Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Journal of Marine Science and Technology not yet been copyediting, typesetting, pagination and proofreading process. Please
DOI: 10.6119/JMST-012-0831-1 note that the publication version of this article may be different from this version.
This article now can be cited as doi: 10.6119/JMST-012-0831-1.
the purpose of analyzing observed real propeller signals. The coefficient α controls the modulation level, which ad-
In this work, the aim is to synthesize an artificial propeller justs the variation in amplitude about the mean level.
signal. In order to generate the artificial signal, the propeller The determination of the harmonic coefficients Ak is rela-
noise spectrum obtained in the previous section is used as input. tively more complex. The ship propeller noise typically exhib-
The propeller noise spectrum may be represented by the its a strong harmonic at the blade frequency fbr defined as
nP Z
N-point discrete frequency domain function X(k). The fre- f br = k Hz
quency index k is related to actual frequency by the relation
60 (16)
k = 1,2,3,...
⎧k ⋅ f s / N , 0 ≤ k ≤ N /2
f =⎨ (10) Hence the value for the coefficient AZ is typically greater
⎩( N − k ) ⋅ f s / N , N /2 < k < N
than the other coefficients.
X(0) corresponds to the DC (zero frequency) component of
the spectrum and X(N/2) corresponds to the Nyquist frequency V. NUMERICAL EXAMPLE
fs/2. The remaining components are complex conjugates of one
The developed model is applied to a four bladed propeller
another such that X(N-k) = X*(k).
whose principal dimensions are provided in Table 1. The off-
To simulate a natural time domain signal, a random phase is
sets of the blade sections and the hub as well as the details of
associated with the spectral value X(k) which results in the
the trailing and leading edges of the blades are obtained from
randomized spectral value Xr(k). The random phase is gener-
[15].
ated using a uniform random variable U on the interval [0,1).
⎧ X (k ) k = 0, k = N
2
Table 1. Main Particulars of the Propeller.
⎪ Number of Blades 4
X r (k ) = ⎨ X (k ) e 2π i U 0<k < N
2 (11)
⎪ * Propeller Diameter (m) 2.100
⎩X r (N − k) <k<N
N
2 Pitch Ratio at 0.7R 0.8464
The time domain signal of the broadband noise is attained by Expanded Blade Area Ratio 0.55
Boss Ratio 0.276
taking the discrete Fourier transform of the randomized fre- Rake 0
quency spectrum. In other words, Skewback (degrees) 40
x(n) = DFT { X r (k ) } Pa (12) Direction of Rotation Right Handed
The modulator function m(t) is a sum of sinusoids and har-
The 3D representation of the propeller is provided in Figure
monics in the form of a Fourier series,
∞
1. The non-uniform wake field in which the propeller operates
m(t ) = ∑ An cos( n 2π f shaft t ) is shown in the form of a velocity ratio contour plot in Figure 2.
n=0 (13)
f shaft = nP / 60
The coefficients of the sinusoidal modulating function and
NET PROPELLER NOISE (MODEL SCALE)
v=3.35 m/s n=16.31 RPS
0.15
150
140
130
0.05
120
110
0 100
Z
Exp
90
Calc
-0.05
80
70
-0.1
60
50
1 10 100 1000 10000 100000 1000000
-0.15
-0.15
-0.1
-0.05
0
Y
X
0
0.05
0.15
Fig. 1. 3D Representation of Propeller. Fig. 3. Model Propeller Broadband Noise Spectrum for Condition 1 in
Table 2.
1.5
3 5 0 .0
1 0 .0
0
20.
.0
.0
30
0 .0
150
32
5
.0
1 0.7
40
31
0.
0.75
0. .0
0
50 140
0.8 0
30 0
0 .0
0.7
0 .0
60
Ls - SPL (dB ; re 1μPa, 1 Hz, 1 m)
0.5 2 90 .80
130
0
70 .
0
0
0.8
.0
0.9
0
0.85
120
5
0.95
0.7
2 8 0.0 8 0 .0
0.
90
r/R-y
110
0.9
0 270.0 90.0
1 0 0.0 100
2 6 0 .0 Exp
.0 1 10 90
-0.5 25 0 .0 Calc
12
0 .0 0 .0 80
24 0 13
0. 0. 70
23 0
0
14
0.95
0.
-1
0 .0
0.
15
22
60
.0
0
160
1 9 0 .0
0 .0
1 7 0 .0
21
180.0
200
50
.0
for the test conditions in Table 2 are presented in Figures 3 and 120
5 for the model and in Figures 4 and 6 for the full-scale pro- 110
peller. In these figures, the logarithmically scaled x-axis 100
Exp
represents the center frequencies (f) in Hz while the linearly 90
Calc
scaled y-axis represents the sound pressure levels (Ls) in dB re 80
120
110
100
Exp
90
Calc
80
70
60
50
1 10 100 1000 10000 100000
f - Center Frequency (Hz)
Fig. 7. Broadband Noise (Time Domain Version of Spectrum).
Fig. 6. Full Scale Propeller Noise Level for Condition 2 in Table 2.
The time domain version x(n) of the spectrum from Figure 4 It is concluded that the empirical prediction of broadband
(Condition 1) is attained using Equation (12), and the result is noise followed by a modulation technique as outlined in this
shown in Figure 7. The graph of x(n) is limited to a one second work may be used to generate a realistic time series audio signal.
period for the purpose of clarity. The full audio signal in WAV or MP3 format may be down-
loaded from the web site [16].
VI. CONCLUSIONS REFERENCES
An empirical prediction model of broadband noise for ma- 1. Atlar, M., Takinacı, A.C., Korkut, E., Sasaki, N., and Aono, T., "Cavita-
tion Tunnel Tests for Propeller Noise of a FRV and Comparisons with
rine propellers is developed. The model is composed of two Full-Scale Measurements," presented at CAV 2001: Fourth International
components: firstly, the empirical prediction of the frequency Symposium on Cavitation, Pasadena, USA, June 20-23 (2001).
domain broadband noise, and secondly, modulation of the noise 2. Blake, W.K. “Propeller Cavitation Noise: The Problems of Scaling and
Prediction,” ASME Symposium on cavitation and Multiphase Flow Noise
in the time domain.
(1986).
The results of the empirical prediction model are seen to be 3. Brown, N.A., "Cavitation Noise Problems and Solutions", International
in general agreement with the available experimental data. In Symposium on Shipboard Acoustics (1976).
order to impart a realistic audio character to the spectrum thus 4. Carlton, J.S., Marine Propellers and Propulsion, 2nd ed., Butterworth
Heinemann (2007).
obtained, a modulation model is also employed. As a result the 5. Hazelwood, R.A. and Conelly J., “Estimation of Underwater Noise – A
data obtained in the frequency domain is converted into an Simplified Method,” Int. Journal of the Society for Underwater Tech-
audible output. nology, vol. 26, no. 3, pp. 97-103, Nov. (2005).
6. ITTC, "Cavitation Committee Report", 18th International Towing Tank
For the future work, it is desired to enrich the empirical pre- Conference, Kobe, Japan (1987).
diction algorithm and the modulation parameters further with 7. Kehr, Y.Z. and Kao, J.H., “Numerical Prediction of the Blade Rate Noise
more empirically gathered data. Induced by Marine Propellers,” Journal of Ship Research, vol. 48, no. 1,
pp. 1-14, March (2004).
8. Kummert, A., “Fuzzy Technology Implemented in Sonar Systems,” IEEE
NOMENCLATURE J. Oceanic Eng., vol. 18, no. 4, pp. 483-490, Oct. (1993).
9. Lourens, J.G. and du Preez, J.A., “Passive Sonar ML Estimator for Ship
A, B Constants (Defined in Eq. (6))
Propeller Speed,” IEEE J. Oceanic Eng., vol. 23, no. 4, pp. 448-453, Oct
A0-n Constants in Eq.(15) (1998).
Ac Mean Cavitation Area on Propeller Blades (m2) 10. Nielsen, R.O., “Cramer-Rao Lower Bounds for Sonar Broad-Band
AD Propeller Disk Area (m2) Modulation Parameters,” IEEE J. Oceanic Eng., vol. 24, no. 3, pp.
285-290. July (1999).
D Diameter of Propeller (m) 11. Odabaşı, A.Y., Cavitation Inception and Prediction of Broad-Band Noise
KTip Constant (Either 60 or 80) Levels, British Maritime Technology, Tech. Rep. W1607, March (1987).
Ls Sound Pressure Level (dB) 12. Okamura, N. and Asano, T., “Prediction of Propeller Cavitation Noise and
Its Comparison with Full-Scale Measurements,” Journal of the Society of
LS P Value of Sound Pressure Level for Peak Freq. (dB) Naval Architects of Japan, vol. 164 (1988).
13. Ross, D., Mechanics of Underwater Noise, Peninsula Publishing, pp.
N Number of Points in Fourier Transform (Eq. (10)) 253-285 (1987).
Ps Static Pressure (Psi in Eq.(2), Pa in Eq.(3)) 14. Szantyr, J.A., “A Method for Analysis of Cavitating Marine Propellers in
Pv Vapor Pressure (Pa) Non-uniform Flow,” International Shipbuilding Progress, vol. 41, no. 427,
pp. 223–242 (1994).
VTip Tip Speed of Propeller Blade (m/s) 15. Takinacı, A.C., Korkut, E., Atlar, M., Glover, E.J., Paterson, I., Cavitation
i
VTip Rotation Rate of Start of Tip Vortex (RPM) Observation and Noise Measurements with Model Propeller of a Fish-
eries Research Vessel, Dept. of Marine Technology, Univ. of Newcastle
X(k) Spectral Value of Broadband Noise at Freq. Index k upon Tyne, UK, Report MT-2000-56 (2000).
Xr(k) Spectral Value with random phase 16. Takinacı, A.C., Propeller Noise Simulation Web Page. [Online] Available:
http://www.gidb.itu.edu.tr/staff/takinaci/NoiseSimulation/PropNoiseSim
Z Number of Propeller Blades ulation.html, (2008).
f Noise Frequency (Hz) 17. Yoshimura, Y. and Koyanagi, Y., “Design of a Small Fisheries Research
fbr Propeller Blade Rate Frequency (Hz) Vessel with Low Level of Underwater-Radiated Noise,” Journal of the
Marine Acoustics Society of Japan, vol. 31, no. 3, pp. 137-145 (2004).
fp Center Frequency (Hz)
fs Sample Frequency (Hz)
ACKNOWLEDGEMENTS
fshaft Shaft Frequency (Hz) (Eq.(13))
m(t) Modulator Function (Eq. (13)) This work was sponsored by TUBITAK BILGEM (Center of
np Rotation Rate of Propeller (RPM) Research for Advanced Technologies of Informatics and In-
k Frequency Index (Defined in Eq. (10)) formation Security). The authors also wish to thank the Em-
r Propeller Radius (m) erson Cavitation Tunnel of the University of Newcastle upon
x(n) Broadband noise time domain function (Eq.(12)) Tyne for the experimental data.
y(n) Final Output Signal (Pa) (Eq.(14)
P,M Indices for full scale propeller and its model
ΔL(P) Frequency Shift in Eq. (7) (dB)
λ Model Scale Ratio (Diameter of model propeller is
always taken as 30 cm)
ρ Density of Fluid (kg/m3)
σn Cavitation Number
σ ni Incipient Cavitation Number