SEPTEMBER 11 2016

Numerical and experimental predictions of underwater

propeller radiated noise 
Taehyung Kim; Jonghoon Jeon; Sunghan Chu; Sunghoon Kim; Wonho Joo

Proc. Mtgs. Acoust. 28, 070004 (2016)

https://doi.org/10.1121/2.0000430

CrossMark

 
View Export
Online Citation

Articles You May Be Interested In

Analysis of entry of additional energy to gunpowder in electrothermal chemical shot

AIP Conference Proceedings (November 2017)

Compact pulse generators with soft ferromagnetic cores driven by gunpowder and explosive
Rev. Sci. Instrum. (December 2015)

Acoustical archaeology - Recreating the soundscape of John Donne's 1622 gunpowder plot sermon at
Paul's Cross

09 October 2023 06:42:37

Proc. Mtgs. Acoust. (June 2013)
Volume 28 http://acousticalsociety.org/

22nd International Congress on Acoustics

Acoustics for the 21st Century
Buenos Aires, Argentina
05-09 September 2016

Underwater Acoustics: Paper ICA2016 - 35

Numerical and experimental predictions of

underwater propeller radiated noise
Taehyung Kim and Jonghoon Jeon
Hyundai Heavy Industries, Seoul and Ulsan, Republic of Korea; [email protected]; [email protected]

09 October 2023 06:42:37

Sunghan Chu, Sunghoon Kim and Wonho Joo
Hyundai Heavy Industries, Ulsan, Republic of Korea; [email protected]; [email protected];
[email protected]

The underwater propeller generates harmful acoustic impacts to marine environments. The propeller tonal
noise was numerically predicted by the unsteady volume variation of sheet cavitation on the blades. This
procedure was validated with model propeller tests in the water tunnel facility. Two experimental
methods are developed for propeller broadband noise prediction. The first method is model scale
estimation through development of semi empirical formula based on tip vortex formation noise theory.
The final form of this formula was modified by subsequent model tests for nine kinds of propellers with
considering the transfer function of cavitation tunnel. The second experimental approach is a full scale
prediction based on transmission coefficient in the stern tube area. The relationship between fluid borne
sound and structure borne sound was obtained through acceleration measurements in the steering gear
room, while impulsive sound was generated by the gunpowder near the propeller during the final stage of
ship construction in dry dock. Finally, the model scale and full scale experimental methods were verified
by underwater radiated sound measure at sea trial and showed good agreement with measured ship's
noise data.

Published by the Acoustical Society of America

© 2017 Acoustical Society of America [DOI: 10.1121/2.0000430]

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 1
T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

1. INTRODUCTION
Underwater propeller noise is directly related to the survivability of naval vessels and is also
closely connected with quality assessments and ship orders for commercial vessels [1]. The
report of the International Maritime Organization stated that noise in the low frequency range of
10 Hz to 1 kHz has negative effects on the marine ecosystem [2]. Accurate prediction of noise
levels of surface ships is an essential factor in preparing a noise control policy and a noise
abatement plan. Numerical investigations of propeller cavitation noise have been mainly focused
on the prediction of blade rate noise related to sheet cavitation. Acoustic prediction using the
Ffowcs Williams and Hawkings equation [3] based on the Lighthill’s acoustic analogy [4] was
performed by Seol et al.[5] They modeled the sheet cavitation volume on the blade as a point
source of thickness noise. Ianniello et al. [6] predicted the cavitation noise using RANS
(Reynolds averaged Navier Stokes) simulation and the acoustic analogy. Hsiao et al. [7] focused
on the tip vortex cavitation inception noise based on the Rayleigh-Plesset equation of bubble
dynamics. Seong et al. [8] calculated hull pressure that fluctuates through non-cavitating
propeller blade loading and thickness noise based on an acoustic boundary element method.
Prediction of propeller broadband noise has been performed mainly using model experiments
and full scale measurements. Yamada et al. [9] and Oshima et al. [10] suggested prediction
methods for tip vortex cavitation noise by experimental deduction of tip vortex cavitation
inception and its scaling exponent. Jeon et al. [11] suggested an indirect prediction method using
sound transmission coefficient measurements in a dry dock and subsequent on-board

09 October 2023 06:42:37

measurement of structure-borne sound at the sea trial. They compared the prediction results with
other commercial vessels’ noise data regarding only the overall sound level.
In this paper, a prediction method of propeller cavitation noise on the basis of a numerical
analysis and an experimental approach has been brought forward. The numerical prediction of
propeller tonal noise was performed with a distributed cavity source model of a propeller blade.
Then a semi-empirical formula for cavitation broadband noise was developed based on the
theory of tip vortex formation noise and was optimized by water tunnel tests with nine kinds of
model propellers. The sound transmission coefficient method was also carried out to predict
cavitation noise for a real ship. This method was validated by the underwater radiated noise
measurement in deep water. The test vessel used in this paper was a commercial 14,200 TEU
container vessel built by Hyundai Heavy Industries in 2015 with a five-bladed propeller.

2. NUMERICAL APPORACH
A. CAVITATION FLOW ANALYSIS
We analyzed a numerical model with a propeller, shaft, rudder and a wake screen for
comparison of the numerical solution and water tunnel test results. The computational domain in
our numerical model was composed of a square cross section of 0.6m x 0.6m and was extended
from the propeller rotation plane to the upstream and the downstream location as 0.6m and 1.2m,
respectively. Figure 1 shows the computational domain used for this analysis in detail. The
software used in this paper was a STAR-CCM+ 10.06, and Spalart-Allmaras 1-equation
turbulence model was applied with sliding polyhedral mesh. The seed density and seed diameter
for the cavitation model were 1.0×1012/m3 and 1.0×10-6 m, respectively. The number of grid
points for the blade was 6.3×105 while the number of grid points for the surround domain was
4.2×105. The symmetry condition was imposed at all boundaries except the inlet and outlet

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 2

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

boundaries. The velocity inlet boundary condition and the initial velocity condition were applied
to the inlet boundary with the axial wake distribution as shown in Figure 2. The pressure outlet
condition was imposed at the water tunnel outlet boundary.

Figure 1. Computational domain.

0.1
0.9
0.8
0.7

09 October 2023 06:42:37

0.6
Z (m)

0
0.5
0.4
0.3
0.2
-0.1
0.1

-0.2
-0.2 -0.1 0 0.1 0.2
Y (m)

Figure 2. Normalized inlet axial wake distribution.

B. ACOUSTIC PREDICTION
Sound generated by turbulence and moving surfaces in arbitrary motion can be predicted by
the Ffowcs-Williams and Hawkings(FW-H) equation [3] which is based on the Lighthill’s
acoustic analogy [4]. The FW-H equation is the rearranged version of the Navier-Stokes equation
which has an inhomogeneous wave equation form. The source terms are composed of a
monopole source, a dipole source on the body surface, and a quadrupole source in the volume
surrounding the body. The differential form of the FW-H equation can be described as:

(⃗ ) (⃗ ) [ ( )] [ ( )] [ ( )] (1)

where 𝑛 is the local velocity of the body normal to the body surface defined as = 0. In
addition, 𝑖 is components of the local force on the body surface, and ( ) and ( ) are the Dirac
delta function and the Heaviside function, respectively. The monopole source is caused by the

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 3

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

displacement of fluid as the blade passes through it. The sheet cavitation volume on the blade
surface is also considered as an additional term of the monopole source. The dipole source
resulted from unsteady motions of force distribution on the blade surface is also included to
predict propeller loading noise. Therefore the final form of the acoustic prediction formula is

(⃗ ) (⃗ ) (⃗ ) (⃗ ) (2)

where

 0  vn  vn    0vn  rM r  cM r  cM 2  
4 pT  x , t     2 
dS     dS (3)
 r 1  M r  ret r2 1 M r
3
f 0 f 0
  ret

1  lr   l l  1  lr  rM r  cM r  cM 2  
4 pL  x , t    dS    r M
c  f 0 
 2  dS    dS (4)
c f 0  r 1  M r  ret f 0
 r 1  M r
2
 ret r 2
1  M
3
 ret
 r

̈( ) ̇( ) ̇ ̂
(⃗ ) ∫ [ ] ∫ [ ]

09 October 2023 06:42:37

| | | |
(5)
̇( ) ( )
∫ [ ]
| |

Equation (3) represents the acoustic pressure fluctuation related to the monopole source of
thickness noise. Equation (4) is related to the dipole source contributing to loading noise.
Equation (5) represents the acoustic pressure fluctuation at the observation time and position
along the sheet cavitation on the blade surface [5]. The integrands with 1/ are the far-field terms
and those with 1/ 2 are the near-field terms. The subscripts 𝑛, and refer to the dot products
with the unit normal vector, the unit radiation vector, and the surface velocity vector,
respectively. As the volume of sheet cavitation is changed according to the rotation of the blades,
the acoustic pressure fluctuation is stronger when the sheet cavitation moves closer to the
receiver than when the sheet cavitation moves away from the receiver. This phenomenon is
caused by the Doppler effects and is valid for the receiver fixed with the same distance from the
center of the rotation of the blades.
A distributed source model developed in this paper was applied to predict the acoustic field
emitted by a complex pattern of sheet cavitation variation. The point source model and the
distributed source models are illustrated in Figure 3. The point-source model assumes that the
monopole source related to sheet cavitation volume is concentrated on a specific point on the
blade surface. However, the resulting sound pressure level can be different depending on the
choice of source position. The prediction error is also increased when the cavitation volumetric
pattern is complex. Therefore a distributed source model is suggested to describe the sheet
cavitation volume more accurately and to reduce the prediction error for the cavitation tonal
noise. The five representative cavitation volumes were assumed to be the distributed sources
along the blade span from the root of the blade to the blade tip.

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 4

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

Figure 3. Schematics of point-source model (left) and distributed-source model (right).

3. EXPERIMENTAL APPROACH
A. DEVELOPMENT OF SEMI-EMPIRICAL FORMULA THROUGH
WATER TUNNEL TEST
The nine kinds of model propeller tests were performed in the medium sized water tunnel
located in Hyundai Heavy Industries. Table 1 summarizes the types of test propellers which were
designed and manufactured by Hyundai Heavy Industries. The noise measurement was

09 October 2023 06:42:37

performed using a chamber type hydrophone B&K8103. The acoustic distortion effect of the
water tunnel walls was calibrated by measuring transfer function using the underwater
loudspeaker, Electro-Voice UW30, at both the water tunnel and the towing tank (210 m × 14 m ×
6 m).

Table 1. Types of test propellers.

No Vessel No Vessel
1 8,500 TEU Container Carrier 6 174,000 CBM LNG Carrier
2 13,800 TEU Container Carrier 7 105,000 DWT Product Tanker
3 14,000 TEU Container Carrier 8 181,000 DWT Bulk Carrier
4 82,000 CBM LPG Carrier 9 320,000 DWT Crude Oil Carrier
5 154,800 CBM LNG Carrier

According to Baiter [12]’s work, the tip vortex cavitation noise shows the dominant
contribution to the continuous sound spectrum than other cavitation types. Therefore propeller
cavitation broadband noise was assumed to be generated mainly by the tip vortex cavitation
formation. A prediction method in this study for the tip vortex cavitation noise was developed by
modification of the aero-acoustic theory performed by Brooks and Marcolini [13]. As a result of
experimental tests using nine model propellers, the semi-empirical formula for the tip vortex
cavitation formation noise was modified and optimized to the hydroacoustic rotating blades. The
following equation is the final form of the developed formula. In this formula, the turbulent flow
noise component is considered near the propeller tip in the local separating flow. The
contribution terms of the tip vortex core diameter, the blade boundary layer thickness, the
Strouhal number and the Reynolds number are considered as follows:

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 5

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

( ( ) ) (6)

(7)
√( 𝑛 ) ( )

(8)

(9)

√( 𝑛 ) ( )⁄
(10)

where is the Strouhal number and 𝑠 is the vortex shedding frequency. is the vortex
core diameter. and 𝘩𝑖 𝑘 are the vortex core scaling factor and the boundary layer thickness at
the trailing edge of the 0.7r/R location, respectively. The chord length 𝘩 was assumed to be

09 October 2023 06:42:37

the quarter of the propeller diameter. means the Reynolds number of the 0.7r/R location in
the propeller spanwise direction. The other constants were optimized by the water tunnel model
tests in this paper.

( ) (11)

Equation (11) was derived from the combination of the measurement formula by Ross [14]
and the developed tip vortex cavitation prediction formula. Here, is the number of blades,
and are the propeller diameter and the rotation speed in RPS (revolutions per second),
respectively. The scaling law used in this paper was based on the cavitation committee method of
ITTC [15]. This law was derived under the assumption that the cavitation dynamics between
model and full scale was identical. The continuous part of the sound spectrum can be expressed
as follows:

𝑛
(12)
𝑛

𝑛
(13)
𝑛

where is the frequency, 𝑛 is the revolutions per second and is the propeller diameter. The
subscripts 𝑚 and 𝑠 refer to the model scale and the real ship scale, respectively.

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 6

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

B. VALIDATION OF SOUND TRANSMISSION COEFFICIENT METHOD

THROUGH SEA TRIAL
The structure-borne noise measurement was performed at the center of the steering gear
room in the 14,200 TEU container vessel of the sea trial as shown in Figure 4. The measurement
equipment was an accelerometer Dytran type 3148E which has a sensitivity of 100mV/g and
frequency response of ±5% in the range of 0.5 to 5,000 Hz. The sound transmission coefficient
of hull structure was measured for the 13,800 TEU container vessel of Hyundai Heavy Industries
which is similar to the 14,200 TEU container vessel. The experiments in a dry dock were carried
out to measure the sound transmission coefficient of each ship’s stern. The structure-borne sound
in the steering gear room was measured simultaneously while propeller and hull were excited
outside by the impact hammers and the impulsive gunpowder. The hull’s structure-borne sound
transmission coefficients were determined based on the measured signals.

09 October 2023 06:42:37

Figure 4. On board structure-borne sound measurement in steering gear room.

The underwater radiated noise from the real ship at specific operating conditions was
measured with the aid of Korea Research Institute of Ships and Ocean Engineering. The test was
performed using a floating type remote-transmitted hydrophone system in accordance with the
ISO 17208 [16]. The number of total hydrophones was three and the deepest one was located
300 m in depth. Table 2 summarizes the specifications of the test vessel and conditions.
Table 2. Specifications of test vessel: 14200 TEU container vessel

Parameter Value
Shipbuilder Hyundai Heavy Industries
Engine power 65% and 85% of the maximum continuous rating
Measuring site East sea of the Republic of Korea
Water depth 400 m

4. RESULTS AND DISCUSSION

The rate of shaft rotation was constant at 1500 RPM for the water tunnel test. The cavitation
number and thrust coefficients of the model propeller were matched with those of the real ship’s
propeller for the similitude of the small-scale model propeller. The angular position of a key
blade was measured from the vertical upward position in a counter-clockwise direction. The
cavitation patterns of the model test and the corresponding numerical results are shown for a
specific angular position in Figures 5 and 6. The calculated cavitation patterns show good

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 7

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

agreement with the experiments for both 65% and 85% MCR conditions. The sheet cavitation is
located in the spanwise range of 0.6 r/R to 0.99 r/R.
Figures 7 and 8 show the comparison of cavitation tonal noise between the numerical
analysis and the water tunnel measurements for the model propeller of the 14,200 TEU container
carrier. Both the point source model and the distributed source model show good agreement with
the measured sound pressure level up to the third blade passing frequency harmonics within 5 dB
error bound. The prediction accuracy is improved up to 3 dB for the second blade passing
frequency harmonics for the distributed source model rather than the point source model because
the first blade passing frequency sound pressure level is affected mainly by the overall shape of
acoustic waveform while the small fluctuation contributes to the second and higher harmonics.

Figure 5. Photograph of the cavitation patterns of the experimental result (left) and the numerical result
(right): Blade angle 6 degrees; 65% MCR

09 October 2023 06:42:37

Figure 6. Photograph of the cavitation patterns of the experimental result (left) and the numerical result
(right): Blade angle 6 degrees; 85% MCR

Figure 7. Spectrum of propeller cavitation tonal noise; 65% MCR

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 8

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

Figure 8. Spectrum of propeller cavitation tonal noise; 85% MCR

The propeller cavitation broadband noise was predicted using the developed semi-empirical
formula and then converted to the full scale source level by the ITTC law [15]. The propeller
cavitation noise was also predicted using the sound transmission coefficient method. The semi-
empirical formula and the sound transmission coefficient method were used to estimate source
levels of the 14,200 TEU container carrier for 1/3-octave band (Figures 9 and 10). The
developed semi-empirical formula shows good agreement with the experimental data within 5
dB in the frequency range of 300 Hz to 3 kHz. However, the level of difference increases in the

09 October 2023 06:42:37

frequency range of 100 Hz to 200 Hz. The difference is caused by the lack of consideration of
other cavitation noise sources such as hub vortex cavitation and cloud cavitation. The
simultaneous consideration of tip vortex cavitation and other cavitation types is difficult to be
embodied in the water tunnel test because the range of cavitation number is different from each
other. The semi-empirical formula presented in this study provides tools to estimate propeller
cavitation broadband noise levels especially for tip vortex cavitation noise. The indirect sound
transmission coefficient method also shows good agreement with the underwater radiated noise
measurement.

Figure 9. Spectrum of propeller cavitation broadband noise; 65% MCR

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 9

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

Figure 10. Spectrum of propeller cavitation broadband noise; 85% MCR

5. CONCLUSION
This study presented the prediction methods of propeller cavitation tonal noise and
broadband noise. The propeller tonal noise was predicted by the distributed source model based

09 October 2023 06:42:37

on the acoustic analogy, and the accuracy of our prediction methods was verified using the water
tunnel tests. The proposed method shows good agreement with the cavitation tonal noise
measurement within 3 dB error bound.
The semi-empirical formula was developed based on the tip vortex formation theory and was
optimized by the water tunnel test for various propeller types. The underwater loudspeaker was
used to measure the transfer function between the tunnel wall and the towing tank. The tank was
assumed to be a quasi-free field of underwater sound radiation. The propeller cavitation
broadband noise was also estimated by the sound transmission coefficient method through the
dry dock test of a similar class container ship and the on-board structure-borne sound
measurements. The proposed semi-empirical formula and the sound transmission coefficient
method show good agreement with the underwater radiated noise measurement in the frequency
range of 300 Hz to 3 kHz. An additional full-scale measurements for various vessel types help
build on the robust prediction method of underwater propeller radiated noise.

ACKNOWLEDGMENTS
The underwater radiated noise measurement for full scale trial of this work was supported by
the Ministry of Trade, Industry and Energy (MOTIE) (project code: 10045337).

REFERENCES
1
The 27th ITTC, “Specialist Committee on Hydrodynamic Noise,” Copenhagen (2014).
2
IMO/MEPC.1/Circ.833, “Guidelines for the reduction of underwater noise from commercial shipping to address
adverse impacts on marine life” (2014).

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 10

T. Kim et al. Numerical and experimental predictions of underwater propeller radiated noise

3
J. E. Ffowcs Williams and D. L. Hawkings, “Sound generation by turbulence and surfaces in arbitrary motion,”
Philosophi. Trans. Royal Soc. 264, 1151, 321-342 (1969).
4
M. J. Lighthill, “On sound generated aerodynamically. I. General theory,” Proc. Royal Soc. London. A. Math.
Phys. Sci. 211, 1107, 564-587 (1952).
5
H. Seol, “Time domain method for the prediction of pressure fluctuation induced by propeller sheet cavitation:
Numerical simulations and experimental validation,” Ocean Eng. 72, 1, 287-296 (2013).
6
S. Ianniello, R. Muscari and A. Di Mascio, “Ship underwater noise assessment by the acoustic analogy, part III:
measurements versus numerical predictions on a full-scale ship,” J. Marine Sci. Tech. 19, 2, 125-142 (2013).
7
C. T. Hsiao and G. L. Chahine, “Scaling of tip vortex cavitation inception noise with a bubble dynamics model
accounting for nuclei size distribution,” J. Fluids Eng. 127, 1, 55-65 (2005).
8
W. Seong, J. Lee, A. Hyun and J. Lee, “Non-cavitating propeller noise source inversion,” J. Acoust. Soc. Am. 130,
2558 (2011).
9
T. Yamada, K. Sato, C. Kawakita and A. Oshima, “Study on prediction of underwater radiated noise from
propeller tip vortex cavitation,” Proc. 9th Inter. Symp. Cavitation (CAV2015), San Francisco, California, USA, July
18-24 (2015).
10
A. Oshima, “Scaling of tip vortex cavitation noise of propeller,” Mitsubishi heavy industries, Technical review, 31,
3, 115-118 (1994).
11
J. Jeon and W. Joo, “Prediction of propeller radiated noise by on-board measurement,” Proc. 2nd Int. Conf. and
Exhib. Underwater Acoust. (UA2014), Rhodes, Greece, June 22-27, 667-674 (2014).
12
J. J. Baiter, “Advanced views of cavitation noise,” Proc. Int. Symp. Propulsors and Cavitation, Hamburg,
Germany (1992).
13
T. F. Brooks and M. A. Marcolini, “Airfoil tip vortex formation noise,” AIAA J. 24, 2, 246-252 (1986).
14
D. Ross, “Mechanics of underwater noise,” Pergamon Press, Oxford (1976).

09 October 2023 06:42:37

15
ITTC, “Cavitation committee report,” 18th Int. Towing Tank Conf. Kobe, Japan (1987).
16
ISO/PAS 17208-1, “Acoustics – Quantities and procedures for description and measurement of underwater sound
from ships – Part 1: General requirements for measurements in deep water,” (2012).

Proceedings of Meetings on Acoustics, Vol. 28, 070004 (2017) Page 11