Unsteady Table 3 Main particulars of DARPA Suboff AFF8
flow solver was employed for the prediction of
Generic Submarine Type DARPA Suboff AFF8
performances of the propeller in wake condition. Results Description Symbol Magnitude
of KT, KQ and η0 values are given in the Table 2 in Length overall Loa 4.356 m
comparison to the experimental values. Length between perpendiculars Lpp 4.261m
Table 2 Comparison of CFD results with experiment Maximum hull radius Rmax 0.254 m
Centre of buoyancy (aft of FB 0.4621 Loa
J KT 10KQ η0 nose)
Experiment 0.61 0.12 0.17 0.66 Volume of displacement ∇ 0.718
Wetted Surface Swa 6.338
CFD 0.61 0.11 0.19 0.59
% Error 1.12 8.93 11.69 DARPA Suboff AFF8 hull was investigated
experimentally by Li & Huang (1998) for Re=12x10 6.
KT values were predicted with a better accuracy than KQ
values. η0 values were also affected by the prediction of
KQ values as expected.

116
 Table 4 Main particulars of INSEAN E1619 submarine
propeller
Propeller Type INSEAN E1619
Advanced Speed 1.68 m/s
RPM 280 RPM
Diameter 0.485 m
Number of Blades 7
AE/A0 0.608
Hub/Diameter Ratio 0.226
Pitch to Diameter ratio, P/D at 0.7R 1.15
Figure 5 Calculation domain of DARPA Suboff

Figure 6 Cut away views of DARPA Suboff Figure 8 Detailed fine mesh for INSEAN E1619

4.2 INSEAN E1619 Submarine Propeller 5 VALIDATION OF THE METHODOLOGY IN UNIFORM

The INSEAN E1619 generic submarine propeller has FLOW
been analysed in open water, within artificial wake and
5.1 Performance Prediction
propelling DARPA Suboff AFF8 conditions. The
Open water performance calculations of E1619 propeller
propeller was a seven-bladed highly skewed submarine
was carried out at J=0.74 and J=0.85. A computation
propeller with an unloaded tip blade design. Open water
domain has been generated with a 1.5m diameter where
experiments were performed in the INSEAN towing tank,
1.5m region left prior and 3m after the propeller
and wake velocity measurements were made, using LDV
geometry. A mesh independence study was performed
in the large circulating water channel of INSEAN. Results
from coarse to fine meshes, using 6,386,638, 8,065,679
were presented by Di Felice (2009). 3-D views and the
and 10,513,205 cells, respectively with non-dimensional
main particulars of the E1619 submarine propeller are
wall distance value of y+≈50.
given in Figure 7 and in Table 4, respectively. In Figure 8
the detailed fine mesh used for the study is presented. The convergence of grid study can be seen in Figure 9
with above grid properties in comparison with the
experimental values at J=0.74.

Figure 7 3-D views of INSEAN E1619 propeller

117
 Figure 9 Comparison of convergence of CFD values and Figure 12 Comparisons of noise predictions for different
experiment results from coarse to fine mesh receiver positions at J=0.74
Results of the fine mesh are presented in Figure 10. It can
be seen that RANS calculations very well captured the 6 FLOW CALCULATIONS FOR DARPA SUBOFF
thrust and torque values at given advanced ratios. Post processes of the solution are also performed by
ANSYS 15 FLUENT by using Fast Fourier Transform
properties. Hanning filter was used and results are
presented for Sound Pressure Level (dB) form for 1/3
Octave band.
Before computing self-propulsion a set of computations
were performed for DARPA Suboff towed in fully
appended configuration without propeller, and results
were shown in Figure 13 in comparison with the
experiment results (Liu and Huang 1998). Computations
were carried out in a cylindrical domain with a diameter
of 9m and a region of 4.5m prior to model geometry and a
9m of after the model. Two different grids were generated
with 7,812,122 and 11,511,137 cells where y+≈50.
Results indicated that the resistance predicted by RANS
Figure 10 Comparison of open water characteristics computations was in a good agreement with the
calculation for E1619 with experimental results by Di Felice experiments.
et. al. (2009)
5.2 Noise Prediction in Open Water
After the steady computations which were performed for
the performance predictions of E1619 generic submarine
propeller, calculations were carried out transiently to
predict the noise characteristics.
The domain which contains the propeller body was
selected as the source of noise and a number of receiver
locations was defined for the calculation. For the
calculations performed for E1619 propeller in open water
condition acoustic predictions were made for a total of 6
receivers located 1m away from the propeller reference
point (0,0,0) and 0o, 45o and 90o angles from the shaft axis Figure 13 Comparison of resistance predicted by CFD
(x,0,0). Position of receivers can be seen in Figure 11. method with experiment (Liu and Huang (1998)
Figure 12 shows the noise predictions for different
receivers at J=0.74. 7 NOISE PREDICTIONS
Noise predictions were performed in simulated (induced)
wake condition and behind DARPA Suboff condition.
7.1 Propeller in Simulated (Induced) Wake
RANS calculations were conducted in order to simulate
the submarine moving straight ahead. The
computationally predicted wake at the location x/L=0.978
was compared to the wake measured by Crook (1990).

Figure 11 Receiver locations for E1619 propeller

118
The hole at the middle of each figure is due to the body of
the submarine. As can be seen from the figures overall
pattern of the wake is very similar. Reduced flow velocity
can be traced downstream of the rudders and hydroplanes,
higher wake field which can observed at the top is due the
wake of the sail. Figure 15 3-D view of DARPA Suboff with E1619 propeller
fitted

CFD

Figure 14 Measured and computed wake of DARPA Suboff

AFF8 configuration
The “V” shaped higher velocity field on the top side
which differs from sides and bottom is because of the
horseshoe vortices which deplete the boundary layer at
the center and send low momentum flow to the sides. In
the experimental wake data, the low velocity field
between lower rudder and bottom is due to presence of
struts which were used during the wind tunnel tests.
For the computations a grid with 8.5m region prior and
9m after the propeller geometry was left and 9,834,381
cells used where y+≈50.

7.2 Propeller behind DARPA SUB OFF

DARPA Suboff AFF8 propelled by the E1619 propeller
has been studied by several authors for self-propulsion
(Chase 2012, Chase and Carrica 2013), cycle-to-cycle
blade loading (Liefvendahl and Töerng 2011) and hull
interactions (Alin. et al. 2011). Similar to the listed
studies, the geometry of DARPA Suboff was kept
constant and E1619 propeller diameter was scaled to Figure 16 Calculation domain and cut away views of the
0.262m. bodies
Simulation of the DARPA Suboff AFF8 fitted with the
E1619 propeller under self-propelled conditions was
performed using RANS. Geometric details are similar to
those listed in Table 3.
Computations were carried out in a cylindrical domain
with a diameter of 9m and a region of 4.5m prior and a
9m after the model geometry. The grid was with
19,777,345 cells where y+≈50. Figure 15 shows the
DARPA Suboff with E1619 propeller fitted and Figure 16
shows the calculation domain and cut away views. Figure
17 also shows the perspective view from the stern where Figure 17 Axial velocity cross-sections around DARPA
contour plots of the axial velocities are shown on Suboff with E1619
transverse and longitudinal cross-sections along the hull.

8 PRESENTATIONS AND DISCUSSIONS OF RESULTS

In Figure 18 longitudinal cross-section views of both self-
propelled and within induced wake conditions of DARPA
Suboff and INSEAN E1619 propeller were presented. The

119
main difference between two cross-sections is velocity
profiles after the hub region which is due to the blunt hub
geometry at the absence of the submarine body. This
affects the region about two diameters long behind the
propeller hub with a wide boundary layer. This can also
be traced on the transverse cross-section views of the
behind propeller region which can be seen from Figure 19
to Figure 22 with 0.005 x/L intervals.

Figure 19 Wake at x/L=1.005

Figure 20 Wake at x/L=1.010

Figure 18 Cross section plan view of the wake for self-

propelled and induced wake conditions
The results of the analysed conditions are presented in
Figures 23 through 25. In each graph: the logarithmic- Figure 21 Wake at x/L=1.015
scaled x-axis represents the centre frequencies in Hz,
while the linear-scaled y-axis represents the sound
pressure levels in dB re 1 µPa (standard reference
pressure for water), 1 Hz, 1m.
Although the characters of the SPL curves of both
conditions are almost similar, there is a difference in
values of receivers located behind the propeller. This is
not identical for the second receiver which is located in
the front of the propeller plane. Computations predict
almost the same SPL spectra in the vortically undisturbed
region (in front of the propeller) but predicting shifting
values on dB scale for the receivers located behind the
propeller. This difference between the receivers in two Figure 22 Wake at x/L=1.020
computations is probably also due to the blunt hub shape.

120
 propeller was also selected for the study. A number of
validation studies and mesh independence studies were
performed to verify the predictions of experimentally
measured drag & wake data of submarine and thrust &
torque data of the propeller. Noise calculations were
performed for two cases; in the first case, calculated wake
profile was induced at the inlet of the domain where
propeller was rotated and, in the second case
computations performed for the propeller rotating behind
the submarine model in self-propelled condition. Some
conclusions drawn from the study are as follows:
 In the validation case, experimental and
Figure 23 Underwater noise predictions for varying receiver computational noise data showed good
positions in the case of self-propelled DARPA Suboff by agreement between 200Hz to 5000Hz. For lower
E1619 propeller and higher frequencies, CFD over-predicted the
noise.
 Steady computations of drag and wake over
DARPA Suboff AFF8 fully appended body
showed a good agreement with experimental
measurements.
 Open water performance of the INSEAN E1619
propeller was well predicted, which was a
preparatory study prior to noise calculations.
 During the induced wake computations, the blunt
hub shape at the absence of submarine body
caused a wider boundary layer behind the
Figure 24 Underwater noise predictions for varying receiver propeller comparing the self-propelled condition.
positions in the case of DARPA Suboff with induced wake A more streamlined hub geometry could
terminate this difference.
 Similar to the validation case where a highly
skewed four-bladed surface ship propeller was
used, the two computations on INSEAN E1619
showed good agreement between 200Hz to
5000Hz while using almost half the mesh
elements for induced wake condition which
changed the computation time significantly.
 Experimental noise data is publicly published for
neither DARPA Suboff submarine body, nor
INSEAN E1619 submarine propeller. Future
studies may include the experimental
Figure 25 Comparison of underwater noise predictions for measurements of noise characteristics for hull
self-propelled DARPA Suboff by E1619 propeller with those only, propeller in open water conditions,
by induced wake at receiver2 propeller within wake and self-propelled
conditions.
9 CONCLUSIONS Studies focused on predicting the flow details around the
A validation study was conducted in a model scale for model scale submarine and propeller. This is due to the
highly skewed fisheries research vessel propeller. The published experimental data regarding the Suboff models
predicted noise, using RANS and employing Ffowcs and E1619 propeller. Community seeks high quality
Williams-Hawkings method, was compared to the experimental data for full scale submarines which can be
experimental data which was measured in Emerson used for validation of computational noise predicting
Cavitation Tunnel of Newcastle University. Calculations methods.
were extended to predict the noise characteristics for a
submarine propeller.
ACKNOWLEDGEMENTS
DARPA Suboff AFF8 fully appended body was chosen as Authors would like to thank Prof M. Atlar of Newcastle
a model scale submarine geometry and INSEAN E1619 University for sharing the geometry, wake and noise data
highly skewed 7-bladed generic model scale submarine of Sumitomo Fisheries Research Vessel. We would also

121
 like to thank M. Felli and F. Salvatore of INSEAN for Groves, N., Huang, T. & Chang, M. (1998) ‘Geometric
sharing the 3D geometry of INSEAN E1619 submarine Characteristics of DARPA SUBOFF Models (DTRC
propeller for use in our academic studies. The authors are Model Nos. 5470 and 5471’. David Taylor Research
grateful to Assist. Prof. O. Unal and Prof. O. Goren of Center Report, Report No. DTRC/SHD-1298-01,
ITU for useful discussions and use of the high speed March.
computer at Numerical Hydrodynamics Laboratory. ITTC. (2008) ‘Propulsion Committee Report’.
Authors also wish to acknowledge Mr. B. Can of Wichita Proceedings of the 25th ITTC, 25th International
State University for his invaluable help and E. Demir Towing Tank Conference, Fukuako, Japan, , Vol. I,
from DATUM Ltd for preparation of 3D geometries. pp. 83-141.
ITTC. (2011) ‘Specialist Committee on Computational
REFERENCES Fluid Dynamics Report’. Proceedings of the 26th
Atlar, M., Takinaci, A.C., Korkut, E., Sasaki, N. & Aono, International Towing Tank Conference. Rio de
T., (2001). ‘Cavitation Tunnel Tests for Propeller Janeiro, Brazil, pp. 337-375.
Noise of a FRV and Comparisons with Full-Scale Kawamura, T., Watanabe, T., Takekoshi, Y., Maeda, M.,
Measurements’, 4th International Symposium on Yamaguchi, H. Numerical (2004) ‘Simulation of
Cavitation CAV2001, Pasadena, California, USA Cavitating Flow around a Propeller’ JSNA. 195.
Alin, N., Chapuis, M., Fureby, C., Liefvendahl, M., Korkut, E., Ozden, M.C., Avci, G.A, Goren, O., Takinaci,
Svennberg, U. & Troeng, C. (2010). ‘A Numerical A.C. & Danisman, D.B. (2013) ‘Numerical
Study of Submarine Propeller-Hull Interaction’. Calculations of Noise Characteristics of Surface and
Proceedings of the 28th Symposium on Naval Underwater Navy Ships Generated by Propeller’.
Hydrodynamics, Pasadena, California. Turkish Scientific and Research Council (TUBITAK )
ANSYS (2009) ANSYS 12 FLUENT Theory Guide. Project, Project No: 110M327.
Barbarino, M. & Casalino, D. (2012) ‘Hybrid Li, D.-Q. (2006) ‘Validation of RANS Predictions of
Analytical/Numerical Prediction of Propeller Open Water Performance of a Highly Skewed
Broadband Noise in Time Domain’. Int. Jour. of Propeller with Experiments’. Conference of Global
Aeronautics, 11(2), pp.157–175. Chinese Scholars on Hydrodynamics, Journal of
Bull, P., (1996). ‘The Validation of CFD Predictions of Hydrodynamics, 18(3), Supplement, pp.520-528.
Nominal Wake for the SUBOFF Fully Appended Liefvendahl, M., Toerng, C. (2011 ) ‘Computation of
Geometry’. 21st Symposium on Naval Cycle-to-Cycle Variation in Blade Load for a
Hydrodynamics, June, pp.1061– 1076. Submarine Propeller, using LES’. 2nd International
Chase, N., (2012) ‘Simulations of the DARPA Suboff Symposium on Marine Propulsors, SMP11, Hamburg,
Submarine Including Self-propulsion with the E1619 Germany.
Propeller’. M.Sc. Thesis, University of Iowa, USA. Liu, H.-L. & Huang, T. (1998). ‘Summary of DARPA
Chase, N. & Carrica, P. M. (2013). ‘Submarine Propeller SUBOFF Experimental Program Data’. Report No.
Computations and Application to Self-propulsion of CRDKNSWC/HD-1298-11, June.
DARPA Suboff’. Ocean Engineering, 60 pp.68-80. Ozden, M.C., Avci, G.A. & Korkut, E. (2012). ‘A
Crook B. (1990) ‘Resistance for DARPA Suboff as Numerical Study on Prediction of Noise
Represented by Model 5470’. David Taylor Research Characteristics Generated By a Propeller’. 10th
Center Report, Report No: DTRC/SHD-1298-07. International Conference on Hydrodynamics –
ICHD2012, St. Petersburg, Russia.
Di Felice, F., Felli, M., Liefvendahl, M. & Svennberg, U.
(2009) ‘Numerical and Experimental Analysis of the Salvatore, F. & Ianniello, S. (2003) ‘Preliminary Results
Wake Behavior of a Generic Submarine Propeller’. on Acoustic Modelling of Cavitating Propellers’.
Proceedings of the 1st International Symposium on Comp. Mechanics, 32, pp.291–300.
Marine Propulsors, Trondheim, Norway. Seol, H., Suh, JC, & Lee, S. (2002) ‘Prediction of Non-
Ffowcs Williams JE, Hawkings DL (1969) ‘Sound Cavitating Underwater Propeller Noise’. J. Sound and
Generation by Turbulence and Surfaces in Arbitrary Vibration, 257(1), pp.131-156.
Motion’. Philosophical Transactions of the Royal
Seol, H., Suh, JC. & Lee, S. (2005) ‘Development of
Society A264: 321–342.
Hybrid Method for the Prediction of Underwater
Gao, FD., Pan, CY. & Han, YY. (2012) ‘Numerical Propeller Noise’. J. Sound and Vibration, 288(1),
Computation and Analysis of Unsteady Viscous Flow pp.345-360.
Around Autonomous Underwater Vehicle with Testa, C. (2008). ‘Acoustic Formulations for Aeronautical
Propellers Based on Sliding Mesh’. J. of Central South and Rotorcraft Noise Prediction Based on the Ffowcs
University of Technology. 19(4),pp.944–952. Williams and Hawkings Equation’. Ph.D. Thesis,
Universita degli Studi di ROMA TRE, Italy.

122

View publication stats