Summary of work on acoustic properties of underwater bubble screens

NPS-61-82-003-PR

"' IOOL

J.

LIBRARY

RESEARCH REPORTS DIVISION
NAVAL POSTGRADUATE SCH:
MONTEREY, CALIFORNIA 93940

//

Wm. POSTGRADUATE SCHOOL

Monterey, California

a

•

.

SUMMARY OF WORK ON ACOUSTIC' PROPERTIES OF
UNDERWATER BUBBLE SCREENS

O.B. WILSON and J.V. SANDERS -
Department of Physics and Chemistry

FINAL REPORT FOR PERIOD JAN 19 81 - DEC 198 2

Prepared for :

Commander, Puget Sound Naval Shipyard

ATTN: Mr. John Kriebel , Acoustic Range Branch, Code 246

Bremerton, WA 9 8 314

FEDDOCS

D 208.14/2:NPS-61-82-003PR

NAVAL POSTGRADUATE SCHOOL
Monterey, California

Rear Admiral J. J. Ekelund D. A. Schrady

Superintendent Acting Provost

This report was prepared as a summary of work supported in part by
funds fron the Puget Sound Naval Shipyard, Job Number N 0025181 WR
10124, dated 30 January 1981.

Reproduction of all or part of this report is authorized.
This report was prepared by:

A. INTRODUCTION

The objective of this report is to summarize the work carried
out this last year by Lieutenants Kelley and Marr as part of their
thesis research, supported in part by funds from your
organization, and to provide some comments, conclusions and
recommendations of our own. The theses (1,2) have been
transmitted separately.

In the design of an acoustically insulating bubble screen,
there are many aspects which must be considered. The work of
Kelley and Marr addressed primarily only two, the acoustic
transmission properties of the screen and the noise which is
generated by the screen itself.

B. NOISE GENERATION

A possible problem in the use of a bubble screen for noise
isolation is the noise generated by the screen. The work of LT
Kelley (1) was directed toward the measurement of the noise power
associated with the formation of a bubble screen. The following
summarizes some of the results in Kelley' s thesis and recasts some
of the results in a form which may be more useful. Some errors
noted in Ref. 1 are also corrected.

The work was carried out in the tanks in the Postgraduate
School's Underwater Acoustics Laboratory, using a reverberant
field method. The steady state spatial average of the sound level
in an enclosure which has at least partially reflective walls is a
measure of the acoustic power generated by a source in the
enclosure. An enclosure, in this case, the tanks, and the
associated hydrophone system were calibrated by using a source

of known power output. A similar averaging taken when an unknown
source is in the enclosure can be used with the calibrations to
estimate the power output from the unknown source. In this way LT
Kelley was able to make some measurements of the acoustic power
created or associated with the bubble screen for several
configurations of bubble generating manifolds and air flow rates.
Efforts were made to approximate bubble densities which might be
appropriate for a screen which would be effective at the low
acoustic frequencies of interest at Carr Inlet. Noise
measurements were made on three different bubble generators,
constructed by drilling rows of small holes in two inch diameter
PVC pipe. They differed in number and sizes of holes. One had
many small holes, the second had about the same hole area achieved
by a smaller number of larger holes. The third had one row of
small holes. The generator pipes were a bit less than five feet
long and were located at the bottom of the tanks which are about
seven feet deep.

Flow rates were controlled by valves and were measured using
flow meters and by timed capture of air of a known volume from the
bubbles. Only approximate estimates of bubble size were possible
using visual and photographic observation. Bubble density was
estimated from flow rates, bubble rise times and screen
dimensions. Tests were conducted to discriminate between the
relative noise generating characteristic of bubble formation,
bubble rise and bubble venting. There were found some errors and
omissions in Kelley' s thesis. Some are trivial and obvious, but

the errors in Table VI are not. Enclosed as an appendix is a
corrected Table VI with additional information tabulated.
Kelley's results support the following conclusions:

(1) The dominant source of noise from the screens produced
in this lab is the bubble formation. Bubble migration appears to
be a measurable source and may contribute as much as ten percent
of the energy. Bubble venting appears to contribute very little
additional noise in the frequency range we used (20Hz to 10kHz).

(2) The production of bubbles by a large number of small
holes is significantly quieter (the order of 10 to 15 dB at some
frequencies) than when the same air flow rate passes through a
smaller number of larger holes. This effect may be due, in part,
to better acoustic shielding provided by the bubble distributions
around the pipe in the case of the larger number of holes.
Quantitative measures of the differences are imprecise because at
many frequencies, the noise generated by the quiet screen was less
than the ambient threshold of the measuring system.

(3) Figures 15 and 16 in Ref. 1, give the source level in one-
third octave bands for the 4.8 foot long screen in dB ref 1 UPa at
1 meter. It can be seen that for the quieter type bubble generat-
or, the source level determination is limited by the ambient noise
threshold. We believe that the peaks in the output from the
noisier manifold may be due to bubble resonances. However, we can
not be really sure that these results would be applicable to
bubble generators at significantly greater depths.

A worst case is from Figure 16 at 200Hz. The one-third
octave source level aiven is about 132 dB . If it is assumed that

the acoustic power is a linear function of the length of the
manifold and that the spectrum is uniform over a one-third octave
band, then for one yard length of screen an acoustic power
generated in a one Hertz band at 200Hz is about two microwatts.
This is for a bubble screen about ten inches thick, one yard long
with an air bubble concentration of about one percent and a total
air flow rate of about 4 SCFM . The quieter screen should be less
noisy than this by a factor of ten or more in sound power.

If one desired to predict the level at some distance from
such a screen, a reasonable source model is to assume that the
screen generator behaves as a line array of incoherent sources.
Transmission loss models appropriate to the geometry of the
problem would have to be chosen and applied.
C. ACOUSTIC TRANSMISSION

LT Marr (2) assumed a geometrically ideal screen (plane,
parallel boundaries) with a uniform distribution of non-resonant
bubbles having diameters much less that the wavelength of the
sound. His computer calculations showed that for realistically
obtainable bubble concentrations (between 10~1 and 10~3
volume percentage of air), stop bands exist within which the
transmitted intensity is significantly reduced for a broad band of
frequencies and a considerable range of incident angles. The
attenuation in these stop bands exceeds 20 dB . Pass bands of
nearly 0 dB attenuation have very small frequency extent.

A real bubble screen is not expected to be geometrically
ideal nor is it expected that the bubble concentration will be
uniform. The screen will increase in width as it rises to the

surface, and its boundaries will be irregular and ill defined.
Furthermore, the bubble concentration will vary with depth
(because of both the increased size of the individual bubbles and
the increased width of the screen) and with distance from the axis
of the screen. The first effect is difficult to predict because,
while the bubble size as a function of depth is well known, the
width of the screen has never been measured except for screens a
few meters deep. The behavior of bubble concentration with
distance from the axis is completely unknown; it has been reported
in the literature that both the concentration and the bubble size
decrease near the edges of the screen.

It is difficult to predict the effectiveness of a non-ideal
screen. While the transmission in the pass bands will undoubtedly
be reduced, it is equally likely that the transmission in the stop
bands will be increased, making quantitative prediction of the
effectiveness of a real screen impossible. Given the distribution
of bubble concentration, sophisticated theoretical approaches
exist that could be used to predict the transmission. However, it
is our belief that the hydrodynamic theory necessary to predict
the bubble concentration for a given bubble-injection population
does not exist.
D. RECOMMENDATIONS

We believe that the work we have studied so far does not
provide enough information to permit design and construction of a
full scale bubble screen with a satisfactory degree of risk. This
applies to both the noise generation and the attenuation. There
is no question that a screen of the kinds considered by Marr would

provide a sizezable amount of attenuation. It is not really clear
that the bubble generation noise would be a problem. We just do
not yet have sufficient confidence to design a full scale screen.

It is our opinion that in the absence of a full-scale basic
research effort the effectiveness of a real bubble screen would be
most efficiently determined from in situ experiments. Such exper-
iments must be carried out on a larger scale that those done prev-
iously where the screen depth never exceeded more than a few
meters. Measurements would have to be made of the bubble concen-
tration as a function of depth and distance from the axis of the
screen. Acoustical transmission measurements should be made con-
currently to allow comparison with theoretical results. Some very
recent bubble screen experiments done in a pond by S.N. Domenico
at Amaco Production Co . ( 3 ) came to our attention this week.
We have talked with the author and expect to get a preprint of his
paper soon. We will send a copy of it as soon as we get it.

We recommend that full-scale bubble screen experiments not be
carried out until after more research has been done to better
define the basic hydrodynamic properties of such screens. Since
such work would be both expensive and slow, we suggest that other
approaches to reducing the effects of noise interference in Carr
Inlet be examined. Signal processing combined with directional
hydrophone arrays might be a productive approach.

References

1. Experimental Study of Noise Produced by an Underwater Acoustic
Bubble Screen, Clark Thomas Kelley, MS Thesis, NPS, June
1981.

2. On the Design of an Acoustically Isolating Bubble Screen For
Carr Inlet Acoustic Range, Kenneth William Marr, MS Thesis,
NPS, June 1981.

3. Acoustic Wave Propagation in Air Bubble Screens in Water, S.N.
Domenico, Paper presented at the October 1981 meeting of The
Society of Exploration Geophysics. (To be published in
Geophysics, 1982.)

cu

- — .

c s

M

O 3

D

-fl

•H i-i

P

CU

P O

(0

.»->»

P

id >

5

-p

r-

co

r-

<d

M

c

"*

ON

r^

m

CO

CN

^

rH

CN

E

P >. 0

cu

•H

c A

p

u

H

rH

rH

o

o

iH

o

rH

rH

+J

cu

U

0}

O 0

u

cu

a

Con
(Rati

■H

IW

0

cu

T3

CU

5

05

0

c

•

•H

(— 1

-H

H

VO

VO

m

vo

r^

<*

o

CN

<D

>

Cn

05

s

E

CU

0)

CN

«*

un

CN

m

CO

o

rH

rH

•H

OS

05

3

i-t

■H
<

+J

id
OS

p

Cm

P

rH

rO

T3

•H

rH

+J

cu

T3

^*

vO

<*

•H

id

p

id

LO

•

•

un

•

Un-

Q

id

e

OS —
E

CN

ro

o
1

^*

in

o
1

CN

en

o
1

ci)

en

•H

cu 0

o

O

CN

o

O

co

O

O

CN

0

O

P

rH ^-^

•

•

•

3

■H

CO

A

O

O

O

•a

P

w

A

cu

05

3

>-4

•H

CQ

M

0

cu

p

p

05

0

05

cu

03

c

0)

05

u

cu

c ~

•H

id

cu

^ e

O

in

o

r^

m

r^

r»

r»

co

>

fi

In-

U 0

H

r-4

CN

rH

CN

CN

^C

0

•H w

p

G

CO

£

en

c

0)

cu

a)

rH

M

X

cu

0

0

E

T3

CO

CQ

•H *-*

*

CU

E* C

r^

r^

-*

CM

«tf

<T>

«tf

CO

m

C

cu

CO

•H

M

1— 1

CO

* e

rn

rH

i— I

ro

H

O

<r>

r^

vO

3

£

Id

rH w

P

A

rH

•H

fl

3

o

En

•

na

•H

X

aa

P

D

s

0

-i

CU

"*>N»

CN

—1

«*

CN

o

O

t>

rH

rH

rO

*-•»

CU

d,p

CT»

o

O

^0

k>

r»

r*»

CO

CO

•

cu

P

>

CO

Cm

i-H

r— t

1— 1

r— 1

r-i

05

H

s

C

05

>

■H

cu ~

id

E-1

rH U

^

o

<T>

m

It

<*

m

o

O

r-H

w

3 1)

o

J

cu ^

I *

<*

ro

CN

CN

CN

m

m

in

CQ

05

Sm — •

i

■K

<

■H

3

'

=H

OS

Eh

-

TJ

.5

p

01

p

1 CD

VO

i— I

o

«*

o

CO

rH

^

M

0)

1 ^

vO

r-»

c?>

o

—1

in

VO

\Q

cu

cu

Cl. P

rH

rH

g

^^

>

CO

Cm

•H

—< — »

E-

r-l U

•H CU

O

^H

r^

\T)

G\

r»

O

*£

m

cu

P 05

0}

CO — '

^

U?

m

«*

ro

ro

r^

vO

vO

Cm

P •

05 0

CU 2

E-1

CN

rn

in

^o

co

cr>

APPENDIX

DISTRIBUTION LIST

Commander

Puget Sound Naval Shipyard

ATTN: Mr. John Kriebel

Code 246

Bremerton, WA 98314

Dudley Knox Library
Naval Postgraduate School
Monterey, CA 93940

Professor J.V. Sanders
Code 61Sd

Naval Postgraduate School
Monterey, CA 93940

Professor O.B. Wilson, Jr.
Code 61W1

Naval Postgraduate School
Monterey, CA 93940

LT Kenneth Marr
4717 Thresher Ct.
Virginia Beach, VA 23464

LT Clark Kelley
1812 Long Meadow Dr.
Montgomery, AL 36106

U200022

DUDLEY KNOX LIBRARY - RESEARCH REPORTS

ii ii 1 1 ii iiiiii illinium

!fl;ll

ii 1 1 ii iiiiii mil
5 6853 0

071516 2