Operation Ivy, Pacific Proving Grounds, November 1952, Project 6.9, Measurement of Material Density with Beta Densitometer

Page 6

This document con&in of 38 pages Report to the Scientific Director MEASUREMENT OF MATERIAL DENSITY WITH BETA DENSITOMETER Lteutemnt Colonel, USAF s 8’3Iw, Alsmos Sctentfflc hboratory University of California STkXt 1446. ATOMIC ENERGY ACT, 3%

Page 7

The objective of beta-densitometer instrumentation at Ivy Mike shot was, primarily, to measure material density near ground surface as a function of time and, secondarily, to test the latest modifications to previous densitometer models. The results showed (1) that thermal or preshock dust is absent at a ground range of about 23,006 it (Sation 690.02), (2) that it is possible to calculate the over-pressure due to air shock alone from *he measured density change, provided that preshock turbulence is not excessive, and (3) that the modifications to the densitometer proved to be satisfactory. The calibration and electronic engineering of the Ivy model densltometer were considerably improved over RESTRICTED DATA -- SECURITY INFORMATION

Page 8

2.2 Problems in Densitometer Design CBAPTER 5 ANALYSIS OF RESLXTS CHAPTER 6 CONCLUSIONSAND RECOMMENDATIONS RESTRICTED DATA-e SECURITY INFORMATION

Page 10

The objective of beta-densitometer instrumentation at Ivy Mike shot was, primarily, to measure material density near the ground surface as a function of time and, secondarily, to test the latest modifications to previous densitometer models.

Page 11

2.1 PRINCIPLE OF OPERATION A detailed explanation of the working principle and descriptions of the working parts of the densltomcter appear in previous reports.‘,’ Briefly, the densitometer consists of a fixed source of beta rays, a detector which responds to changes ln beta-ray intensity, and a recording lnstrument which converts the detector output into an oscilloscope-beam deflection that 1s photographed by a strip camera for permanent record. The attenuation of the beta rays, and hence the detector output, is a function only of the densfty of material between source and detector. Densitometer components are shown ln Figs. 2.1 to 2.3. Three l-curie amounts of Sr” were used as beta sources in each station, except at Station 690.02, where two l-curie amounts were used. 2.2 PROBLEMS IN DENSITOMETER DESIGN The Greenhouse densitometer used 6 curies of Sr” at 50 cm from the detector. The ampllfler was a balanced d-c type, designed to receive the total-gamma and total-beta signal from one RCA 5819 multlplter phototube at one grld :nput and the total-gamma signal only from a second phototube at another grid input. The subtracted outputs of the two plates of the ampllfler were to give only the beta signal with all gamma or background signal canceled out. Disregarding for the moment the difficulty of balancing a d-c ampbfler and the varlatlon of each photomultlpller gain with fatlgue, the llmltatlon of the design itself presented a formidable obstacle agalnst ob&lnlng a beta record ln the presence of any appreciable background (malnly gamma) radlatton. This obstacle vises from the fact that the grld input signal must be limited between the saturation vaj&+&ually a value of plus I or 5 volts) and the cutoff value (usually a value of minus 10 to 15 velte).;Z&Uaearity of ampllfler out&w&&id input is desired, an even smaller spread of grld input must be adhered to. Consequently, to get any usable betasignal component ln the output, the combined gamma and beta signal must be restricted to the usable portion of the ampllfler characteristics. For practical reasons thls means that the magnitude of the gamma signal muat be comparable to the magnitude of the beta signs?. Large gamma background wlll saturate both halves of an ampllfler designed to accomodate a relatively small beta signal. With such an input system, therefore, only weak gamma backgrounds will be subtracted. Moreover, because of the difficulty of malntalnlng a balanced output, subtraction may not tmcessarlly give zero gamma signal.

Page 15

Another aerhus source of difficulty in subtracting the gamma signal with the amplifier used at Greenhouse was the fact that the 5819 photamultipllers were turned on asd were fatiguing several hours prior to shot time. The gain of two photomultipliers, carefully selected as they may be, could vary considerably after several hours of fatigue. Gf the two records obtained at Greenhouse, one was a pure gamma w) record without superimposed beta signal. The other was the n& output of the balancing mystem. In the Buster model a mechanical chopper cutting off the beta flux 400 times per second was used with a single lP2I photomultipller. A band-pass filter and ampllfer displayed the 4OO-cps variation of the beta signal and the superinqosed noise. This was a sound design that could have worked, except for the high spurious noise picked up that made the record unreadable. The amplifier was noisy when disturbed by shock, but ctherwlse it operated sattsfactorily. The excessively long leads, about 20-ft loops, between successive dynodes and photomultiplierpower supply picked up and amplified spurious Sign& of magnitude comparable to that of the detected beta signal. As in Greenhouse no system of field calibration was incorporated in the instrument. Laboratory calibntiou with a pressure tank was relied upon to relate detected beta signal to material density along the beta path. Both the Greenhouse d-c model and the Buster a-c model with completely redesigned electronic systems were used at Tumbler with the idea ofz 1. Eliminating deficiencies from both systems. 2. Testing a field calibration method using removable foils in the beta path. 3. Comparing the systems for best results. 4. Determining feasibility of use at Ivy. 5. Obtainiq density-time measurement for Tumbler. Both the d-c and the a-c systems, &scribed in the Tumbler report, worked satisfactorily. The d-c system is favored over the a-c system for the following reasons: 1. It smoothes out statisUcs and therefore produces a cleaner record. 2. The d-c system requires simpler instrumeutation, eliminating the mechanical chopper The amplifier was used in the linear and nonBne.ar range; this required accurate knowledge of the ampIifler characteristic in the interpretation of results. A completely linear reproduction of the beta signal would Mve been preferable, especlaIly because the variation of beta signaI with material density was logarithmic. An amplifier with a larger voltage variation could be used, but this would require more batteries and more circuitry. A more attractive solution, achieving both Iinear response and simplified circuitry, is to use a. high-sensitivity fastresponse phosphor with the photomultiplier. . 1. F. B. Panel and J-7 Blast Measurement Group, Measurement of Air Density in a Shock Wave, Greenhouse Preoperational Beport, Project 6.4.1.

Page 16

The densitometer used at Ivy Mike shot was the d-c type. Aside from having the desirable features of previous models, it had rn added simplification in lid it did not use a second photomultiplter for balancing the background gamma signal. This simplification was made possible by the special situatton expected to prevail at Mike shot. Calculations by F. B. Ponel predicted negligible background radiation at the tnstrument stations at the times of blast-wave With only one photomultiplier to receive both the beta and the background signal, background radiation (matnly gamma) would increase the total signal and could, if sufficiently strong, drive the electron beam, which ls supposed to respond to the beta signal only, off the face of the osctlloscope. Any thermal* or preshock dust, on the other hand, would diminish the signaL Miscalculation of the gamma-radiation rate could lead to confusion in the tnterpretation of the beta trace, even if the beta trace between zero time and time of blast arrival appears cornParable to the preshot trace. It could mean that there was no detectabxe gamma nor preshock dust. It could also mean that the effects of gamma background and preshock dust neutralized ench other. Tumbler records showed that the first possibility is more likely than the second, especially if the trace returns to its preshot level rapidly. Preshock dust raises the beta trace to a fairly steady level. Decaying gamma, on the other hand, brings the trace down conUnucnisly. As it turned out, calculations of gamma-radiation rate and predictfon on the behavior of the beta trace were correct. Three l-curie SrW beta sources were used at each statfort, except at Statfon 600.02, where two l-curie sources were used. Sources were 100 cm from the 2-mil aluminum fotl which covered the detector. Bach l-curie source of S#’ was contained in a cylindrical stainless steel shell, 1.0 in. in diameter and 0.3 in. deep. The active face was sealed with a 1-mil stainless steel foil. Details of the source holder are shown in Fig. 3.1. Vhen a nuclear bomb is detonated over a dry ground surface, dust is literally exploded from the surface shortly after detonation and usually before the arrivP; of the blast wave. The exact mechanism of this phenomenon is complex and is believed to be due to the lntense burst of .’ thermal energy released by the bomb, hence the term thermaI dust. RESTRICTED DATW SECURITY INkORMATION

Page 18

The detector con&tS of an RCA 5819 photomultiplier and a phosphor glued to the cathode end of the ti. The phosphor is made up of 5-nail transtilbene crystal glued on top of a ‘/(-in. lucfte disk. The disk is formed to fit the contour of the tube face. The phosphor, being far more effective in attenuating beta than gamma radlatlon, la therefore more sensitive to beta than to gamma radiation. The phosphor was effective in increasing detector output to such an extent that it was possible to eliminate the d-c amplifier (i.e., to feed the detector output directly to the oscilloscope) and also to double the distance between source Mauidetector. Doubling the beta path meant doubling the material sample whose density The detector circuit is shown in Fig. 3.2. The lxnver supply is made up from the 30-volt cells of 300-volt Eveready No. 493 batteries. Each power sup$y consists of duplicate 1260volt hanks cd batteries, paralleled for greater reliability. Construction of the power supply in this manner resulted in very short leads, which la an essential requirement for minimizing or eltminatfng spurious signals. The voltages applied to the various pins are as follows: 1. Cathode to dynode No. 1: 180 volts. 2. Between dynodes (No. 1 to No. 10): 90 volts each. 3. Dynaie No. 10 to plate (across load): 270 volts. The plate load was variable between 200,000 and 400,000 ohms with the center of the load grounded. Thie point was selected as ground to permit a balanced-line output to the recording The photomultipller power was turned off by relays which opened to remove voltage from dynodes No. 2 and No. 4. This method was effective. When the power was turned off in this manner, there was no photomultiplier output nor fatigue. This method was parttcularly convenient in that no b&h-voltage circuits had to be turned on or off. The photomulti~lier was made light tight by housing it in a steel tube (Fig. 3.3), which also afforded magnetic shielding. The cathode end was sealed with a 1.5-mil aluminum foil. The recording unit fxdsted of a dual-trace cathode-ray tube, a strip camera, timing lights azxl oaciUators, control and delxy circuits, and associated power suppltes (Fig. 2.2). The hdire asrsembly was housed in a watertight aluminum cylinder (Fig. 2.1). The output (current) of the photomultiplter was fed to a variable (resistance) load, b8lanced about ground and located in the record-ing unit (Fig. 3.4). The voltage drop acroee tbir load m &played dtrectly on the screen of a dual-gun cathode-ray tube. Any 6purioua noire picked up by the beta-signal leads in cable D would also be picked up by 8n add&nml pair of leada terminated across a 600,ooo-ohm load in the battery box (Fig. 3.1) between terminals d and 8. In this way spurious gignals that might be mistaken for betaaigoal var!!Ua~ would be displayed in greater amplttude by the secad or noise-monitor@ trace of the cathode-ray tube. The cothodt-ray tube M a 3U2Pll manufactured by the Electronic Tube Corporation. The parer suppiy was made up of dry batteries as follows: Gromck anode8 No. 2 and No. 3, unused deflection plates, and center tape of signal leads. -600 to -900 volts: focusing or erode No. 1. 180 and 120 rolls: positioning voltages.

Page 22

AR batteries x.&!ng up these voltages were !n duplfcate and connected !n yeraRe as fnsurance NT6 Willard B-vc~i wet cells were used for filament power !n l!eu of dry cell8 because of the greater current requirement which these cells could furnish The ground points of the beta-signal circuit and of the cathode-ray tube were connected to the recording unit case ulrough a lOO-megohm resistor to dra!n any static charges and to provide a h!gh impedance to any possible short to the ca8e from any pprt of the recordingA M-mm Kodah Clne E strip camera photographed the var!at!m of the signals on the cathode-ray-tube screen and generated a ttme base proportional to the f!lm speed. Since the speed of the camera varies noticeably, accurate time interval8 could not be determined by measuring film lengths. Oscillator circuits were used to generate 5- and 1004zycle pulses to neon bulb8 mounted on the oscilloscope screen. Thus a series of 0.200 and 0.01~see t!m!ng marks were included in the film record, directly below the beta-signal trace. The O.%O-set The camera was powered by a bank of 221/,-volt dry batteries. With approximately a 2-amp dra!n, the batteries were reliable only for two to three run8 of about 1 min each. However, for the measurement time involved the power wa8 sufficient. For repeated trial NIUI the camera was powered by external batteries connected through mon!tor!ng cable C (Fig. 3.5). Welnbrldge oscillator and peaking circuit. The oscillator output was a voltage spike great enough to flash the GE51 neon bulb and sufficient to give sharp t!m!ng marks on the record!ng film. Filament and plate voltages were supp!!ed by dupl!cated dry batteries. The control system consl8ted of two exteinal ma8ter Kdgerton, Germeshau8en 6 Crier, photomult!pl!er battery box. The master relays, ncrmally open, closed to give timing signah at E-l min and H-5 set, respectively. The H-l nu May activated the &cults of a hold-downrelay and five swttch relays (Fig. 3.7), which turned on the power supplies to the photomult!pl!er and the f!laments of the cathode-ray tube and timing oscillators. The H-5 set retay activated a hold4own rehy, a 8w!tch relay for the camera-motor power, the foil cal!bration ckcuit, and the power for the delay relay. After a predeterm!ned time the delay relay closed and burned out a fuse which turned off power to all internal relays and to the came- motor. Although the relay8 were rated at 24 volts, 45 volt8 wa8 u8ed to g!va tbam a greater closing force and prevent the cmtacts from. Chattering when hit by a shock. The heater voltage for the delay relay (Amperite 6NO60) was obtained from the (Fig. 2.1). covered by a removable plate, wa8 cut on the side of the alum!num cylfnder which housed the recording un!t. The tolloaring checks and adjustments could be accompltshed through the windows 1. Check functioning of Uming l&S. 3. Xn8ert or replnce control fuse. 4. Adjust electron beams on CRT screen for focus, brlghtne88, and po8!t!on. 5. Set photomultipler load to locate output-signal level at desired po!nt on screen (f!na.l RESTRICTED DATA 1. - SECURITY INFORMATION

Page 26

ray tube correqxn&ng to …