DEVELOPMENT AND TESTING OF AN ACOUSTIC SYSTEM FOR IN SITU DETERMINATION OF MICROBUBBLE CONCENTRATION IN THE OCEAN Leonard Arnold Wiens United States Naval Postgraduate Schoo m Development and Testing of an Acoustic System for In Situ Determination of Microbubble Concentrations in the Ocean by Leonard Arnold Wiens Ensign, United States Mavy S., United States Naval Academy, 1970 Thesis Advisor: H. Medwin June 1971 AppA-Oued {^on. pub tic n.zlzcii>t; dOitnAbation u}itanltQ.d. T139818 Development and Testing of an Acoustic System for ^O. Situ Determination of Microbubble Concentrations in the Ocean by Leonard Arnold Uiens Ensign, United States Navy B. S., United States Naval Academy, 1970 Submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN OCEANOGRAPHY from the NAVAL POSTGRADUATE SCHOOL June 1971 /AL POSTGR-CDUATE SCHOOL STSREY, CALIF. 93940 ABSTRACT The effect of bubbles on sound propagation in the ocean has long been known. More recently, bubbles rising to the ocean surface have been related to salt nuclei in clouds. Yet due to the lack of suitable instrumentation little data are available on the concentration of bubbles in the ocean. A one-dimensional standing wave system was con- structed and evaluated to determine bubble concentrations by measuring the effect of bubbles on the system damping constant (Q" ). The system was comprised of a mylar electrostatic transducer built into a plane reflector facing a parallel reflecting plate, and associated electronic components. The system was designed to measure bubbles of radius from approxiniately 500 microns to "^O microns utilizing frequencies from 20 to 100 kHz at depths to 30 meters. The system accuracy was determined 3 to be + lOHz in resonant bandwidth corresponding to 500 bubbles/m at 6/ kHz. The system was used to successfully measure bubbles in a laboratory tank. TABLE OF CONTENTS I. INTRODUCTION 7 II. BUBBLE THEORY 8 III. ACOUSTIC THEORY 9 IV. THEORY OF ACOUSTIC rCASUREMENT 12 V. EQUIPMENT DESIGN 15 A. SIGNAL GENERATOR 15 B. POWER AMPLIFIER 15 C. POLARIZATION 16 D. TRANSDUCER, REFLECTORS AND FRAME 16 E. HYDROPHONE 18 F. PREAMPLIFIER — - 18 6. AMPLIFIER AND RECORDER 19 VI. EQUIPMENT EVALUATION - 20 VII. GENERAL CONCLUSIONS 25 A. SUMMARY 25 B. RECOMMENDATIONS 25 BIBLIOGRAPHY 40 INITIAL DISTRIBUTION LIST 41 FORM DD 1473 42 LIST OF TABLES I. System Electronics 30 II. Transducer Specifications 32 III. Bubble Concentrations Measured in Laboratory Tank 39 LIST OF FIGURES 1. Bubble Radius as a Function of Resonant Frequency and Depth for Air-Filled Bubbles 27 2. Damping Constant as a Function of Frequency 28 3. Diagram of Pressure Amplitudes in Two of the Standing Wave Modes 29 4. Block Diagram of Standing Wave System 31 5. Transducer Diagram 33 6. Diagram of Frame with Plates 34 7. Samples of System Response for Two Transducer Backs 35 8. Time versus Q Curves for Plexiglass Transducer. No Bubble Removal 36 9. Charging Curve of Transducer 37 10. Time versus Q Curves for Plexiglass Transducer. Bubbles Removed 38 ACKNOWLEDGEMENTS The author wishes to express his appreciation to Dr. Herman Medwin for his guidance and patience during the preparation of this thesis and to Mr. William Smith for his assistance in the construction and maintenance of equipment. I. INTRODUCTION The process of scattering and absorption of sound energy by air bubbles in water is well known. However, except for bubbles observed in ship wakes or in the surf zone, little is known about the presence and persistance of bubbles at sea. Although there has been some notable work done in the laboratory concerning bubble action in the water, there have been very few attempts to make "in situ" measurements of bubbles at sea. Until 1964, when work commenced at the Naval Postgraduate School to develop a system to measure bubbles at sea, there were no suit- able instruments to determine bubble concentrations at sea. The acoustic system used in this project was basically developed from projects starting in 1964 [2,3,4]. A knowledge of bubble concentrations in the ocean is important for reasons other than their effect on sound propagation. A bubble in water is a cavitation nucleus. As it rises to the surface, it collect parti- culate matter and chemicals in solution, and when it bursts at the surface, it produces airborne salt nuclei which can be linked to tnunder- storm activity [1]. The particulate matter and organic skin may form surface slicks, which in turn could.be evidence of the mechanism which produced the bubbles in the first place, such as plant and animal life or possibly internal waves and chemical processes in the sea. Therefore, a greater knowledge of ocean bubbles could lead to a greater understand- ing of other ocean processes. This project was undertaken to provide suitable instrumentation to gather "in situ" data on bubble concentrations at sea. ■ II. BUBBLE THEORY Classical theory suggests that bubbles have only a transitory exis- tence in water and last only long enough to rise to the surface or dif- fuse into the surrounding water. Hov/ever, experiments have shown that small bubbles may remain stabilized in the water for long periods of time. Experiments have indicated that bubbles of radii less than about 30 microns may remain suspended indefinitely, and bubbles have been observed to persist for over 100 hours [8]. Several bubble stabilization models have been proposed to explain the persistence of microbubbles , but none satisfy all conditions. How- ever, in the bubble sizes of interest in this research, (30-500 micron radius) the compressed wall model provides the most satisfactory '' explanation for bubble stabilization. In this model, when a bubble is formed in the water, it begins to capture solid particles. At the same time the bubble shrinks by diffusion of gas into the surrounding water. The bubble continues to collect particles and to shrink in size until a continuous wall of particles is formed around the bubble. This wall would be capable of supporting a compressive stress, thereby maintaining the bubble size. Gas would continue to diffuse out of the bubble until the bubble's internal pressure equaled that of the partial pressure of gases in the surrounding water. The mass of the particles forming the bubble's wall would supply counter buoyancy to the bubbles [8]. In this manner, bubbles could attain various degrees of stability in the water column . III. ACOUSTIC THEORY A gas bubble, because of its elasticity and effective mass, forms a resonant system which are able to scatter and absorb sound to a much greater degree than its geometrical size would indicate. The scattering cross section for a single bubble is [5]: power scattered 47Ta' Incident acoustic intensity The absorption cross section is given by: power absorbed by bubble 2 2 + 6 (1) (l 's ^k^a ■1) (2) Incident acoustic intensity and the extinction cross section is defined as the sum of (1) and (2) ^a ■" <^s (3) where: a = resonant bubble radius w = resonant angular frequency of bubble pulsation w = angular frequency of incident sound 6 = bubble damping constant 2tt = wave number at resonant frequency, The resonant frequency for a clean air bubble in water is given by [5] . = _L (3y£ t j2 ° 2Tra P ^ Rearranging terms and noting that f = c/x and k = Zi\/\ gives: 1 (4) ka = ( 3yP t\2 where: Y = ratio of specific heats of bubble gases p = ambient pressure at bubble depth p = density of surrounding water T = surface tension factor. The factor t is very nearly unity for bubble sizes of interest in this study, and the quantity y/^ is in the range 1.6 < y^"^ < 1.40 for bubbles greater than 20 microns. Using these values and representative values for the ambient pressure and water density, the value of ka is on the order of 10~ .■ Thus, using equations (1) and (2) and taking into account the value of ka, the expression for the extinction cross-section of a bubble is given by: °e ^/[(^)2 -if^ 6 (5) At resonance, equation (5) reduces to: Since: k - ^ - ? ^0 - 2ac ^e. "o^ -0 t.5 where c = velocity of sound in water, m/sec. A 150 micron bubble has a resonant frequency of approximately 25 kHz at the water surface (Figure 1). Using c = 1500 m/sec and 6 = .07 (from Figure 2), the resonant extinction cross-section for this bubble would be 2.58 cm^. This represents an acoustical cross-sectional area 3550 times the geometrical cross-sectional area. In comparison, a rigid scatterer of the same size would have an acoustic extinction 10 cross-sectional area at most equal to the geometrical cross-section and at the bubble's resonant frequency would have an acoustic cross-sectional o area 10" times its geometric cross-secional area [5]. Therefore it can be assumed that sound attenuation by solid particles can be ignored. Because of the large acoustic cross-section of bubbles, acoustic measuring techniques have to date proven to be the most practical. Three assumptions are necessary in order to determine bubble concentra- tions acoustically. First, it is assumed that particulate matter form- ing the bubble wall has a negligable effect on the extinction cross- section; second, the bubble gases are assumed to be the same as atmos- pheric air; and third, it is assumed that all sound attenuation at a given frequency is due to bubbles of resonant size [5]. 11 IV. THEORY OF ACOUSTIC MEASUREMENTS When an acoustic signal is transmitted between two rigid, parallel reflectors, a standing wave will be established if the spacing of the two reflectors is an integer multiple of a half wavelength for a given frequency (Figure 3). This can be expressed as: L = ^ n = 1, 2, 3, .... (8) where: L = spacing between plates. If the plates are fixed in position, the resonant frequencies for a given plate spacing is given by: nc ^n K " = ■•' 2. 3, ... (9) where: f = ^ The efficiency with which a resonant system stores energy is termed the "quality factor", Q. The Q may be expressed as [4]: Q = ^ (10) where: fo = resonant frequency in Hz Af = half power bandwidth in Hz, or by: Q = -^ (11) where: a = spatial attenuation, nepers/m X = wavelength at the resonant frequency. 12 When bubbles are present, the system v/ill not be as efficient in storing energy; therefore, the system Q will be lower. By comparing the system Q in bubble free water and in water with bubbles, the sound attenuation due to bubbles can be determined. If measurements in bubble free water are denoted by the subscript 1_ and measurements in water with bubbles are represented by the subscript 2^, the attenuation due to bubbles can be bound by rearranging equation (11): IT a - — QX The increase in attenuation due to a bubble is 'Q2 Ql Aa - ap - a-] - y (— - -^ ) . Recalling Equation (10): «„ = { h'^ r " Af -1 ^ npnorc/m au. ^_ v-i'v '-'']/ ..<-(--. -^/i.. ■Converting to db/m gives: Aa = M8l (Af2 - Af^) _ (12) Aa can be related to the extinction cross-section, a , as [6]: Aa = 4.34na db/m (13) where n = number of bubbles per cubic meter. Setting (12) equal to (13) gives: ^^ (Afo - AfJ = 4.34 na r. ^ \ e Solving for n gives the expression: n = '^ (Af. - AfJ 13 Substituting for a^ from (7) gives the expression for bubble concentrations: n = ^^y^ (Af2 - Af, ) bubbles/m • (14) c a 14 V. EQUIPMENT DESIGN The basic design of the standing wave system developed in this pro- ject was the result of earlier projects [2,3,4]. However, previous sys- tems either did not give consistant results or failed to give any results at all. In this project, the standing wave system was redesigned and tested to insure reliable performance in the frequency range of interest (20-100 kHz). A block diagram of the entire system is shown in Figure 4. Each block will now be considered separately. A. SIGNAL GENERATOR The first oscillator used to provide the input signal to the systeir. was a WAVETEK 114. The WAVETEK was used in the sweep mode with the sweep rate kept below 100 Hz/sec in order to allow Lhe resonant pattern to develop completely [4]. The time sweep of the WAVETEK is linear, provided it is properly warmed up. This oscillator is limited to a maximum sweep time of five minutes, so several runs were necessary to cover the frequency range of interest. To facilitate data taking, an HP3590A Have Analyzer was used in place of the WAVETEK. The main advantage of the wave analyzer was that it had a sweep time of almost 11 minutes, enabling it to cover the frequency range in one run. In addition, the filter characteristics of the HP3590A eliminated a large part of the interfering noise. B. POWER AMPLIFIER The power amplifier used was a HP 467A power amplifier. Its pur- pose was to increase the signal power applied to the transducer and 15 provide a more variable output to the transducer for test purposes. A 10 volt peak-to-peak output was commonly used. A larger output caused distortion in the received signal. C. POLARIZATION A polarizing voltage of 300 volts DC was applied to the mylar trans- ducer to keep it closely attracted to the transducer back, yet free to oscillate. The level of polarization proved to be very critical to the operation of the system. Polarization will be considered in more detail in Section VI. D. TRANSDUCER, REFLECTORS AND FRAME The design of the transducer used in the standing wave system is shown in Figure 5. It consisted of a 30 inch diameter sheet of aluminum coated one mil (0.025 mm.) mylar gjued to a 30 inch diameter plexiglass backing. Mounted in the plexiglass was an four inch wide annular ring of aluminum to provide the second electrode for the electrostatic mylar transducer. The mylar was left unglued over the aluminum ring and this ring was the active face of the transducer. Since the mylar transducer operates essentially as a piston transducer, an air slot was provided to maintain an air film between the mylar and the aluminum plate. The air slot provides a degree of pressure equalization when the system is operated at different depths in the sea, preventing the mylar from stretching. Other characteristics of mylar transducers are discussed by Keller [4]. The mylar was coated with approximately a one mm thickness of Shell Epon 828 epoxy resin mixed with 30 percent Shell Epon curing agent by weight for water proofing. Liquid neopreme was added for additional 16 protection. Uater in contact with the aluminum film would cause the film to separate from the mylar, disrupting its transducing ability. Careful handling was necessary so as not to subject the epoxy to shocks which might crack the epoxy and cause water seepage. Rapid exposure to large temperature variations was avoided for the same reasons. The plexiglass-backed mylar sheet served as one reflector in the standing wave system, while a 30 inch diameter steel plate (one fourth inch thick) served as the second plate. The original design utilized steel plates for both reflectors, since steel has a higher reflectivity than plexiglass, but the plexiglass provided less problems with elec- trical radiation. Also, the plexiglass proved to be easier to mount and easier to water proof. Mounting a full sheet rather than a partial sheet of metallized mylar over the plexiglass tended to improve its The reflecting plates were backed by an epoxy-sand mixture to damp the plates against a standing wave being established in the plates. The backing, designed by Donaldson and Macfarlane [3], was shaped into one and one fourth inch high tetrahedrons for optimum performance. However, if the signal exceeds approximately ten volts rms, the plates still seem to cause enough reflection to cause large distortion in the received signal . In order to maintain a standing wave between the plates, the plates must be kept rigidly parallel (maximum deviation - 2mm). Therefore, the design of the frame was critical to the operation of the standing wave system. While the frame must be sturdy enough to hold the plates parallel, it must not expose an appreciable reflecting area to the acoustic field being generated. The final design of the frame is shown in Figure 6. Sections of the frame located behind the plates out of the 17 acoustic field were constructed of slotted angle iron. Where the angle iron extended beyond the circumference of the plates, rubber was applied to reduce any reflections. Flat iron bars (one fourth inch thickness) were used to maintain the separation of the plates. A coil spring shock absorber was designed to protect the system from sharp jerks when used from a boat at sea. Plate alignment was accomplished by adjusting four mounting bolts equally spaced around the circumference of the plates. Optimum plate spacing (45 cm) was determined by Donaldson and Macfarlane [3] so that sound loss by diffraction was compromised with sound loss by absorption at the reflecting plates. E. HYDROPHONE A one fourth inch long, one fourth inch diameter barium-titanate ceramic cylinder was used as a hydrophone [4]. It was mounted through the back of tfie plexiglass plate and extended through the plate far enough to be flush with the reflecting side of the plate. The probe was then held in place by silicon rubber. By not extending the probe into the acoustic field, averaging of the received signal over the length of the hydrophone is avoided and scattering from the probe was minimized.. F. PREAMPLIFIER Since the hydrophone is a very low magnitude signal and it must travel through 100 feet of coaxial cable to the amplifier, a hign input impedance-low output impedance preamplifier was used near the hydro- phone to improve the signal-to-noise ratio. An NUS model 2010-30 pre- amplifier was used. It provided 30 db. gain and was small enough to provide easy mounting. The preamplifier was clamped to the frame behind the plexiglass plate approximately one foot from the hydrophone. To protect the preamplifier from sea water, it was placed in a copper tube, 18 sealed and then the entire container was encased in epoxy to insure that it was watertight and electrically insulated. G. AMPLIFIER AND RECORDER An HP 466A amplifier was used to provide a usable signal level for the recorder. It provides gains of 20 or 40 db., has a wide frequency range, stability, and low distortion. A B&K model 2305 voltage level recorder was used to directly record the system response. When making a run, the initial and final frequencies were recorded from the frequ- ency counter, and the time of the sweep was determined by measuring the length of recording paper used. From these three measurements, the sv/eep rate was calculated, and subsequently the paper tapes could be calibrated in terms of frequency. Using this method, bandwidths could be measured within +10 HZ for well formed resonant peaks. 19 VI. EQUIPMENT EVALUATION Although the basic design of the standing-wave system described in Section V had been constructed in a previous project [3], no evaluation of the system was possible because of equipment failures. In this pro- ject, the causes of equipment failure were eliminated to enable evaluation tests to be conducted. Leakage of water into the transducer was found to be the most common cause of failure. Besides causing the metal film to separate from the mylar, the water caused a short circuit between the two plates of the transducers. This was especially critical when using the steel plate for one electrode, since the slightest water penetration could short the two plates. In addition, when using the steel plate as one electrode, there was a wery large electrical signal present which tended to mask the acoustic signal in the system. This large electrical signal was attributed to the large area of both electrodes and the close proximity to the hydro- phone. Shielding the transducer produced little improvement, nor did electrically insulating the entire steel plate and the frame to wnich it was attached. Because of these -difficulties, a new transducer was constructed using plexiglass as the transducer backing. See Section V-D for a description of this transducer. The electrical radiation was greatly reduced with the plexiglass transducer, providing an excellent acoustic-electrical signal ratio. An unexpected benefit was also observed during test runs of the new trans- ducer. With the steel backed transducer, instead of a single signal peak at each resonant frequency, there v/ere as many as four peaks 20 observed (Figure 7). These peaks v/ere originally believed to be due to reflections from surfaces other than the main reflector faces, such as portions of the frame. Hov/ever, with the plexiglass transducer, only two peaks were observed, and the secondary peak was of a much smaller amplitude. Although the reason for this improvement is not known with certainty, it appears probable that a portion of the signal is trans- mitted through the backing plates. The epoxy-sand tetrahedrons were designed to transmit this sound energy into the water instead of allow- ing it to reflect from the back of the plate. Because of the large acoustic impedance mismatch between the steel plate and the epoxy-sand mixture, the epoxy-sand backing lost most of its effectiveness and reflections still occurred. The plexiglass has an acoustic impedance .much closer to the epoxy-sand mixture, causing higher effectiveness and weaker reflection from the back of the 'plate. Tests v/ere made of the system performance when operated under various orientations in .the water. In order to determine the orienta- tion which provided the optimum operating characteristics, four orienta- tions of the plates were tested. The plates were first orientated parallel to the water surface, with the transducer on top, and then inverted so the transducer would be on the bottom. The plates were then tested while at right angles to the water surface and finally when the plates were angled 30 degrees from the horizontal plane. With the transducer on the bottom, resonant bandwidths of approxi- mately 700 Hz (frequencies 24-33 kHz) were obtained. With the plates at an 30-degree angle, bandwidths were slightly greater (800 Hz), while when the plates were positioned vertically, bandwidths of over 1100 Hz were measured. All of these resonant peaks were low magnitude, flattened peaks and rather difficult to measure due to distortion. 21 With the transducer placed on top, large, symmetrical peaks were recorded with bandwidths between 700 to 800 Hz. Although the bandwidths were not particularly better with this plate orientation, it was found that the bandwidths remained essentially constant with depth, although tests were limited by the depth of available tanks (eight feet). With the plates in the other orientations, the bandwidths were found to vary with depths due to the change in hydrostatic pressure. Presumably this pressure change caused a change in the spacing between the mylar and the metal plate, altering the transducer characteristics and causing larger sound absorption by the air space behind the mylar. However, when the transducer was lowered face down, the hydrostatic pressure tends to press the mylar against the metal plate. Since the mylar is already attracted tightly to the metal plate due to the polarization, the press- ure changes cause very little change in operating characteristics. Therefore, all subsequent measurements were made with the transducer on top. The system stability (its ability to produce identical results under identical conditions) was evaluated next. This evaluation consisted of a series of runs over the entire frequency range (20-100 kHz) being made over a period of 45 hours. The results were then put in graphical form for ease of analysis (Figure 8). The system Qs were found to vary during the duration of the test. The maximum variation of the 33 kHz curve would correspond to a bubble concentration of 430 bubbles/m^ of 100 micron bubbles. This variation, although not large compared to bubble concentrations expected at sea (several thousand bubbles/m ), was considered to be too large for calibration purpose's. Up to this time, the bubble concentration in the tank water had bean considered small and relatively constant. However, due to the system 22 variation with time, it appeared that the tank's bubble population was large enough to affect the system's operation, possibly due to photo- synthesis by fresh water micro-organisms [5]. It was found that because of the plate's orientation in the water, the upper plate would tend to collect bubbles, causing a greater affect than the actual concentration of bubbles in the tank would indicate. In order to keep the system free of this bubble build-up, a "bubble broom" was built. Sweeping the trans- ducer face before each run, a new time evaluation was made. The results are shown in Figure 10. The curves are now seen to stabilize quickly after the system is immersed, and to remain within \fery narrow limits. The system Qs are also improved over the previous evaluation, showing further evidence of the elimination of bubbles. System bandwidths range between 500-700 Hz over the frequency range. Using these curves, the system's inherent error was determined to be ± 10 Hz in bandwidth corresponding to 500 bubbles/m'^ at 57 kHz. This error is due to the inability to measure the Q curves closer than + .1 mm. Because of the effect of the tank bubbles on the system, a re- evaluation of the polarization was made. Previous investigators believed that the mylar characteristics changed with time when a DC voltage was applied. This change had been attributed to a slow polariza- tion of the mylar transducer [4]. Donaldson and Macfarlene [3] designed a switching network which would alternately apply positive and negative polarizing voltages to the transducer in an attempt to operate with a constant output. This alternating-polarization system was initially used in this pro- ject. However, it was found to be unsatisfactory due to the long charge time of the mylar transducer. The transducer charge time was found to be about ten seconds (Figure 9), while the lowest frequency at which 23 the sv/itching network could be operated v/as about ten Hz. Therefore, the transducer was constantly being operated with the transducer under- going rapid charging. Since the system characteristics change while the transducer is charging, the alternating polarization system was deemed unsatisfactory for this system. Instead, the polarization was manually switched before each run, allowed to charge to a stable level, and then a new run was started. In this manner, the transducer was constantly being operated with the same time of polarization. However, due to the effect of bubbles in the tank, a new evaluation of the effect of a continuous polarization was made to determine if the changes mentioned above were actually due to changes in the transducer characteristics or due to bubbles. Runs were made over an eight hour period with constant polarization. The plates were swept before each run to remove any bubble accumulations. There was no change observed in the system Qs or in the output level. Therefore, it was concluded that a continuous polarization applied to the transducer would have no adverse effect on its performance. However, the magnitude of the polarizing voltage was found to be very critical. A decrease of 15 VDC was found to cause as much as 100 Hz change in bandwidth. As small a change as four VDC still causes up to a 50 Hz change in bandwidth. Therefore, the polarization must be kept within about t 1 volt DC in order to insure consistant results. In order to test the system's ability to measure bubble concentra- tions, bubbles were allowed to accumulate within the system. Runs were then made with the bubbles present. The bandwidths were determined and compared with bandwidths measured when the plates had been swept free of bubbles. The results are tabulated in Table 3. 24 VII. GENERAL CONCLUSIONS A. SUMMARY Although the standing v/ave system has not been evaluated under operating conditions at sea, tests conducted in this project indicate that the standing wave system is a practical and dependable instrument for bubble measurements at sea. The system error of + 10 Hz in band- width is well within the limits which can be tolerated in actual meas- urements. The frequency range with which this instrument can be opera- ted (20-100 kHz) is not as broad as desired. However, further research into the operating characteristics of the standing wave system may provide the answer to this problem. The following characteristics were determ.ined experimentally: 1. System responses were able to be analyzed over the entire frequency range. 2. Best system responses were bandwidths on the order of 500-600 Hz (maximum Q = 150) using the plexiglass backed transducer. 3. The plexiglass backed transducer provided less distortion in the system response. 4. System characteristics do not change with time when a constant DC voltage is applied. System response does change if polarization varies in magnitude. 5. Best system response was obtained with the transducer operated over the reflector. B. RECOMMENDATIONS: 1. Conduct calibration runs of the system in a tank fitted with dust covers and filled with chemically treated water •to kill algae in order to reduce bubble formation in the tank. 2. Design a suspension system for the instrument so that it will remain stationary in the water column and not be affected by the rolling of a research boat. 25 3. Investigate the possibility of battery power for the elec- tronic components in the system in order to reduce ship noises affecting the system operation. 4. When constructing a transducer of the design used in this project, insure that the mylar sheet is cut back away from all sharp edges. This provides a more secure bond for the epoxy to prevent water leakage. Also insure that the epoxy layer is thick enough to provide protection but not too thick to hinder the transducer operation. A one mm thickness was found to be satisfactory. 26 500- 200- o t/1 l-H Q a: 50 - FREQUENCY, kHz Figure 1. Bubble Radius as a Function of Resonant Frequency and Depth for Airfilled Bubbles (3). 27 < to 2: o <_> 100 FREQUENCY, kHz Figure 2. Damping Constant as a Function of Frequency (3) 28 O) E > QJ fO > 3: CO •1— II CD CD C c fC "^ •» "O 1— S- C II o fC c 4- 4-> (/) .< Sl t/) -M cu s- cn x: o c 4-> 4-> OJ u r^ M- O) 0) O r— > 4- fC O O) 3 S ctl \- <4- E O) fO <_) O) n3 03 < -o S- Q. 3 fO 4- t/0 •1— Ql r— •» - CL <_) cC cn.,- UJ •1- 4-> _J (U m t— U- i- :3 UJ rs o E cc f Press ween Tw nteger O -l-> -r- >> CD U E CQ C E 03 fO OJ S- W Z2 CD a> II cr fO "O O) •r- O E s_ Q ^-— ' M- CO (U s- cn or oo CO Cl. 29 TABLE I. SYSTEM ELECTRONICS Signal Genera tor Hewlett Packard 3590A Wave Analyzer Power Ampl ifier Hewlett Packard 467A Power Amplifier Polarization 300 Volt Battery Transducer See Figure 6 Hydrophone One Fourth Inch Barium Titanate Hollow Cylinder Preamplifier NUS Model 2010-30 (30db gain) Aiiipl ifier Hewlett Packard 466A AC Amplifier Recorder B&K Type Voltage Level Recorder Oscilloscope Tektronix Type 545 Oscilloscope 12 Volt Power Sup jpiy 30 a: UJ E 0) +J > CO > rtS 1 Cd LU ct; 1 — 1 LU Ll_ 3 >— < O _J Cu D- cC ca o _i h- cC <: 2: d; o UJ 1 — 1 2: oo UJ 0 CD •r— ■o c rO +» CO en o o CO CTl 31 TABLE II. TRANSDUCER SPECIFICATIONS Plate Diameter 30 Inches Plate Thickness 1 Inch Annular Ring o.d. 26 Inches Annular Ring i.d. 18 Inches Air Slot 1/2 Inch by 1/2 Inch Pickup Access Diame ter 1/2 Inch Tetrahedron Backing Thic kness -1-1/4 Inch Transducer Capacitance .1 yf 32 1/32 Inch Air Holes Aluminum Ring Neopreme Metal ized Mylar Air Slot Epoxy-Sand Backing 3S^ Epoxy V VWvVvVVvV VWWVvWWX A Plexiglass Hydrophone Plate Acess Hole Figure 5. Transducer Diagram, 33 Shock Absorber REFLECTORS ■6" Long, 3/8" Holding Bolts Flat Iron Bar Slotted Angle Iron Figure 6. Diagram of Frame with Plates 34 (31V0S 3Aiivi3y)aa (31V3S 3Aiivi3y) ea Figure 7. Sample's of System Response for Two Trans- ducer Backs-Top: Plexiglass Plate; Bottom: Steel Plate. 35 f 00 N o CO in CD o CO OsJ o o . tn CO > o E Q) CC (1) J3 CO 0) o ■o <£) s_ (T3 r— O) •r- X o (U > s- o cy i/i 3 to S- > 00 (U i-. 3 _ Ln 36 (J Z5 -o t/) E •4- o > S- CD CD s- o cri en O o CO o 0 o s: 0 CJ o 1— 1 1— 1— C/0 < h- ivi _J 1 — 1 o ex: > eC _J 2: O H-H Ci- 37 o r ^1- o o Q O Q O tsj CO N 6 N zr. in LO 6 N zn IX) Lf) 6 0 N CO 6 6 6 6 O O 6 0 6 6 6 6 6 -T" O CO — r o — r o 6 o ■e- LO LO CO O CO LO C\J _ o LO ■a > o E O) OJ J3 CO (U o ■o (/) £= i- LO X 4- o- (/) =5 > cu E