A
image
.
"
>
.
.
•
.
.
ORNL P
1986
-
.
45
$6
FEEFEREE
2
|.25 1.1.4
1
.
.
.
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS -- 1963
SIS
cs
1566
CFSTI PRICES
ORNU p 1980
CONf- 660206-8
MAR %
H.C. $ 1.00, MN 50
THE ACOUSTIC CHARACTERISTICS OF THE OAK RIDGE REACTOR
"MASTER

R. F. Saxe
Presented at
and intenzione
La .
_--..*.'
Symposium on Neutron Noise, Waves and Pulse Propagation
. February 141-16, 1966
Gainesville, Florida
4+
***.. twisdom
The detection of the onset of nucleate boiling in a nuclear reactor is
hindered by: (1) the nuclear environment in which thc boiling takes place,
(2) the uncertainty regarding the probable location of the boiling, (3) the
location of the reactor core, vhere the boiling occurs, vithin a pressure ves-
sel and a biological shield. As a result of these restrictions, it was sug-
cestead, and that the observation of the fluctuations in the neutron population
in the reactor might lead to the detection of boiling. There are only a few
instances in the literature of the detection of boiling by neutron-flux noise
measurements, and in these cases the amount of boiling appears to have been
considerable. The dravbacks of this method of boiling detection are: (1) The
effect of a bubble on the ncutron flux in a nuclear reactor varies depending
on the position of the bubble in the reactor and in some positions a bubble
may have no effect at all. (2) £. neutron detector is receiving neutrons from
a large part of the reactor core and the effect of a single bubble is effectively
"smeared-out" over the whole, or a large part of the core. (3) The effects
of bubbles are coupled to the detector through the reactor core which acts as
fram
* Research sponsored by the U.S. Atomic Enercy Commission under contract with the
Union Carbide Cooperation.
WRC
,
LEGAL NOTICE
.,
RELEASED FOR ANNOUNCEMENT
fi
IN NUCLEAR SCIENCE ABSTRACTS
This report mi prepared a un account of Government sponsored work. Neither the United
states, aor the Commission, nor any person acting on behalf of the Commission:
A, Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, comploteness, or usofulness of the information contained in this report, or that the wo
of any information, apparatu, method, or procos, disclosed in to report may not infringo
privately owned righto; or
B. Asmumos may liabilities with rospect to the une of, or for damagos resulting from the
use of any information, apparatuu, motbod, or process diaolond la to report.
A, uued in the abova, "person acting on behalf of the Commission" includes any on-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Commission, or employs of much coatractor preparos,
disseminates, or provides accous to, any information pursuant to his employment or contract
with the Commission, or his employment with such contractor.

.
a "101-pass filter" and will therefore probably attenuate the bubble effects.
Another suggested method for the detection of boiling in a nuclear reactor
is that of measuring the scoustical radiations emitted by bubbles and at least
one successful measurement exists in the literature in this method, the ef-
fect of a bubble should be approximately independent of its position in the
reactor and the attenuating effects of the reactor core may vell ve suall.
In both these methods, however, the measurement must be conducted in the
presence of a background noise, and the value of the method vill depend on whether
the desired effect can be detected in the presence of this background noise or
not. Accordingly, in order to assess the promise of the acoustical method for
detecting the onsei of boiling in say ORR, the background noise in that reactor
was investigated.
Experimental
To detect and measure the spectrum of the acoustical radiations from ORR,
a microphone, (crystal contact microphone), vas attached to a metal pipe which
rises vertically from the top of the containrent tank and ends in a flat top
a short distance above the pool water surface. The attachment vas made by
pressing the microphone onto the flat surface by compressing a piece of expanded
plastic onto the back of the microphone rrith adhesive tape. This metal pipe,
which is braced to the tanks wall in two places, as used as an elementary a-
coustic wave guide to eliminate the need for placing the microphone under water,
The signal from the microphone vas amplified, and analyzed by a Hewlett-
Packard Model 300A vave analyzer.
To plot the frequency spectrum of the acoustic noise, the meter on the
front of the wave analyzer vas read for each frequency setting. At most fre-
quency settings the deflection of the meter needle fluctuated in an apparently
mer.-
num
:
.
.
- 3
maridom manner, and the mean value was assessed by eye. Repeated readings on
different days reproduced the same frequency spectrwa curve quite well, but
the accuracy of individual readings is not high.
The modulation of any desired frequency banci in the acoustical noise was
displayed by passing the output of the vave analyzer through a semi-conductor
diode detector. The wave analyzer operates by heterodyning the signal to be
measured with a local oscillator signal adjusted to give a difference frequency
of 20 kc/s which is then amplified by an amplifier tuned to 20 kc/s. Any
modulation of the input signal, therefore, appears as a nodulation of the 20
kc/s output of the tuned am:lifier. This modulation was extracted by means
of a circuit consisting of a semi-conductor diode i series with the parallel
combination of a capacitor, capacitance 0.01 uF, and a resistor R. The value
of R was ad justed to provide signals of suitable amplitude for the input of
the tape recorder on which the modulations were recorded; its value was of
the order of 104 ohms giving a time-constant of the order of 10° seconds which
was found to be adequate to give good recording of the modulation without af-
fecting the modulation frequencies of interest. The recorded modulation ave-
the neutron flux amplitude recordings.
Simultaneous recordings of the acoustic signals on one channel and of the
neutron flux signals on an adjacent channel enebles, in principle, the computa-
tion of the cross-correlation function or of the cross-spectral density curve
to be performed.
Experimental Results
The frequency spectrum of the acoustic radiations from the reactor, trans-
mitted up the pipe to the microphone, were plotted for the reactor for full
*
*:7.5*
power (30 MW) and with full water flow rate and is shown in Fig. 1. It vill
be seen that it consists of several large peaks, particularly large ones being
at 1.4 - 1.5 kc/s and at 2.2 kc/s.
The frequency spectrum of boiling measured in vater at atmospheric pres-
sure has been shown to consist also of peaks usually in the frequency regions
of 0.9 - 1.2 kc/s, 2.2 kc/s and at higher frequencies. The frequencies at which
bubbles radiate are pressure dependent and the frequencies to be expected
from ORR would, therefore, be a little higher thɛn those quoted above.
It vill be realized from this that the chances of the acoustical detection
of boiling in ORR are considerably reduced by the presence of this background
noise having peaks at or around the frequencies at which boiling bubbles would
..
.-....
..
.
be expected to radiate.
If, therefore, acoustical measurements are to be con-
sidered as a possible boiling detection method, it is desirable to attempt to
understand the cause of this background noise with a view to its reduction or
elimination. Further, it would be of interest to know if this background noise
is peculiar to ORR or is likely to be found in other reactors.
The frequency spectra of the acousvic radiations vere plotted for different
values of reactor power and for different values of water-flow rate. It was
found that the spectrum of the radiations is independent of reactor power level
but 18 very dependent on the value of water-flow rate as is shown in Fig. 2.
It will be seen that the higher frequencies increase very rapidly with increase
of water-flow rate, while the lower frequencies increase relatively little. Re-
cordings of the modulations of the 1.4 - 1.5 kc/s peak and of the 2.2 kc/8 peak
were made on the tape deck at a recording speed of 3.75 inches per second and
using a bandwidth setting on the wave analyzer of 30 c/s. The frequency setting
of the wave analyzer was adjusted in each case to give maximum amplitude of the
:
.
.
.
.
'
S
"
-5-
modulation. Later, recordings vere made of the modulations of the 3..4 - 1.5
kc,'s peak and of the 2.2 kc/s peak, at a bandwidth setting of 145 c/s, and,
simultaneously, on an adjacent channel, of the neutron flux variations. The
modulations of the two above mentioned acoustic frequencies were also recorded
on adjacent channels.
Fig. 3 shows the spectral density curves for the neutron flux variations
erid for the modulations of the acoustical frequencies at the two values of
bandvidth. It will be seen that the neutron flux spectral density curve falls
steadily from low frequencies to high frequencies with a peak centered at about
11 c/s. The 1.5 kc/8, 145 c/s, modulation shows an approximately constant
low frequency response, a fall-off at higher freq:encies and a peak at about
14 c/s. The 2.2 kc/s, 145 cis, modulation shows approximately the same behavior
with a broad peak at the range 20-17 c/s.
The 1.5 kc/s, 30 c/s, modulation shows a more pronounced low frequency
response with a small peak at 7 c/s while the 2.2 kc/s, 30 c/s, modulation
shows an approxima tely flat low frequency response with a rapid fall-off at
frequencies above about 7 c/s.
Discussion
The frequency spectrum of the noise detected by the microphone in ORR
gives a high frequency recion which has prominent peaks, the amplitudes of which
increase rapidly with increase of water-flow rate. In order to generate the
three values of water-flow rate used in the experiments, one, two, and three
water purps were used. In general it would be expecteå that, if the acoustical
noise was being generated by the pumps, then the measured noise for the three
cases would be in the ratio 1:2:3. The observed rapid increase of measured
noise vith the number of purps renders it unlikely that the noise is being

generated at the pumps. It also seems unlikely that the noise is being generated
in the large smooth-bore vater pipes connecting the pumps to the reactor. The
reactor itself consists of an assembly of narrow channels into which the flow
must be constricted, and it appears likely that the acoustic noise generator is
vithin the reactor core.
D. H. Jorgensen has made a study of the acoustical noise from cavitating
water jets. In these mea surements, the first evidence of cavitation as a rather
sudden increase of noise level in the high frequency bands, the cavitation
being heard as infrequent, randomly intermittent bursts of noise. For condi-
tions where cavitation is vell advanced, the noise spectra depend on the physical
properties which determine the flow.
Jorgensen uses the cavitation index
1/2 p U2
where P. is the static pressure at the nozzle, o is the density of the liquid,
and V is the liquid velocity in the nozzle, as a parameter indicating the amount
of cavitation likely to be present. Using cylindrical nozzles, the smallest
being 3/8 inch in diameter, he shows that cavitation noise is present for values
of o less than about 0.6 and is well developed for values of o less than about
0.6 and is vell developed for values of around 0.2 or less.
Concentrating attention on the coolant channels between the fuel: plates of
ORR, recognizing tiiut the nozzle is of rectangular shape with large aspect
ratio and not a cylinder, assuming, nevertheless, that the small dimension
is the dominant one, we calculate that for ORR at full flow (20,000 gal/min)
020.08, while for flow from one pump (10,000 gal/min) 00.3. It follows,
therefore, that for one pump flow there may be little cavitation and that for
fwl flow there may be vell-developed cavitation.
s
.
.
-7-
If cavitation 18 present, it will represent a voidage in the core region
possibly giving a reactivity change compared with the condition of no voidage.
If the aivount of voidage is approximately constant for any given conditions,
the reactivity effect vill be constant and will be offset by the control rods
to maintain criticality. However, if the amount of and possibly position of
the voidege is varying, it will cause the reactivity of the reactor to vary
and could give rise to ncutron flux noise.
There is also the possibility that the presence of cavitation could cause
a flow change and could give rise to neutron flux noise by coupling through
the temperature coefficient.
The spectral density measurements on the modulations of the two prominent
acoustical peaks show increased response within the range 7-14 c/s as does
the neutron flux spectral density curve. This suggests that either:
or
1. The cavitation (assumed to be present) modulates the neutron flux
and is itself modulated, for example, by hydrodynamic instability,
2. the cavitation modulates the neutron flux and is itself modulated
by, for example, variations of flow rate,
3. the same phenomenon modulates both the neutron flux and the acoustical
or
radiations and that there is no coupling between the acoustical genera-
tor and the neutron flux.
Several possibilities and conclusions spring to mind as a result of this study:
2. That the fluctuations in the neutron flux may be caused, partly at
doen
least, by hydrodynamic phenomena.
2. That the acoustic radiations generated in the reactor are very depend-
ent on the water-flow rate.
3. The nature of the acoustic radiations under different conditions are
.....
3-
consistent with the hypotesis of their generation by flow induced
hearino
cavitation
time
4. Such cavitation, if present, may affect the reactivity of the reactor
state
with
and may give rise, at least in part, to the neutron flux noise. The
reutron flux fluctuation spectral density behavior with variation of
way to master the
reactor power and with water-flor: rate is consistent vith this hypothe-
sis.
This is not, of course, evidence or causality alone but indicates
either causality or common oricin.
5. Assuming only common origin, the acoustical radiation measurements
.
--
-
provide another means of studying the mechanism giving rise to the
worn.
observed effects.
des termee
minaimbit
6. The behavior of the acoustical radiations when the reactor is approach-
ing boiling may be useful as an incipient boiling detector in that
a
one vould expect that flow induced cavitation vould be more readily
s t
ermatitis met sterker
started at higher temperatures.
transmit
7.
ent
in the internet ortam
ehdottoma
in naisten kommenter on line
In any boilirg experiment in which throttling of a channel is cmployed
to induce boiling in that channel, there is the possibility that the
acoustical radiations from the reactor vill be altered and that cavi-
tation may be caused in the channel by the throttling and may affect
the boiling in that channel.
In any event, the background noise against which boiling acoustical
8.
noise must be detected is very water-flow dependent and may be modified
if the water-flow conditions are altered. At present the background
noise in ORR is such as to make the acoustical detection of boiling
doubtful.
.
AR
2. 4
-16-11-15
Wir
Ackno:ledgements
The work described herein was performed while the author was a summer par-
ticipant at the Oak Ridge National Laboratory, and the autizor wishes to thank
the Union Carbide Company for permission to publish the se resuits. The author
also vishes to acknowledge the generous assistance given by members of the
..-
-
Controls and Instrumentation Division and in particular Messrs. E. P. Epler,
-.-
.
-
D. P. Roux and D. N. Fry.
References
1. KAPL-M-IRB-3
2. Boyd, L. R., Nucleonics 11, 96, 1959
3. Thie, J.A., Reactor Noise, Roiman & Littlefield Inc., New York, 1963
4. James, L.C., Nuclear Engineering, p. 18, January 1965
5. Strasberg, M., J. Acoust. Soc. Amer., 28, 20, 1956
6. Jorgensen, D.17., J. Acoust. Soc. Amer., 33, 1334, 1961
.
...
...
...
po?.
.
s
,uwii
.V
i
2
lot
400 T
10
+
FeG!
AMPLITUDE.
Oo!
FREQUENCY (2/5).
1000
10 000
L
A
RA
V
i
ctoria en
el
mediante che
non ci th
com Katowania
in the most the money and other towar writia.
---
-


400
..AMPLITUDE.
- 20 000 galls/min.
- 16800
to 0000
2 то
loo
1000
10 000
FIGZ
FREQUENCY (c/s).
.
*
-.-
.
.
.
-
.
.
.
A
..1
*
.
- ..
.in
....
......
·
.
..
.
.
:
!
*
.
*
1.5kes (3014
NEUTRON FLUX.
.
.
2:2 kels (30c/s).
2
tot

15 kels (1454)
:-.2.2 ke/s (145c/5).
10
FREQUENCY (4)
مست
.. انهم من شده نفت نیمه : نه مخبودند که منشمننمللمفشخ باسنمتنتمكنه
.
.
.
-.....
.
..
.
..
.
END

DATE FILMED
6 | 16 /166
-
-
-
-
-