USVI Public Records

Document text

Technical Memorandum No. 493 5 ) I I Results of Air Gun Tests (St. Crolx):' I November 1978 ,V.A Mamtl)en, C.M. Gogos. and J.E. Barger I •LOQLQ~oflSOO CIJanuary 1979 I Prepsred for: Dlefnse Advanced Research Project* Agency Reproduced From Best Available Copy H9 2 .0761.1 4SOON The vo aNO conclusions contained In •this Oocunww wo "W of the authors W4l should not, be Interpreted m• rep~re- IWtk .9t0 OftfilW I6WOtie either" ex .~qmW Or im0144I of the Defensew Adaie Flsec Prjc~ts A c or ft U.& : • .,, -• J . . .. - . . ii i i inll i iiii | Technical Mem,4-4A.ý, No.!493 "- FESULTS OFAIR GUN TESTS (ST. CROIX): _NOVEMBER 1978m, -W W.R. amblen)1 I!E./Barger Con rA.t No. N00039-78-C-e--62 ,A O rde ,rs-..3532 I BRN Proj 7 No1. 0086 Prepared by: Bolt Beranek and Newman Inc.r" 50 Moulton Street Cambridge, MA 02138 Prepared fort Defense Advanced Research Projects Agency 1400 Wilson Boulevard Arlington, VA 22209 THE VIEWS AND CONCLUSIONS CONTAINED IN THIS DOCUMENT ARE THOS'. OF THE AUTHORS AND SHOULD NOT BE INTERPRETED AS REPRESENTING THE OFFICIAL POLICIES, EITHER EXPRESSED OR IMPLIED, OF THE DEFENSE ADVANCED 'RESEARCH PROJECTS AGENCY OR THE U.S. GOVERNMENT. :1: • i .'• •--i.••. " •-]J _____ .. . .......... .. . ! Tech Memo No. 493 Bolt Beranek and Newman Inc, I I TABLE OF CC•NTENTS page I LIST OF FIGURES ................. .......................... iv SECTION 1. INTRODUCTION ............................ 1. 2. PROCEDURE . .............................. 2 2.1 Equipment .............................. 2 2.2 Texting ................................ 6 3. RESULTS ............................. . 8 3.1 ESL ........................ ............ 8 3.2 Timing ... ................................. 8 :3.3 Frequency-Depth Dependence .............. 9 .4. ANALYSIS .... ............................. 12 5. SUMMARY AND CONCLUSIONS .... ...... 16 APPENDIX A.. ..... .... .................................. 18 t ,o 'Ds Il Tech Memo No. 493 Bolt BeranLk and Newman Irc. LIST OF FIGURES page SFigure 1. Block Diagram of Experimental Apparatus ................................ 2. Energy Source Level for the Western Geophysical Air Gun as a Function of Frequency ................................ 10 3. Example of On-Line Data Analysis: Gun Failed to Close ......................... 14 4. Example of On-Line Data Analysis: Proper Gun'Operation .................... 15 A-1. Shuttle Trajectory for Four Values of the Coefficient of Friction ............. 22 Table I i............ iv Tech Memo No. 493 Bolt'geranek and Newman Inc. j 1. INTRODUCTION "This technical memorandum presents the results of source I level measurements made-on a Western Geophys'cal (WG), Walker- type airgun. The objective of these tests was to determine if this type of airgun could meet the requirements Adeveloped-in &-Ref--1' for a single element of an active array. The energy source level (ESL) required is 223 dB -Pe--uieý in the fundamental I frequency band.-- SEarlier tests (Refs.2 and 3), with air guns designed and built by Bolt Associates showed that this source level was T achieved only at low fundamental frequencies (shallow depths). As the depth was increased, the ESL's rolled off sharply.' It - was felt that this was due primarily to extended ipen port dura- * tions which allowed cushioning of the collapsing bubble. The - WG air gun has several features which indicated that the open port duration would be much shorter than previous guns and that this time would be relatively independant of depth. Thus it was hoped -' that high ESLs could be achieved at the required fundamental frequencies. hr .......... SA.7." . . , * , ', Tech Memo Nc. 493 Bolt Beranek dnd Newman Inc. 2. PROCEDURE 2.1 Equipment The measurement program was conducted during the period, from 6.to 17 November 1978 in the deep ocean channel 6 miles north of Christiansted, St. Croix, USVI. The measurements were staged from the laboratcry barge YFN 1126, which is operated by the Key West Detachment of the Naval Air Development Center. The barge was outfitted with the experimental apparatus shown schematically in Fig. 1. The WG air gun is configured quite differently from previous air guns and this necessitated slightly different handling. The WG design consists of two concentric cylinders with sets of matching holes or ports. As tne gun fires the inner cylinder is slid longitudinally, opening the ports as the holes align and then, as the inner cylinder continues to move, closing them. To fire the gun again, the inner cylinder is slid in the opposite direction, repeating the process. The volume of the gun is about 1,100 cubic inches. This design requires two air supplies and two firing circuits; one air supply and one firing circuit for each end of the gun. In addition, the present design must be fired with the axis of the cylinder horizontal. To achieve these requirements, WG supplied a bridle which supported the 'gun in a horizontal'position and split the air supply in order to supply both'ends of the gun., The bridle-air gun assembly was connected to the cable bundle with quick-disconnect type of connectors for ease of assembly. .2 4 . Tech Memo No. 493 Bolt Beranek and Newmin *nc. I F14 w c -J- I 11 4 0 z cc 0 LL 400 cc 'I5 ~ifl i Ill U- Tech Memo No. 493 Bolt Beranek and Newzn Inc. The cable bundle consisted of three eIectrical cables; two firing cables and one pressure trpnscucer cabie, plus the air hose. This bundle was taped to the strength member as the zun was deployed. The gun was deployed from an oceanographic winch on the 01 deck of the barge with the aid of a Pettibone hydraulic crane for handling the gun at the surface. In this respect, the handling was identical to previous tests, (Ref. 3). The anticipated testing at 5 and 6 thousand ps:. predicat'ed a new air supply system. The air compressor was an Ingersoll Rand, having 100 scfm capacity at 6,'000 psi. The air compressor, fed a new air supply manifold built and supplied by WG. The manifold provided precise control over compressor pressure and air gun pressure. The air gun pressure gauge on the manifold was calibrated with a dead weight calibration just prior to installation on the barge and all pressures were referenced to this gauge. Finally, 2,000 feet of 3/8 Synflex air hose, rated tc 7,500 psi was supplied by WG. A dynamic pressure transducer, PCB Model 111A22, was installed in the Rir gun to measure the air-discharge history of the chamber. The electrical cable of the pressure transducer provided dc power to the transducer amplifier via a power supply topside, and the signal line up to the measurement equipment. The nominal sensitivity of the prez3ure transducer is 1 mV/psi with a full-scale range of 6,000 psi. The transducer was located at the end of an 18-inch airhose .that-was connected to the gun's air chamber.' This mounting system provided shock isolation for the transducer. 'Earlier direct mountings had shown a high inci- i dence of shock-related transducer failures. Tests conducted by WG indicated that no artifacts were introduced in the pressure trace with this mounting system. F i V .: Tech Memo No. 493 BHlt Beranek and Newman Inc. jThe air gun was firld from a dc power supply and a switch box. The switch box -inrply provided a means to select which side of the gun to fire. The measurement hydrophone, Type F-50, Serial No. 88, was L lashed to a weighted nylon line that was run through a block on a davit located near the stern on the port side. The hydrophone t sensitivity, based on calibrations made in the BBN hydrophone calibration facility before testing, was -218.5 dB v/re uPa. Frequetit checks on this calibration were made during the test program with a G-19 hydrophone calibrator. The hydrophone was lowered to a depth of about 300 ft and about 80 ft aft of the air gun line. The gun-to-hydrophone distance was obtained by measuring the elapsed time between the pressure-transducer air-discharge pulse and the direct acoustic arrival. Both the hydrophone and the pressure signals were "" displayed on a 2-channel oscilloscope and tape recorded. In addition, the firing current was also recorded. The hydrophone signals were spectrum-analyzed upon arrival. Each shot was analyzed on-line, in addition to being tape recorded. Each scurce-level data point required the following SL processing:, * Capture of the acoustic pressure waveform - Fourier transform of each waveform - Summation of the, energy in all anaiysis bands that comprise the fundamental-frequency band - *Measurement of source-to-receiver acoustic transit time and computation of distance i5 Tech Memo No. 493 Bclt Beranek and NJewman Inc. • Calculat4on of transmission loss on the basis of spherical spreading * Calculation of the energy source level in the fundamental- frequency band. In addition, the •hamber pressure-amplitude time history was also photographed on-line along with the acoustic pressure waveform for the purpose of both gun performance analysis and as a measure of the acoustic transit time. 2.2 Testing The test was conducted in accordance with the objectives out- lined in the test plan, Ref. 4. These objectives were to establish as quickly as possible the required parameters (depth and timing insert) for obtaining the highest ESL at the desired fundamental frequency and to then obtain that data point. The depth parameter Prises because the WG air gun is cylin- drical in shape and previous theoretical and experimental work indicated that the resonant frequency of a cylindrical bubble is higher-than the resonant frequency of a spherical bubble of the same total volume at thesame depth. Thus, a new frequency- depth relation had to be obtained for t gun. It was hoped that the 100 Hz fundamental frequency cou.Lu be obtained at a sig- nificantly shallower depth'than previously. This would provide important system (and source level) benefits. It was hoped to vary the timing in order to prevent.bubble cushioning as had bee..i seen in other air guns. To obtain'the' optimum time of 5 to 7 msec a series of.timirng washers were to be inserted in'the gun. These washers varied in thickness and * 6 Tech Memo No. 493 Solt Beranek anC Niwman In;c. essentially spanned the range from 0 to 5/i6 In. in /176-in. steps. Actually, it was discovered, that these washers were ""much less critical to the timing than anticipated. The major impediment to data acquisition was the reliability Lof the gun when operating at depth. Frequently, at depths over 400 feet, the gun would fire once then hang "open,," necessitating retrieval and reworking of the gun. These failures typically appeared to be one of two types. Either water would infiltrate . behind the open end of the piston (inner cylinder) preventing complete piston travel thus causing the gun to remain open, or compression of the (ambient) air behind the open end of the picton .. would cause the piston to rebound, leaving the ports open. Once the pcrts were open, there was no way to reseal the gun or to move the piston, short of retrieving the gun to the deck. Thus, toward the latter stages of the test when the gun was being fired at depth, frequently only one data point per lowering was obtained. The only other difficulty experienced were breaks "* in the electrical splices. -. t4 7 Teh Memo N o. 4)3 3o•t .Bera,-ek anid hiewman Inc. 3. FESULTS 3.1 CiL -•I le malor result of ;his test is that the , ar g':n an-ears to rcvdean 'Z f 2. to 219 dB -Pe ;jas Ier1nto frequency (depth) over the range fron 25 t4 1- Hz .--. 1hHs is demcnstrated 7ost clearly by Fig. 2, whic,, plots the measured ESL as a function o'f frequency . Table 1 lists .•,e average -S'z..s ottained during the course of the test.•/The. fact that the ESLs are independent of depth confirms the contention that-the param- eters of this gun (when it operates) are Idependent of depth. if this indeed Is the case, then the ESL would not tegin to roll off until the fundamental frequency is on the order Of the reciprocal of the open p.or, duration. For this gun that would be approximately 150 to 170 Hz. 3.2 Timing There are two concerns with r(,ard to the timing of this air gun. The first concern is tha the open port duration lie in the 5 to 7 msec range. It was found that the duration (fortuitously) fell within this range irrespective of what timing Insert was used. The inserts we.e varied between 0 (no insert) and 5/16 in.' (the maximum) and while some changes in gun per- formance were' --9ted (mainly source level changes at the low frequen,-3s), oasically the open . tt duration remained constant at about 6 msec. Consequently, timing insert "B" (1/4 'in.) was used for the remainder of the shots. iF Tech M~No. 493 Bolt Be-an ek and Newma~n 1111c. The second concern, while not immediate to th-s test, the amountz of Jitter i.n tb-e gun firing~. In ztost ar~aly343, the t I. e bDetween the inftlation of the firing zýrent and the inltiatlzri of th-P inressure dr-oo in tLhe :;un was 7easured and c071oared for shot.'s I-a')en under ientioa-. oon:I4ti ons. VWhile tlh-e data base is too small to 'Le cor'clus-Ive, it -Ices indicate t-1at. this t."-e may vary by several- mlil.seconds. Th-ýIs will impact the ability to array this aIj..gun. 3.3 Frequency-Depth Dependence A5 had been hcped, the WG achielred a fundamental' frequency of 100 E~z a!; a snallower depth than previous 1,000 cubic in. air guns,,approximatel~y 1,400 ft rathe I than 1,6'0 ft. A linear, reg.-esslIon on the data incidates that the frequency is propor- tion~al to the depth to the 0.929 power. This is virtually identicall to the dependence of the other air guns, however, the constant of proportionality is less.,. For this gun, the fLun- Iament-al'frequency is given by f *0.13 (d31, 929e is the depth In feet. 1 ; : Tech Memo Nio. 493 Bolt Beraneik and ?Netian Inc. 0 0 LAJ ~LiJ Uo. LbJ 0La LAJ 0 01 I.% Tech Memo No.. 493 Bolt Beranek and Newan"ins. I TAILE'I I , Depth ,;ft) ESL, dB re IAPaz sec 200 219.5 at 20 Hz average of 4 shots I 40oo 218.3 at 35 Hz average of 7 shots 800 219.1 at 65 Hz average of 6 shots 1,200 218.2 at 92 Hz average of 2 shots -1,40o 218.8 at 110 Hz average of 1 shott tested at 5,000 psi, timing insert Stested at 6,000 psi, timing insert D. i I , . . . * - Tech Memo No. 493 Bolt Beranek and Newtman ;nc. 4. ANALYSIS Setting aside the question of gun reliability, the analysis of the data is straightforward. The most important oolnt is that the open port duration at 5 to 6 msec, is com- patible with fundamental frequencies of at least 100 Hz. Thus, no roll-off in ESL was seen. It would be desirable in a successor test to prove the exlstence'of such a roll-6ff and to quantify its location in frequency. The theory of the gun timing is more complex. Presently, the best hyposthesis is that the value of the static friction between the inner and outer cylinders is the controlling factor in the open port duration. The following facts and observations support this. First, the volume of high pressure air that initiates the piston movement is so small that once the piston begins to move the force decrease so rapidly (due to the expanding volume, dropping the pressure) that the piston only sees an impulse. Thus, there is no effective way to vary the driving force. Tnis is sup)orted by the data. The timing inserts which control the rate at which high pressure air flows into the volume behind the piston had little effect on the open port duration (piston velocity) but effected the time between the firing current pulse and the pressure dump markedly. Thus, while ft may have taken longer for the air to flow behind the piston with a timing ins3rt in place, the force sufficient to break the piston loose wts the same as the force in the absence of a timing insert. Second, experiments by WG indicated' that the piston velocities are approximately constant during the open port duration. This 'indicates that the inner cylinder is effectively free from external forces during this time. Our 12 Tech Memo No. 493 3olt Beranek and Niew,•n Inc. T test indicates that this is true only on the surface where the compression effects of the ambient air in t-e "dead" end of t0e gun are minimal. As the gun went down in depth, the higher ambient pressures caused a premature stopping of the piston, causing the gun to hang open. Generally little can be presently envisioned to increase the I gun's perfcrmance. It was noted, however, that the l/e time of a total air dump (which occurred when the gun hung open), was somewhat longer than desired. This may bi neen in Fig. 3, which displays the photographs obtained froir a 5000 psi, 1200 foot deep shot. The gu:n hung open as may be seen from the pressure trace. This %ould indicate that by increasing the port area a greater percentage of air could be released in the 6 or so msec available. An effective increase in the port area could be obtained by drilling out the ports, smoothing the edges to increase flow, or both. Note in Fig. 4, a "good" shot that only 68' percent of the presRure was dumped; opening the ports should raise this figure. Figures 3 and 4 provide a good comparison between a k. cushioned shot and a normal shot. L Ii I 13 II Tech Me~mo No. 4*93 Bolt Beranek and Newman Inc. a -a 3 CD 4Cw r(A Tecn Me mo No. 493 Bolt Beranek and Newman Inc. t-ti 5 15 Tech Meino No. 493 Bolt Beranek a2.d Newman Inc, 5. SUMMARY AND CONCLUSIONS 1. The maximum ESL for the WG gun is 218.8 db re uPa 2 sec at 110 Hz generated at a depth of 1,400 ft and a pressure of 6,000 psi. 2. The timing of the gun appears to be very good and is relatively insensitive to changes in the gun configura- tion. 3. Performance of the gun could be slightly improved by increasing the effective'port size. 4. The reliability of the gun operation is poor and must be addressed in future tests. 16 Tech Memo No. 493 Bolt Beranek and, Newman Inc. I REFERENCES 1. J.E. Barger and P.W. Smith, Jr., "Description of a Long-Range Impulsive Source Surveillance System (U)," BBN Tech Memo 1o, 358, 30 June 1977 (CONFIDENTIAL). 2. J.E. Barger and C.M. Gogos, "Results of an Air Gun Test at, Key West, December 1977," BBN Tech Memo No. 423, January 1978. 3. J.E. Barger, C.M. Gogos, and W.R. Hamblen,"Results of Air Gun Tests at St. Croix, April 1978," BBN Tech Memo No. 457, August 1978. 4. W.R. Hamblen and C.M. Gogos, "Test Plan for St. Croix, November 1978," BBN Tech Memo No. 471, October 1978. I 3-T .i .1i "* 7 ':3 Tech Memo No. 493 Bolt Beranek and Newman Inc. APPENDIX A The following is a simple analysis of the initial shuttle dynamics, immediately after firing, of the Walk'_er-type air gun. This analysis describes only the expansion of firing-chamier air, and not the compression of recoil-chamber alt. The analysis, therefore, appli.s only during the period from firing to port- opening. The system is modeled as shown in the sketch. The in- stantaneous piston volume is •. linear function of the piston PISTON.AREA A POSITIVE PISTON TRAVEL FIRING CHAMBER ; VOLUME Va . INITIAL PISTON VOLUME Vi !I travel distance x. The total volume of the system before firing is V0 . After the valve opens, tne total system Volume is V. 17 V - 0 + V, +Ax . (1) We will simplify by taking the 'air expansion to be adiabatic 1 'and assume no throttling losses at the valve. Then the pressure P acting on the piston is given by8 J .......... : ... . .... ' I Tech Memo No. 493 Bolt Beranek and Newman Inc. [V P/P 0 = (V0 /V)V , (2) where v is the rati, of specific heats. The force F acting on the piston is partitioned into a S"" friction drag force Fd and an acceleration force to move the "shuttle, having mass M. F F AP A MR+Fd + (3) * - The equaticn of piston motion is obtained by combining Eqs. 1, 2, and 3. S... a(bu-\-l) , (4) where ýu' u + Ax/V0 Uo = 1 + V /V a - AFd/VM b - APo/Fd The equation for piston velocity v - u is obtained by II integrating Eq. 4 with respect to the normalized piston dis-. placement u. a(bu--l) du J dv 7 vz S 2 v' - 2a[ i- u•' u. ulruJ*-u(5) S-2 19 + 19 I' ~~- - - - - -- ---- Tech Memo No. 493 Bolt Beranek and Newmia n Inc. The position of the piston u is obtained as a function of time by integrating Eq. 5 with respect to u. (2aYJjl u. -u + U3 u du (6) 0. We have simplified the expression by taking v = 1.4 for air. Equation 6 was solved for a series of parameters by using Simpson's approximation. The paarameters used were: V0 - 1.46 10-4 m' V1. 1.J7 10-1 m3 A - 9.0 I0-1 m2 PO 3.45 10' N/m2 (5000 psi) M - 3.17 kg (70 lb) Fd - Cf 9.56 l0' N (hole area totals 43 in.2 ) Cf(drag coefficient) - 1/4, 1/6, 1/8, '1/16. The calculated values of piston displacement x are plotted f on the vertical scale, in mm. The time-after firing is plotted on the horizontal scale in ms. The holes begin to uncover at ,a displacement of about 10 mm, and then close again at P. dis- placement of about 21 mm. These two displacements are marked on the graph. . When the dragcoefficient is as large aa 1/4, the initial air charge is insnffic~ent'&o fire the gun,'as the piston comes to rest after 3.4 us at a position of 6.5 mm short of hole overlap. The length of the air passages from tU "iring chamber 2 Tech Memo No. 493 Bolt Beranek and Newman Inc. back to Lhe main chamber is abo%;t 6 in. Theref'_re, a new charge oi firing air will begin to replenish air In the firing chamber behind the piston after about 1 rns. Thus, the gun would eventually fire, but it would not reseal. When the drag coefficient is smaller, and equal to 1/6, the gun will fire after 2.3 Ms, but will not reseal- When the drag c6efficient is even smaller, and equal to 1/8, the gun will fird after 1.9 ms and reseal at 3.1 ms. When the drag coefficient is only 1/16, the gun will fire after 1.5 ms and reseal at 2.3 ms. Since the firing times are all larger than the replenish- ment time, more air will be available for firing than was assumed to be available in the simple model. Therefore, the gun should be expected to start firing sometime in the range of 1 ms to 2 ms. It appears that the shuttle inertia is such that the open durations would be in the range of 1 ms to 2 ns. Since these durations are shorter than the observed durations, the retarding force of the recoil chamber must be very important when the shuttle reaches about mid position. Nevertheless, the jrange of initial firing times is apparently about 1 ms for coefficients of friction in the range of 1/6, to 1/16. "I!. ;., 21 Tech Memo No. 493 Bo't Seeanek and Ne,.v',,! Inc. L~ui zC 3c3 CN 0 ~00 I-UJ IN34O-SC RO-S OiOi, I