M+*

]

MARTIN MARIETTA ENERGY SYSTEMS L

3 4145b nasanq 3

ORNL 1479In strumrtiwifciHon

A PROTON RECOIL TYPE

FAST-NEUTRON SPECTROMETER

OAK RIDGE NATIONAL LABORATORYOPERATED BY

CARBIDE AND CARBON CHEMICALS COMPANYA DIVISION 0^ UNION CARSIOC ANO CARBON CORPORATION

QQ3POST OFFICE BOX P

OAK RIDGE. TENNESSEE

•-N

ORNL-1^79

This document consists of 36 pages.Copy^ of 311 copies. Series A.

Contract No. W-7l»-05-eng-26

Physics Division

A PROTON RECOIL TYPE FAST-NEUTRONSPECTROMETER

R. G. Cochran

K. M. Henry

Date Issued

APR 2 19St

OAK RIDGE NATIONAL LABORATORYOperated by

CARBIDE AND CARBON CHEMICALS COMPANYA Division of Union Carbide and Carbon Corporation

Post Office Box P

Oak Ridge, Tennessee

••"A"~—v uanAwEs

3 44*b 03S3n4

- 2 -

ORNL 1V79Instrumentation

INTERNAL DISTRIBUTION

1. C. E. Center 29. K. Z. Morgan

2. Biology Library 30. J. S. Felton

3. Health Physics Library 31. A. S. Householder

k-5. Central Research Lflp&ry 32. C. S. Harrill

6. Reactor Experimental,^ 33. C. E. Winters

Engineering Library 3U. D. S. Billington

7-12. Central Files 35. D. W. Card-well

13. C. E. Larson 36. E. M. King

Ik. W. B. Humes (K-25) 37. E. 0. Wollan

15. L. B. Emlet (Y-12) 38. A. J. Miller

l6. A. M. Weinberg 39. R. B. Briggs

17. E. H. Taylor kO. L. D. Roberts

18. E. D. Shipley 1»-1. R. N. Lyon

19. A: H. Snell k2. W. C. Koehler

20. F. C. VonderLage U3. W. K. Ergen

21. R. C. Briant kk. E. P. Blizard

22. J. A. Swartout I4-5. M. E. Rose

23. S. C. Lind k6. D. D. Cowen

2k. F. L. Steahly 14-7. P. M. Reyling

25. A. Hollaender k&. G. C. Williams

26. W. M. Good 1+9. K. M. Henry

27. M. T. Kelley 50-7li-. R. G. Cochran

28. G. H. Clewett 75. M. J. Skinner

EXTERNAL DISTRIBUTION

76. R. F. Bacher, California Institute of Technology77-311. Given distribution as shown in TID-l)-500 under Instrumentation

Category. *' '

DISTRIBUTION PAGE TO BE REMOVED IF REPORT IS GIVEN PUBLIC DISTRIBUTION

- 3

Table of Contents

Page

Abstract c

Introduction g

Equipment q

Proportional Counter g

Electronic Components lg

Method of Operation 21

Calibration Data and Calculations 22

Spectral Measurements 20

Conclusion ^c

Acknowledgements og

- k -

List of Figures

Figure

I Hydrogeneous Radiator .........•••••<>•'••• •

2a Proton Recoil Neutron Spectrometer, Schematic Diagram ... 9

2b Photograph of Counter, Radiator-Absorber Wheel, andPreamplifier Assembly ..........••.••••••• 10

2c ^trirasr^iwtf^c^^ n

3 Fast-Neutron Proportional Coincidence Counter. ....... 13

k. Triple-Coincidence Proton-Recoil Proportional Counter. ... Ik

5 Integral and Differential Bias Curves for Alphas. =..„».. 15

6 Counter Voltage vs. Gas Amplification for 3-Section Counter 17

7 Counter High-Voltage Power Supply .. . . ....... - - 1°

8 Coincidence Circuit ..........•.•»•••••• 20

9 Stability Test; Alpha-Source Counts from Front Sectionof the Counter, Set on Alpha-Peak ........... 23

10 Van de Graaff Data .......... ,.......•« 27

II Polonium-Beryllium Spectrum Source PB-236 ......... 3°

12 Polonium-Beryllium Spectrum Source PB-256 .......-• 31

13 Polonium-Boron Spectrum Source PB-235 .......•••• 33

Ik Polonium-Boron Spectrum Source PB-281 3*

Page

- 5

A PROTON RECOIL TYPE FAST-NEUTRON SPECTROMETER

R. G. CochranK. M. Henry

ABSTRACT

A triple coincidence type proton recoil spectrometer has been

developed for measuring neutron spectra of interest in shielding design.

The instrument is composed of a triple-section proportional counter with

suitable radiator-absorber wheels, amplifiers, electronic pulse height

selectors, and a triple-coincidence circuit. The energy resolution is

comparable to photographic plates which are at present the best method

for neutron spectra measurements. Considerable effort has been made to

obtain maximum sensitivity by optimizing the available parameters. The

spectrometer calibration was checked by measuring monoenergetic neutrons

from the ORNL Van de Graaff generator. Excellent agreement with calcu

lations was obtained for both energy and intensity.

Several spectral measurements of polonium-beryllium and polonium-

boron sources have been included in this report as examples of the type

of measurements that can be obtained with this instrument.

- 6 -

INTRODUCTION

In order to design shields for mobile reactors optimized for minimum

weight and maximum absorption, an accurate knowledge of neutron spectraand angular distribution is needed. Therefore, the development of a

neutron spectrometer at the Bulk Shielding Facility was undertaken to

provide the shielding group with asuitable instrument for measuringneutron spectra from reactors under various conditions. Several methods

of measuring neutron spectra were investigated. It was eventually con

cluded that a "proton recoil" type instrument using amultisection

proportional counter gave the most promise of fulfilling the requirements

of the shielding program.

The basic principle of most neutron measuring devices is that of

counting recoil protons from ahydrogeneous radiator (Ft*, l).

(1) J. H. Coon and R. A. Nobles, Hydrogen RecoiTProportional Counterfor Neutron Detection, MDDC-96 tJune 20, 19^6).

(2) K. M. Case, Tb£pry of Neutron C^u^tersUs^MDDC-92 (Feb. 15, 19^5)-

Beam of

Neutrons

- 7 -

Recoil Protons

CountingDevice

Hydrogeneous Radiator

Fig. 1

Protons from a hydrogeneous radiator have an isotropic distribution in

the center of mass system and have the following simple energy relationship:

Ep =En cos2 9

In order to utilize the proton recoil method in a neutron spectrometer, a

number of design features must be incorporated. The radiator must be thin

enough to insure that self-absorption is small compared with the energy

resolution desired. The thinner the radiator the better the resolution;

however, the sensitivity of the instrument is approximately proportioned

to the number of hydrogen atoms in the radiator and the hydrogen cross

section. In order to realize maximum sensitivity for a given energy

resolution, radiator thickness, counter solid angle, and detector width

are optimized for each neutron energy to be measured.'3/ Even so, this

type of instrument is at best an insensitive device requiring at least

1000 neutrons/cnr/sec at the radiator in an energy interval to register

counts above the normal background.

(3) B. R. Gossick, General Principles of a Proton-Recoil Fast-NeutronSpectrometer, ORNL-I283 (Aug. Ik, 1952).

8 -

EQUIPMENT

The Bulk Shielding Facility fast-neutron spectrometer consists of

a three-section proportional counter with a suitable radiator-absorber

wheel between the counter and the source of neutrons (Fig. 2a, 2b, and

2c). Each section of the counter is connected to a preamplifier which

is in turn connected through 75 ft of coaxial cable to a pulse amplifier.

The output of each amplifier is connected to a pulse height discriminator

(referred to as P.H.S. in this report), the trigger tube of which is in

turn connected to a monitor scaler. The P.H.S. of each amplifier is

also connected to a section of the three-fold coincidence circuit.

Coincident counts are recorded on a scaler connected to the output of

the coincidence unit. A well regulated power supply furnishes voltage

for the proportional counter tube. The radiator-absorber wheel is

rotated and positioned by use of a selsyn system.

Proportional Counter

The proportional counter used in this instrument is probably the

most important single component. The counter must have good energy

resolution and must produce pulses with a fast rise time. In addition,

the counter must give a low background from neutrons and gamma rays

passing through it.

The three-section counter used in this instrument was developed

after many experiments with counter shapes, sizes and gas fillings. <•The first counters used in the spectrometer were cylindrical with a

(!+) B. R. Gossick and K. M. Henry, Neutron Energy Spectrometer,ORNL-711 (June 12, 1950).

Fig.2a. Proton Recoil Neutron Spectrometer, Schematic Diagram

UNCLASSIFIED

DWG 14706

10

I

Fig. 2b. Front View Showing Radiator Absorber Wheel Assembly.

Fig. 2c. Side View Showing Counter, Pre-Amplifiers and Selsyn Motor.

12

mica window on one end (see Fig. 3). The radiator-absorber wheel was

placed directly in front of the counter window with the appropriate

radiator-absorber combination or an alpha-calibration source. It was

found that these axial counters gave the best pulse rise time when

filled with an argon -2$ CO2 gas mixture at a pressure of 10 cm Hg.

Since these counters gave rather poor resolution, a three-section counter

was developed in which the protons are brought in through the side (see

Fig .!»•). A comparison of the resolution of the first counter design

and the present counter is given in Fig. 5- These curves were obtained

by placing a plutonium alpha source in front of the counter and obtaining

an integral bias curve. The integral curves were then differentiated to

obtain the differential bias curves. The point of inflection of the

integral pulse height curve corresponds to the peak of the pulse height

distribution for monochromatic alpha particles from a thin plutonium

source. This fact is used throughout the calibration of the spectrometer.

It can be seen in Fig. &i that the entire neutron beam passes

through the counter. Since this method normally causes an excessive

background, every effort was made to minimize the background effect.

(Background measurements are made by using a position of the radiator-

absorber wheel that introduces only an aluminum absorber and no radiator',)

The background is minimized most successfully by lining the counter with

10 mils of platinum and filling the counter with spectroscopically

pure krypton gas instead of the argon-C02 mixture previously used.

Further purification of the gas was accomplished by freezing out impurities

KOVAR SEAL

UNCLASSIFIED

DWG. Q-1072-IA

•COPPER FILLING TUBE

BRASS COLLAR

SILVER SOLDER

A„,„r,<>,£2£A h'mvww,,,,,,,,,,,,,^,,,,^,,,,^,^,,,,,,)

PLATINUMLINER \

15-mil NICKEL WIRE

4-mil TUNGSTEN WIRE-

= >*?;;;;/***

v«««««'"Wf«««/»;»/,,,,,,,,,,;,»,,,,,,,,,,,,;;;;,.„£„,;;;;;,,;,;;;;;„„„J

KOVAR SEAL

35-mil STAINLESS STEEL CAPILLARY TUBE

I BRASS COUNTER SHELL

Fig. 3. Fast Neutron Proportional Coincidence Counter.

I

INCHES

Fig.4. Triple-Coincidence Proton-Recoil Proportional Counter.

UNCLASSIFIED

DWG. Q-1265-1A

I

10

8

£ 7rr

& ^? 6

5

4

3

3oo

oLUN

<

oro

UNCLASSIFIED

DWG. 17710

•>

^/INTEGRAL BIAS CURVE

^^o***---^,

A THE PRESENT THREE-SECTIONTRIPLE-COINCIDENCE COUNTER

1 i i

o F RST EXPERT/1ENTAL COU NITER

7.5% \/A \ y

.DIFFERENTIAL OF BIAS (:urve

A-14.5%\

°"X16

0 10 20 30 40 50

PULSE HEIGHT SELECTOR

60 70 80 90

Fig.5. Integral and Differential Bias Curves-Alpha Curves.

i

ai

- 16

with liquid nitrogen before filling the counter. The use of krypton in

the counter lowered the background effects by about 50 percent. At least

a part of the background effects in argon-C02 is due to a known n-a

reaction in argon. Figure 6 shows the gas amplification versus counter

voltage for the three-section counter used in the present model of the

spectrometer. The counter is normally operated with a gas amplification

of about 20. The steepness of the gas amplification factor-voltage curve

requires that the counter voltage-be extremely well regulated.

Electronic Components

The high-voltage power supply found satisfactory for use in this

spectrometer has a half-wave rectifier with an LRC type filter system

(Fig. 7). Extremely good stability is obtained by first regulating the

high voltage with a string of 0A2-0b2 voltage regulator tubes, and then

feeding a string of voltage reference tubes with this regulated voltage.

Regulation of the order of 80 to 100 parts per million per day is obtained(5)

with this power supply. The prototype of this supply is NEPA Model E-109,

with such modification as coarse and fine controls, polarity reversing

switch, and voltmeter.

The output terminal of each section of the three-section proportional

counter is connected to an Atomic Instrument Company preamplifier which

has been slightly modified to reduce gamma-ray pileup. The cathode follower

(5) J. K. Bair and E. E. St. John, An Extremely Stable Power Supplyfor Scintillation Counter Use, NEPA-I265-SER-8 (Jan. 20, 1950).

O

en<

UJ>

_jUJor

100

-17-

200 300 400 500

COUNTER VOLTAGE (volts)

600

UNCLASSIFIED

DWG. 17 711

700 800

Fig.6. Counter Voltage vs. Gas Amplification for 3-Section Counter.

SIDPST

TOGGLESWITCH

115v 60

Fig. 7-Counter High-Voltage Power Supply.

UNCLASSIFIEDDWG. Q-1356-1A

S3 SWITCH3 POLE 3 POSITION CERAMIC WAFER

WESTON 741\ )o-100/j.a METER

WITH 0-1 kv SCALE

PC 11800

19

output drives a terminated 75 ft of RG-62/U cable to three ORNL A-l

amplifiers.' ' The band switch on the OML amplifiers is set on the

0.2^1-second position. The cathode follower output stage of the A-l

amplifiers was changed to improve the linearity.* It was also found

necessary to remove the differentiating network in the P.H.S. output

of the amplifier units associated with the front and middle sections

of the counter to simplify the design of the coincidence circuit.

The coincidence circuit designed for this spectrometer is relatively

simple because the counting rates involved are low, the input pulses have

a magnitude of 10 to 15 volts, and the dynamic impedance of the A-l

amplifier P.H.S. output is of the order of 100 ohms. The low impedance

of the input to the coincidence circuit is very desirable when using

crystal diodes, since this factor prevents large variations in the diode

characteristics from affecting the performance. The circuit consists of

three diodes with their anodes tied together through a common biasing

resistor to the B+ voltage (see Fig. 8). The coincident output is taken

from the anode junction and provides a signal to an amplifier stage, which

in turn drives a scaler.

(6) W. H. Jordan and P. R. Bell, Instructions for Use of A-l Amplifierand Preamplifier, MonP-323 (Feb. 15, 19^917

* A similar revision by Bell and Jordan *t. al, has been incorporated inthe latest model A-l amplifiers.

FRONTAMPLIFIER O

MIDDLEAMPLIFIERO

REAR

AMPLIFIERO

-20-

COINCIDENCE CIRCUIT

12 AT 7

ALL RESISTORS ONE HALF WATT

EXCEPT AS NOTED.

UNCLASSIFIED

DWG. 17709

B +

OUTPUT

U T ©

I MEG

MODIFICATIONS TO THE A-l AMPLIFIER PHS SECTION

V 15

6AK5

VI6

6AK5

Fig. 8

- 21 -

METHOD OF OPERATION

The spectrometer is put into operation in the following manner. The

radiator-absorber wheel is rotated to the position that places the

plutonium alpha source directly in front of the counter. (The alpha

source is spaced so that the plutonium alpha particles end their range

in the rear section of the counter, so that advantage can be taken of

the peak of the Bragg specific ionization curve). The P.H.S. dial is

then set to about mid-scale and the gain on the A-l amplifier is adjusted

until the scaler begins to register counts. Then an integral P.H.S. or

bias curve is obtained similar to those of Fig. 5-

The peak of the derivative of this counting rate curve establishes a

correspondence between pulse height and maximum stopping power of the

plutonium alphas in the counter gas. From the relation between stopping

power of protons and that of alpha particles the appropriate discriminator

pulse height selector settings are calculated for the desired window

widths. This same procedure is carried out for each of the three-counter

sections.

Various radiator-absorber wheels have been constructed to cover the

energy range from 1.3 Mev to 1^ Mev for the spectrometer. The 1-3 Mev

limit is set by the mica window thickness and the gas of the three-section

counter. Each radiator-absorber wheel has a plutonium alpha source

mounted in it for calibration purposes.

22 -

This spectrometer gives good stability considering the amount of

electronic equipment used and the rigid stability requirements. A

typical plot of the alpha peak counting rate versus time is shown in

Fig. 9. Figure 9 certainly indicates that the electronic equipment is

sufficiently stable for a 7- or 8-hr run. Since the data in Fig. 9

was taken without controlling the air temperature around the counter,

even better stability would be expected in a temperature controlled

room. During a spectral measurement, the calibration is usually checked

several times a day.

CALIBRATION DATA AND CALCULATIONS

The spectrometer has been checked for energy calibration and intensity

by measuring a monochromatic beam from the ORNL Van de Graaff generator.

Protons were used to bombard a tritium gas target. Neutron energies of

2 Mev, 3 Mev, and k Mev were measured with the spectrometer. Figure 10

shows a plot of the Van de Graaff data. It should be pointed out here that

data was not taken at energies much lower than the peaks owing to the small

amount of Van de Graaff machine time available. However, during a

previous Van de Graaff test of the spectrometer, considerable data was

taken at energies lower than the neutron peak, and it was found that the

tail of the curve went to zero very sharply. This previous data did not

have a sufficient number of points on the peaks to merit inclusion in this

report.

tz

oo

1100

1050

1000

950

900

850

-23-

UNCLASSIFIED

DWG. 17712

—o—

6 o <

c

o

)

>

o

O>

o

O

9

OUR DR

o

IFT LE 3S THA

' O

°

)

7"H M O /O

Ci—

<j?

°

0 4 5 6

TIME (hr)

8

Fig.9. Stability Test: Alpha-Source Counts From Front Sectionof The Counter, Set On Alpha-Peak.

10

- 2k

For example, to show the correlation between calculations and

Van de Graaff data, consider the 4-Mev peak: the point at the peak

was taken with wheel B, position 10. Wheel B, position 10, has

6.3O mg/cm2 polystyrene radiator and 21.94 mg/cm2 of aluminum absorber.

In these calculations everything is converted to equivalent range in

mg/cm2 polystyrene.

Converting the aluminum absorber to an equivalent amount of

polystyrene:

21.9% mg/cm2 (aluminum) xr^r =14.9 mg/cm2 polystyrene•

1.06 is the stopping power for polystyrene and I.56 is the stopping

power for aluminum. Now to obtain the total proton range,

14.9 mg/cm2 +2^30 ing/cm2 +2.84 mg/cm2 =20.9 mg/cm2.

The radiator thickness is divided by 2 because it is assumed that on

on the average the proton must penetrate half of the radiator

thickness. The 2.84 mg/cm2 is the equivalent range in the counter,

considering the mica window and the gas path.

From proton range energy curves'^ it is found that 20.9 mg/cm2

range in polystyrene corresponds to an energy of 4.0 Mev. (Range in

paraffin rather than in polystyrene is actually used - this con

tributes a very small error). Now the resolution can be calculated

from Geiger's Law:

(7) J. C. Hirschfelder and J. L. Magee, Efficiencies of NeutronCounters Using Recoils of Protons in Argon and Xenon, MDDC-115(July 23,T$55).

SsasBew-aawW'-?

- 25 -

dR = g dER 2 E

From the data for 4-Mev protons, it is seen that the energy resolution

becomes:

dE = 2 x 6.30 = O.I69, or 16.9$E 3-55 x 20.9

where

R = range in mg/cm2 polystyrene, E = energy in Mev, dR = radiator

thickness in mg/cm2, and g = 3-55 (this is obtained from the energy-range

relation).

The calculated resolution can be compared to the measured value

of 18.7$. The above simple calculation represents only a very approximate

approach neglecting several rather large errors.

A more precise calculation of resolution can be made by using a

method devised by B. R. Gossick^ ' to calculate the total range error

which is denoted by:

(8) B. R. Gossick, ORNL-I283 op. cit., p. 28.

c 25 P tot

where 6 p' 2 /RexS jf H,

Rr

6 p%, = 29.1 ft)t2l + t22

5 P t =12 (R)2

For 4 Mev

- 26

= 6 pc8:, + 5 p2a + 5 p2t

2 (9)"straggling error''

"angular straggling error"

'error due to radiator and detector width"

5p\ot =(2'5 x10~^ +^6.35 xlO"3) +(3.28 x10^) . 82.! xio^

Now (82.1 x10-lf) ("Y"J (2.36) =12.040 half-level width.Since this calculation is based on a perfect instrument, the:

discrepancy between 120 and 18.70 half-level width indicates the deviation

of this instrument from an ideal spectrometer.

Consider the intensity N(E) in Ffg. 10 (Van de Graaff data). This is

calculated from the expression:^10'

2 2

bl ^PN(E) = I „ fr

2n S a^-fiLS.

tj %2 E, x C neutrons/sec/Mev/unitn 1 solid angle

(9) H. A. Bethe, The Properties of Atomic Nuclei, II Range-Energy CurvesBNL-T-7. '

(10) B. R. Gossick, ORNL-I283 op. cit., p. 19.

27-

2.0 3.0 4.0 5.0

NEUTRON ENERGY (Mev)

UNCLASSIFIED

DWG. 16821

Fig. 10. Van De Graaft Data for Neutron Spectrometer, 9/26/52.

O

c

o

•o

o

c

3

>CD

O<D<n

\

(/)

c

oi_

-*—

0)c

28 -

where

C = counting rate in counts/sec,

R = range of the proton (in cm polystyrene),

g = Geiger's law constant,

•b-, = distance from source to radiator,

bp = distance from radiator to collimator,

S = number of hydrogen atoms/cc in radiator,

a = radius of collimator (the counter itself acts as a collimator),

t^ = radiator thickness (in cm polystyrene),

t2 = uncertainty in end of range (in cm polystyrene),

E^ = average energy of range.,

<T = hydrogen cross section at energy E.

The expression for N(e) is derived for an ideal point source but

should hold reasonably well for the experimental setup used in obtaining

the Van de Graaff data. Since the Van de Graaff generator was monitored

with a calibrated long counter, the total neutrons per second emitted

from the target can be calculated. For example, at 3-Mev the long

counter gave 2.62 x 10^ neutrons per second emitted from the target.

Now, by integrating the area under the 3-Mev peak, it is found that the

spectrometer data gives 2.42 x 10? neutrons per second emitted from the

target. A comparison of these results shows the spectrometer data to

be 9.250 low. Checking the other calibration energies at 2 Mev and 4

Mev gives equally good comparison.

(ll) Neutron Cross Sections, "A Compilation of the AEC Neutron CrossSection Advisory Group," AECU-2040 (May 15, 1952).

- 29 -

The statistical fluctuation at each data point is calculated by:

~2 _ ^N = Ql - 02

Ql + Q2

where Oh = total foreground counts plus background counts and Qg =

total background counts, both taken in the same time interval.

SPECTRAL MEASUREMENTS

Since the spectrometer gave good results when checked with the Van

de Graaff generator, it has been used to measure the more complex

spectra of a number of neutron sources. These spectral measurements

(12)compare favorably with measurements made at other laboratories.v '

Figure 11 is the spectrum of PB-236 - polonium-beryllium neutron source.

This source was of 10 curies strength at the time the measurement was

made.

Figure 12 is the spectrum of PB-256 - also polonium-beryllium neutron

source. This source was three times as strong as PB-236 at the time

of measurement. Much better counting statistics could have been ob

tained had time permitted. From the four polonium-beryllium sources

recently measured at this Laboratory and spectra measurements of polonium-

beryllium made at other laboratories'^) it appears conclusive that the

polonium-beryllium spectrum varies considerably from source to source.

(12) J. H. Perlman, H. T. Richards, and Lyda Speck, The NeutronSpectra of Po-B and Po-Be. MDDC-39 (July 9, 1955).

(13) B. G. Whitmore and W. B. Baker, "Energy Spectrum df Neutronsfrom a Po-Be Source?, Phys. Rev., 78V 79$ (June I95O)."""

700

600

500

e

•5 300

Uj

^

200

100

UNCLASSIFIED

DWG. 17720

1 /"^ V —i

1 0 ) ll- ^3 )^—

^•V

1\H

1 —y5 V-" H

1—( )—11- (t )

j—L_ —g)— —1

1 ( V ' 1

•)—1

/1

1 1 i— —<r'— 1

1

11

11

11

1r

3 4 5 6 7 8 9 10 11 12

Mev

Fig.11. Po-Be P. B. 236 Spectrum.

I

oI

1.7

1.6

1.5

1.4

1.3

1.2

1 .1

1.0

~ 0.9

o

o 0.8

c

p 0.7ki

2«0.6

0.5

0.4

0.3

0.2

0.1

31-

UNCLASSIFIED

DWG. 17719

!• •"•<•!

•-

c i

-ft) |

H-V >^—

1 rtiV i -^

®lT i)

1 ••

i :i

••

1 —I

2 3 4 5 6 7 8 9 10 11 12

NEUTRON ENERGY (Mev)

Fig. 12. Polonium —Beryllium Source. P.B.-256.

32 -

Thus, one cannot expect the average energy of one polonium-beryllium

source to be the same as that of another.

Figure 13 is the neutron spectrum of PB-235, polonium-boron source

used by the Health Physics Division for calibration purposes. The

spectrum drops to zero above 5 Mev as would be expected from the

Q-values of B10 (a>n) and B11 (a,n).

Figure 14 gives the neutron spectrum of PB-281, a very intense

polonium-boron source recently acquired by the Biology Division.

(l4) T. W. Bonner and L. M. Mott-Smith, "The Energy Spectra of theNeutrons from Disintegration of Fluorine, Boron, and Berylliumby Alpha Particles", Phys. Rev., 46, 258 (1934).

CD

O<U

1400

1300

1200

1100

1000

900

800

E 700o

</>

I 600•t-

0)

X 500ki

400

300

200

100

0

-33-

UNCLASSIFIED

DWG. 14678

t f 7"w

Po- B SOLJRCE NO. 2 35

1 V. i

|—(^—H

•M

F= —1

ST •— — 1 T>)^^

\y——

1— "t ^^"r^i

0 1/ I 5 *\ £> 15 7 £J $3 10 11 12

NEUTRON ENERGY (Mev)

Fig. 13

200

180

160

X

to

O

x 140

a>

o>co

1 <20to

c3

% 100

o

CDCO

CO

c

ok_

-»-

c

80

!S 60

40

20

-34

UNCLASSIFIED

DWG. 16820

£M \

1 c 1

li

J ®\A*J 1

(?)

r®)

©

f j©r

(3f

r i

r»i

(?y

(•1

\?)

©

2 3 4 5

NEUTRON ENERGY (Mev)

Fig. 14. Po-B Spectrum. Pb 281 10/6/52

- 35 -

CONCLUSION

The neutron spectrometer has been found satisfactory thus far in

fulfilling its design requirements. It mainly utilizes electronic

units commercially available with a minimum of modifications and is

reasonably stable considering its electronic complexity.

While considerable effort was made to obtain the maximum sensitivity

by optimizing the available parameters, more sensitivity would be

desirable. However, since the sensitivity is dependent on the number

of hydrogen atoms in the radiator and the hydrogen cross section, a

natural limit is established.

- 36

ACKNOWLEDGEMENTS

The above instrument is the result of continuing the research

initiated by Mr. B. R. Gossick and Mr. K. M. Henry in 1948.

The authors gratefully acknowledge the valuable assistance and

counsel given them by Mr. B. R. Gossick of Purdue University.

We wish to thank Mr. R. E. Zedler of the ORNL Instrument Department

for assistance in designing and constructing counters for the

spectrometer.

Thanks are due Dr. J. L. Meem for his encouragement and support.

We are also indebted to Dr. J. L. Fowler and Dr. Harvey Willard for

Van de Graaff machine time.