Text

Evaluation of Techniques for Dynamic Measurement of Fuel Motion in LIQUID-METAL-COOLED FAST-BREEDER Reactor Safety Experiments
	ML20002E023
Person / Time
Site:	Clinch River
Issue date:	10/31/1980
From:	Evans A, Orndoff J; LOS ALAMOS NATIONAL LABORATORY
To:	; NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
References
	CON-FIN-A-7046 LA-8301, LA-8301-MS, NUREG-CR-1398, NUDOCS 8101260127
	Download: ML20002E023 (67)
	v • d • e

.

TEC/L NUREG/CR-1398 LA-8301

}

55 s-Vi l

=-

l v

a M

M Evaluation of Teshniqtles fdr Dynamic Measurement of Fuel Motion in Liquid-Metal-Cooled Fast-Breeder Reactor Safety Experiments i

CO C

d u

O l

2 I

E i

O

'+-

O I

'5 u

l'

.2 i

C LOS ALAMOS SCIENTIFIC LABORATORY j

Post Of fice Box 1663 Los Atamos. New Mexico 87545 P10 126 0 l'F14 4

l l

l An Affirmative Action /EqualOpportunity Employer 1

l l

Edited by Nancy Shera Photocomposition by Cleo H. Gutierrez I #

M Tens eevoet een prepesed as en ocemens of wert spensesed by se eseecy of the Unsed 5 eees Goveremset. Noaher the M 5 esses Geseresses mes any assory thmeef es say of shee empl8yees, auske any warraney,empreemd er esophed.

er steelse any legal kr ekty er..

" ~ for say thed party) est, es the geseht af such see, of any amfermessee, e

y epfereses. Freduct er peeces W en ehes separt.or sepsenets that es ses by sect thed party oceld ese tefrege M eweed #gets

NUREG/CR-1398 LA-8301 R7 Evaluation of Techniques for Dynamic Measurement of Fuel Motion in Liquid-Metal-Cooled Fast-Breeder Reactor Safety Experiments Albert E. Evans John D. Orndoff*

Manuscript submitted: March 1980 Date published: October 1980

Consultant. 997 B 48th St., Los Alamos, NM 87544 l

Prepared for Division of Reactor Safety Research Office of Nuclear Regulatory Research US Nuclear Regulatory Commission I

NRC FIN No. A7046 l

l l

tJNITED STATES DEPARTM ENT OF ENERGY CONTR ACT W-7409-ENG. 36

FOREWORD l

This report covers work performed in fiscal years 1976-79 under Projects R-283 and R-414," Review of Safety Test Facilities for Fast Reactors." The work constitutes a portion of Task 3 of this project," Experiment Diagnostic Systems Evaluation." Work commenced in October 1975 under the direction of John D. Orndoff; upon his retirement in June 1977, the work was continued by Albert E. Evans.

In the spring of 1978, after a decision had been made to defer or cancel plans to build new safety test facilities for liquid-metal-cooled fast-breeder reactors (LMFBRs), the Nuclear Regulatory Commission was directed to cease suppcrt of research aimed toward providing design data for such facilities. Accordingly, the program elTort was redirected to support development of the upgraded Transient Reactor Test Facility of Idaho National Engineering Laboratory and to provide assistance in the development of coded-aperture fuel-motion monitoring techniques for use at the Annular-Core Re-search Reactor of Sandia Laboratories, Albuquerque. The redirected program was supported under Program R-414 for the purpose of determining whether presently planned fuel-motion diagnostics techniques for scheduled destructive tests of LMFBR fuel assemblies would produce data adequate for verification of the SIMMER code.

Experimental work terminated on September 21,1979.

f, iv

CONTENTS FOREWORD.............................................iv ABSTRACT I

1.

INTRODUCTION

'II. TIIE FACILITY 2

A.

PARKA 2

B.. Provisions for Ex-Core Imaging 6

C. Test Assemblies....................................10-III. INSTRUhlENTATION

...................12 A.

Neutron Detectors

.................12

1. llornyak Buttons

.. 12

2. *lle-Recoil Detectors

. 13

3. Stilbene Detectors

................14 1'

B.

Gamma-Ray Detectors

....... 16

1. Nal(FI) and NE-102

............. 17

2. Bismuth Germanate

........ 18 IV. IIODOSCOPE SCANS

.....24 A. Single-Pin Scans

.................24 B.

37-Pin Assembly Scans

. 31 C. 91-Pin Assembly Scans

.......40 D.

127-Pin Assembly Scans

......42 E.

EITect of Test Capsule Wall Thickness

...............44 F.

Studies of Clad-hlotion Detection

. 47 G.

EITect of Energy Selection on Gamma-Ray Imaging of Fuel

. 50 II. Summary

.... 50 V. CODED-APERTURE IhlAGING STUDIES

... 51 1

VI. IN-CORE DETECTOR STUDIES...................

53 VII. TRANSIENT OPERATION OF PARKA

...........55 ACKNOWLEDGhlENTS

.58 REFERENCES 58 APPENDIX. PUBLICATIONS AND PRESENTATIONS OF WORK COVERED BY Tills REPORT...

........59 v

..~

1 TABLES

. I. Reactivity Worths of Various Objects in PARKA...................... 7 II. Results of Test for Intermixing of Detector Counts Induced by Gamma Rays and Neutrons...................... 34 Ill. EfTect of a 21-mm-Thick Steel Casing on Hodoscope Images of a 37-Pin Assembly... 45 lV. Observed and Calculated 7.6-McV Gamma Rays from a 37-Pin Assembly and Steel Jacket in PARKA...................50 V. Fractional Decrease in Hodoscope Image Intensity Caused by Center Void in Various-Sized Test Assemblies.....................51 VI. - In-Core Detector Response to Fuel-Pin Removal

...........54 Vll. In-Core Detector Response to Removal of Complete Fuel-Pin Rings

..............55 Vill. PARKA Yields from Godiva Bursts

..........56 vi

EVALUATION OF TECHNIQUES FOR DYNAMIC MEASUREMENT OF FUEL MOTION IN LIQUID-METAL-COOLED FAST-BREEDER REACTOR SAFETY EXPERIMENTS by Albert E. Evans and John D. Orndofi ABSTRACT He purpose of the Los Alamos Scientific Laboratory program for evaluation of fuel-motion diagnostics instrumentation is to determine whether fuel-motion monitoring techniques for planned liquid-metal-cooled fast-breeder reactor safety tests can yield information adequate for verification of reactor-safety calculations. Neutron and gamma-ray hodoscope scans of fuel assemblics containing from 1 to 127 pins were obtained. It was found that a single-pin void can be detected in fuel assemblies containing up to 127 pins and that, for both 37-and 127-pin fuel assemblies, the image of a single-pin void varies with the depth of the void within the assembly. The effect of a thick steel casing on image quality and the use of in-core detectors as fuel-motion monitors were investigated. It was shown that PARKA is a suitable driver reactor for static testing of coded-aperture fuel-motion monitoring systems.The use of PARKA in a pulsed mode to study transient phenomena in safety tests is discussed.

I. INTRODUCTION destructive tests could be performed on a full sub-assembly of LMFBR fuel (217 or 271 pins) or on more Complete evaluation of the safety of lig-than one subassembly.

uid-metal-cooled fast breeder reactors (LMFBRs) re-The ability to monitor fuel motion in a disintegrating quires facilities where quantities of LMFBR fuel pins, test assembly is vital to the understanding of the test ranging from a single pin to multiple subassemblies results and to the verification of reactor-safety calcu-containing thousands of pins, can be tested to destruc-lations with these results. Criteria have been established' tion under conditions simulating those that may exist for the performance of fuel-and clad-motion measure-during a reactor accident. Such tests have already been ment systems. These criteria are based upon the sensi-performed at the Transient Reactor Test Facility tivity, accuracy, and time and spatial resolution neces-(TREAT) of Argonne National Laboratory, located at sary to derive useful information from tests intended to Idaho National Engineering Laboratory (INEL), on simulate various types of core-disruptive accidents in assemblies containing from one to seven pins. This LMFBRs, and include specifications for field of view, facility is being upgraded to permit tests on assemblies of spatial resolution, time duration and resolution, and 19,37, or 61 pins. Complementary work is also being mass resolution. In general, the requirements vary with done at the Engineering Test Reactor ofINEL and at the the size of the assembly undergoing test. For instance, Annular-Core Research Reactor (ACRR) of Sandia horizontal spatial resolution requirements vary from 2 Laboratories. Albuquerque (SLA). In addition, require-mm for tests of a few pins to 50 mm for tests of multiple ments have been established for larger facilities' 2 where subassemblies that may contain over 1000 pins.. The 1

fuel-motion measurement system will also be required to II. THE FACILITY have a capability for depth measurement comparable to its horizontal resolution and to detect the motion of as A. PARKA little as 0.04 g of fuel in small-bundle tests or of 50 g in multiple-subassembly tests. Time-resolution require-The PARKA critical assembly at Los Alamos Scien-ments vary from 0.2 to 100 ms, depending on the type of tific Laboratory (LASL) has been modified to serve as a test.

driver for arrays of from 1 to 127 highly enriched UO2 Since the test assemblies are immersed in liquid fuel pins representing LMFBR fuel. PARKA is a sodium, one is constrained to monitor fuel motion with R o ve r-Projec t Kiwi reactor loaded with the image formed by neutrons or gamma rays emitted by graphite-urenium fuel elements. These are one-hole the assembly or by x-rays transmitted through the bead-loaded hexagonal elements that were used for assembly. The quality of this image is strongly influenced critical-assembly studies for Rover reactor design. En-235 by scattering and absorption within the test assembly of riched-uranium (93%

U) loadings of 100, 200, 300.

radiation emitted not only by the test assembly but also and 400 kg/m' are distributed to give approximately flat by the reactor that drives the test assembly to destruc-fission density across the fueled diameter. Figure 1 is a tion. The time resolution (the minimum time during photograph of the PARKA core with a 37 fuel-pin test which information must be gathered to obtain an image assembly at the center. The active core, 0.91 m in with the required spatial and mass resolution)is limited diameter by 1.32 m high,is surrounded by a 15-cm-thick ultimately by the statistics of radiation emission from the beryllium reflector containing 12 rotary beryllium con-test assembly. In practice, however, time resolution is trol drums faced on half their circumferences with B C.

limited either by the data-acquisition system in the case Rotation of the control drums is controlled to the nearest of radiation-counting system or by mechanical or elec-0.01 by stepping-motor-driven actuators. Control tronic considerations in the case of photo-imaging sys-drums may be operated singly or in unison. A differential tems.

power-level sensor operates one control drum to main-Techniques being applied to or being studied as tain a power level constant to within 0.1% during runs candidates for fuel-and clad-motion monitoring include lasting several hours.

flash-x-ray cinematography,* single-and multi-The control system has a reactivity worth in excess of ple-pinhole self radbgraphy,' radiography with Fresnel 9 S. A shutdown reactivity of-5 $ is required to assure and other coded apertures,' use of in-core radiation safe facility operation. If an assembly is to be tested that detector networks,' and use of multi-aperture collimating will cause the reactor to have.a shutdown reactivity hodoscopes.' Acoustical holography has also been con-greater than -5 $, the reactor is shimmed by the removal sidered as a means to monitor fuel motion.' The selection of 6.5-mm-diam graphite rods from some of the of a fuel-motion monitoring system must be done before 10 mm-diam holes in the centers of the fuel elements.

design of the test facility because the system selected will Conversely, more graphite rods may be added to the strongly influence facility requirements.

core if the test assembly ha: negative reactivity.

The purpose of the work described herein was to ONETRAN

transport-code calculations have been provide a facility where the applicability of the various run for a 37-pin test assembly in the center of PARKA.

proposed fuel-motion monitoring techniques could be The 16-group neutron spectra at the center and at the determined and where instrumentation could be tested edge of the test assembly are shown in Fig. 2. Figure 3 and optimized before installation in full-scale test facil-shows the fission densities at these two positions as ities. The facility has been applied to the study of functions of the energy of the neutron causing fission.

hodoscopes, coded-aperture imaging systems, and The median energy of neutrons causing fission at the in-core detectors. Some guidance is provided for the edge of the test assembly is between 0.55 and 3 kev. At future application of these techniques.

the center this shifts to between 3 and 17 kev. These 2

pp-Figure 4 shows the results of a ONETRAN calcu-

~'

lation of the fission density for a 37-pin test assembly in PARKA with water, sodium, or void between the pins.

I Note that water as a filler improves the FOh! by a factor se,,s -

-..- h.... a f 2. an advantageous situation when only gamma-ray 2

P information is desired. For fast-neutron self-imaging of a EY I

test assembly, the presence of water would probably be s

e

- ] # 7 ~ '.;

intolerable. For single-pin experiments, it is possible to yg /.

A g

g increase the FOh! to as high as 80 with a polyethylene q

f, j

-?

f,1 flux trap just outside the test region. It should also be f,'.[

'q'- g [, h and void as fillers. Aluminum grid blocks have been used 0..$

y noted that there is very little difference between sodium p,

i~-

- 7 j. g -.../F.

as a substitute for sodium in most of our tests.

+-%?j g.'

Figure 5 shows the radial fission distributions for 37-

.m gg

..r yy.

to 127-pin hexagonal fuel assemblies in the test region of e

/

PARKA. These measurements were made by irradiating fresh fael pins in selected positions as the test assembly was built up from minimum to maximum size. The

.g selected fuel pins were removed and the gross fis-b' sion-product gamma activity of each pin was compared

'l,f to the activity of the central pin, which was used as a reference. The results show that variation in power I

density across a test assembly, although correctable in most cases, will be a factor in interpreting fuel-motion Fig.1.

diagnostic data. Gra&d fuel loading or graded fuel Photograph of the top of the PARKA core with a enrichment could be used where a more uniform power 37-pin LMFBR test assembly at the center. The density is required in the test assembly. The data shown fuel elements in the light-colored band leading n Fig. 3 suggest that filtering oflow-energy neutrons by from the center to the lower right of the core have inserting a cadmium or boron sleeve around the test been cut to form a riewing slot. The penetration section could improve fission-density uniformity within through the beryllium repectorfor this viewing slot this region.

can be seen at the lower right of thefgure.

Consideration has been given to how large a test bundle could be accommodated in PARKA. Fis-figures show that PARKA is an intermediate-spectrum sion-density calculations for four subassemblies in reactor with neutronic characteristics typical of those PARKA approximate proposed Safety Test Facility required of a driver reactor for large-bundle LN1FBR (STF) test geometries including appropriate sodium i

safety tests in which a test assembly must be reasonably content and stainless steel containers. For this 868-pin uniformly irradiated throughout its volume.

array, the fuel was assumed to be highly enriched in addition to the requirement for uniform irradiation UO.The large self-multiplication of the four sub-2 is the requirement that the ratio, referred to as figure of assemblies increases the PARKA system reactivity by l

merit (FOht), of minimum fission density in the test 23 $. The test assembly depresses fissions in the adjacent specimen to maximum fission density in the driver be PARKA fuel by approximately 25%. The FOh! is large, so that the test specimen can be driven to approximately iI, with a peak-to-minimum ratio across I

destruction without harming the driver reactor. / high the test assembly of 1.33. The 23 $ of excess reactivity in FON1 also improves the signal-to-background (S/BJ ratio PARKA can be reduced easily either by enlarging the in fuel-motion diagnostics measurements. A high FOh!is hole in the hexagonal fuel elements to increase core void obtained by fueling the test object with fully enriched fraction or by adding boron-loaded graphite rods near uranium,either as pure metal or oxide, and maintaining a the central hole. Boron should harden the PARKA 3

relatively low fuel density (400 kg/m near the center of spectrum to give both a smaller peak-to-minimum vari-PARKA)in the driver.

ation across the test assembly c.nd an increased FONI.

3

NEUTRON GROUP NUMBER

.g s

.., IS 64, 13 86

,.' 12 41 10, 9 8

7 6

S

.3

,4 3

I

'l j' f

--- EDGE CF TEST BUNDLE m

l

- CENTER OF TEST BuhDLE to'3

s m

~

3 io 4

3

?

x so* 5 m

.g.

$C*

W F

. in

-a g

1.f

...t

...I

_f

..I

[

C'2 O* 8 4

0 502 gg3 104 30 5 06 of NEUTRON ENERGY (eV)

Fig. 2.

Sixteen-group ONETRAN calculation of the neutron spectrum at the center and at the edge ofa 37-pin test assembly in PARKA.

NEUTRGN GROUP NUMBER 16 IS 14 13 42 si 50 9

8 F

6 5

4 3 2 i

IO'I

..,.ij

... _j

..g

.,)

1 I

I l

..ECAiE OF TEST BUNDLE l

- CENTER OF TEST EluNDLE

'M

_ so' *

-~(

r y

c d

~

$ 10' 3.r--~..

sib O

2 io r u

so"7

-8

..f

..t

.f

.t

..]

g g-2 c1 g

e c2 c3 c4 os g6 IOP NEUTRON ENERGY (eV)

Fig. 3.

Calculatedfission density, as afunction ofthe energy ofthe neutron causingfission, at the center and at the edge of a 37-pin test assembly in PARKA.

4 3

" }'

m o-o 66 6

Ju.

3-

10 0 g g

g g

g

=

= FUEL I

PINS '

Ho a

Void c lO =

8 g

e-DRIVER (Centrol Region) 5 Ho-Fig. 4.

e E,

N* -

Calculatedfission density in the test and driver

~

yg regionsfor a 37-pin test assembly in PARKA with rariousfillers.

I I

o 20 40 so 80 10 0 12 0 Radius (mm) l I

I I

i i

l i

p 9

/

--127 Pins, Across corners f

l 1.30

--127 Pins, Across flots i

f a-- 91 Pins, Across corners

/

A-91 Pins, Across flots

/

l.25 0--61 Pins, Across corners 9

p

/

l e-61 Pins, Across flots

/

O--37 Pins,96 mm tesi hole j

f

/

- l 20 m-37 Pins,76 mm test hole

/

p f

N I

/

j

/ [l l

,, 1.15

/

b

/

l/

l

/

e

/

w

/

/ /

, 1.10

/

'/

b

/

" 1.05

/

y 1.00 i

l l

I I

i~

O 5

10 15 20 25 30 35 40 45 50 Rodius (mm)

Fig. 5.

Measuredfission distributions in various-sized test assemblies in PARKA.

5

100 ;,

, =

Z Fuel Pins 50-2 20

Without 8

$ 10 With 8 Q

=

2 is E E

E Driver 2-

.oy l =.

---~~

Fig. g.

~

=

0.5 :

Caltulatedpssion density in the test and drim regionsfor afour-subassembly array in PARKA, without baron and with boron umformly dis-02 tributed over the driver region.

0.1 I

I l

I I

l i

O 10 20 30 40 50 Rodius (cm)

This was confirmed by calculations with boron uniform-Figure 7 shows an overall view of PARKA. It has ly distributed over the driver portion of the PARKA been necessary to add 12 mm oflead to the sides and 50 core. The boron required to reduce the reactivity the mm to the top of the reactor to permit access to the required 23 $ is equivalent to 6.3 g per PARKA fuel reactor after irradiation.

element. The calculated fission densities with and without boron are shown in Fig. 6. Although the boron poison causes a large perturbation in the driver fission prof le, B. Provisions for Ex-Core Imaging the resulting FOM is approximately 20 and the peak-to-minimum fission ratio in the fuel pins is only A 4.4-cm-wide by 10.2-cm-high slot has been cut 1.10. Maintaining a fiat fission-density profile across the transversely through the fuel and the beryllium reflector driver in PARKA is of no particular concern. It thus of PARK A to permit viewing the test region with either a appears that PARKA can be adapted to serve as an hodoscope or a coded-aperture imaging system. Another excellent driver for a four-subassembly array of fuel pins.

4.5-cm-wide by 5-cm-high slot, located in the reflector 30 Initial measurements on the 37-pin assembly in cm below the primary slot, is available for future PARKA show that addition of the UO fuel to the experiments.

2 stainless-steel cladding tubes increases reactivity by The reactor is a dj a e e n t to an un-1.1 S. The UO, length of 0.91 m is centered in the mortared-concrete-block shielded instrument room. The 1.32-m-long PARKA driver fuel. ONETRAN calcu-viewing slot in PARKA is aligned with a stepped hole lations of this geometry predicted a reactivity for the 37 leading into the shielded area. This hole,30 cm square at pins of 0.9 $ in a voided test cell and 1.1 $ in a cell filled the reactor side and approximately 75 cm square at the with sodium.

inside face of the instrument room, can accommodate a Reactivity worths of test bundles of various sizes, the variety of imaging devices. The overall facility layout is PARKA central region fuel elements, shim rods, and shown in Fig. 8. The combination of 75-cm-thick other additions to the reactor that have afTected system shield-room walls, the 45-cm-thick concrete wall of the reactivity are listed in Table I.

critical-assembly bay, and 38 cm of unmortared concrete 6

TABLEI REACTIVITY WORTHS OF VARIOUS OBJECTS IN PARKA Reactivity item (S)

Central region PARKA fuel element,400 kg/m' 93% enriched uranium 0.0610 Graphite shim rod,7-mm diam 0.0060 Empty axial steel shell,96 mm i.d. by 7.1 mm thick 0.280 Single simulated FFTF fuel pin,172 g 93% enriched UO in steel shell 0.0273 2

Assembly of 37 fuel pins in steel shell 1.300 Assembly of 61 fuel pins in steel shell 2.140 Assembly of 91 fuel pins in steel shell 3.300 Assembly of 127 fuel pins in steel shell 4.980 Lead shield on top of core,3 cm thick 0.800 Lead shield outside Be reflector,1.25 cm thick 0.800 91 pin EBR Il assembly,30 cm long,5 kg $2% enriched uranium 0.690 Slot halfway through reflector and core,102 mm by 44 mm 0.580

. ~

nca+-

l 4

Fig. 7.

F7ew of the PARKA facility set upfor evaluation of f :~,"

fuel-motion diagnostics instrumentation. The as-

'f sembly at the top of the core is an actuatorfor remote rotation of test assemblies and withdrawal of individualfuel pins. The rotary control-drum actuators can he seen on the underside of the reactor.

A t

D F

]D

'D'T A

s e J\\\\ 'JLA1

=

l t

O

- Critical ossembly boy Working platform Test region,9.5 cm 4 Core.OSI m 4 x 1.32 m high

. --Viewing slot, 7.5 cm 4 4.4 cm widex 10.2 cm high thru hole 30 cm square hole CH,38 cm square j y,g 7

a l Concrete wall ff fPb,78 cm square

' iiii!l 7 ' r>

i Auxiliary

'h'ir'i

!l l i

equipment 6

room e

25C,NHodoscope Detectors

__-~

colkmotor, y

15 cm square Hodoscope

'-~

9 scan drive l

1 i

i i !

i f

{

Corridor T

<! l ; ; ; ; l Concrete block g

Layout of thefuel-motion diagnostics instrumentation evaluationfacility.

l I

l 9

im 2,m

/

Exterior block stacked between the reactor working platform and The field of view of a collimator slot is considered in the assembly-bay wall provides an attenuation factor in Fig. II. lf radiation is detected uniformly across the face the shield room for PARKA neutrons and gamma rays of a sensor at the collimator exit (a) and if collimation is of approximately 10 000. The shield room has been lined

" perfect," the detector senses radiation uniformly from a with 25 mm of boron-loaded plaster (Reactor Experi-field at the object (b) of the same dimensions as the slot, ments, Inc. #284) to suppress thermal notrons.

plus a fringing area defined by diagonals across the slot As shown in Fig. 8, a four-channel stee: hodoscope from its entrance (c) to the exit. If O is the angle between collimator was installed in the large hole in the these diagonals and the walls of the slot, d the width of shield-room wall to view the central portion of the test the slot and Iits length, we have region in PARKA. This collimator,15 cm square and 1.83 m long,is a lamination of three 45-mm-thick steel tan 0 = d/t (1)

. slabs into which the viewing slots have been milled. A photograph of the collimator is shown in Fig. 9.

If r is the distance of the object plane from the front face Details of the collimator and its relationship to the test of the collimator, then the total width y of the viewing region are given in Fig.10. The slots through the area is given by collimator are 4.8 mm wide by 12.7 mm high. The front face of the collimator is 0.91 m from the center of the y - d + 2r tan e - d[1 + (2r/t))

(2) i reactor test region.

l 8

i t

i

r:-

A_'

f_:

.e Fig. 9.

Installation of the four-channel hodoscope col-fl limator.

D s-r F

(.

.z

'=

183.2 91.6

=

I i

15.2 --4----________________

y

____________.c___

20.3 m Drive point 30.4 PARKA Swivel point Steel Hodoscope Outside of Collimator Be reflector Front Back j

Slits.

-O O O

F O.48 x 1.27 4.42+

+-

4.76

- 30 0

0 j

i l

10.2

~4.44 I

i g,9 -.

i Si t thru PARKA Collimator Slits core end reflector Dimensions in cm Fig. 10.

Details of the hodoscope collimator and test region geometry.

i F-w 0

2 COLLIM ATOR O

~ ra y w arecer m s w w w rt e,

l7

+

Fig. 11.

~

d *

'~

Field of view of a collimator slot.

I I

(a)

(c)

(b) 9 D

]D D h \\[ hA maM o f.1 AIfb

The horizontal resolution y' of the collimator slot is Fuel pins are assembled into test bundles with taken to be the width at which the sensitivity of the aluminum grid plates drilled to locate pins in a hexagonal detector mounted at (a) falls to half of its maximum, pattern with 7.26 mm between centers. In the region of Therefore, the viewing slot, the grid plates are stacked to provide a solid aluminum matrix for the test bundle as a crude nuclear substitute for sodium. Measurements described y' = dil + (r/t)1 (3) below showed that the presence of the aluminum did not materially atTect fuel-motion monitoring.

In a similar fashion, one derives an expression for the A test well was formed on the axis of the PARKA vertical resolution. From the dimensions given in Fig.10, core by removing 37 of the hexagonal fuel elements and a horizontal resolution of 7.2 mm and a vertical resolu-inserting a 110.7-mm-o.d. by 103.25-mm-i.d. steel liner.

tion of 19.05 mm are determined.

An inner lining 96.5 mm i.d. by 101.6 mm o.d. is mounted to rotate within the outer liner so that the test assembly can be oriented at any angle with respect to the C. Test Assemblies viewing slot. Remote orientation is provided by means of a stepping-motor and rim-gear drive at the upper end. A Simulated fuel pins of Fast-Flux Test Facility (FFTF) drawing of a test assembly in the test well is shown in dimensions were fabricated by loading 4.8-mm-diam by Fig. 13.

50.4-mm-long extruded pellets of 93.09%-enriched UO To facilitate the handling of test assemblies after 2

into stainless steel tubing. The tubing, rejected FFTF irradiation, a shielded cart has been built. This cart fuel-pin casing obtained from Battelle Northwest Labo-provides the lower body of a person working on test ratories, has an inside diameter of 5.08 mm and a wall assemblies the protection of 50 cm oflead shielding. The thickness of 0.38 mm. Eighteen pellets, each containing cart is shown in use in Fig.14.

7.83 g of nU, were loaded into the central 91.4-cm Test assemblies in TREAT and in future safety test 2

region of the 137-cm-long fuel pins. The ends of the pins facilities must be encapsulated in steel as thick as 45 mm.

were filled with 25-cm-long sections of 4.75-mm-diam This much steel is bound to seriously affect the ability to drill rod. The bottom ends of the pins are rounded to monitor fuel motion in the test assembly. Accordingly,it assure easy insertion into grid plates. A threaded fixture was necessary to measure the effect of steel on ex-core is provided on top to allow attachment to a pin-motion imaging techniques. For these tests, a 37-pin test as-actuator. Details of a test fuel pin are given in Fig.12.

sembly was surrounded in the viewing region by a FUEL PE LLE T stb!

F UE L C AslNG 0 m aam a 2 00 k"'s ST AINLE ss sTE E L HEDL SUPPLIED CoVP R E ssION END C AP O 200 i d = 0 230 o d SP R IN G p pg g y s T AIN L E ss ST E E L SPHE RIC AL R ADIUS SE AL PLUG DRILL ROD HOLL g

0187 d.am 018 7 44m

/

j

\\.

I e n NC

\\

( W ll!

//

W'l AhRil]

Fig. 12.

l l

' 6 0 0m Demils of afuelpin.

8 125

- - - 36 00 %

sT AINL E ss l

l STEEL NON ROT ATABL E

% 00 DWE NsioNs IN INCHFs 10

1 d h '

.N 0

Wo.h

/uo:':t O o o&'l:)

I X:

o^

n 4,o p

o V' o

' c' :I o

' c:

Xc o c; 1 i 127 HOLES HEX b'c,,,

o J-'/

PATTERN

\\ 'Q"'%,,gjg,,

BETWEEN CENTERS 6.09 mm diom., 726 mm N

Fig.13.

Details of afuel-pin assembly in test llllllll57-PIN FUEL ARRAY well.

CAPSULE WALL

/ SIMULATOR LIFTING A

RING rmm mm 41 TEST CAPSULE Q="

WALL SIMULATOR TIE ROD (3)

L'FTING ROD (3)

ORIENTATION PIN d

N TOP SUPPORT PLATE 8:

MY

M (STAINLESS) 15 cm & BRASS GEAR TEFLON BUSHING 37-PIN GRID PLATE (12)

! / 50 mm & x 25 mm THICK p:F=

7 I ;:pSTEEL TEST CAPSULE

$3 WALL SIMULATOR t

524 s d. x 95.O o d x 203 g

I LONG n

n g

N

.. ~ _

. %~~~-'~["~

' [127-PIN GRID PLATE 95 mm 4 x 25

~

SUPPORT ALUMINUM ROD (3)sN SNAP RINGS co

-n--r s

'[127-PIN GRID PLATE

( ALUMINUM )

o i

4 a

l j

,plNNER TEST CELL LINER J

lPOUTER TEST CELL LINER g

s h?

Yhlh l

! ra---965 N ROLLER BEARING

~10325

10.7 DIMENSIONS IN mm 11

i

.p -.,

3 n

g p

b j

F y

[" % 3

(

l

. {y;3C a

1 j

)

f pf j

1 U

L JJ -

M&W 9 l

, N R T^#

}

y

\\

g, h xl.

~

n I jr

~

t-4 Fig. 14.

y VQ Shielded test assembly cart.

~-s-c

)

$\\

cylindrical casing 52.4 mm i.d. by 95 mm o.d. by 203 l

mm long. This casing was remotely raised from and Fis 15-lowered into the field of view by means of steel cables Actuator turretfor raising and loneringfuelpins attached to a stepping-motor-driven drum.

andfor rotating the test assembly.

)

Up to six pins in a test assembly may be raised from or lowered into the field of view by means of nylon stilbene scintillation detectors have been used. The ideal strings attached to stepping-motor-driven spools. The neutron detector for this purpose is one that is as lifting assemblies for the fuel pins and the steel test insensitive as possible to gamma radiation and that can capsule are mounted on a turret that is attached to the be biased for neutrons with energies of from I to 2 MeV.

inner liner of the test well. This turret is shown in Fig.15.

To minimize backgrounds, the detector should have a Stepping motors controlling motion of the fuel pins sensitive area closely matching the slot size. Hornyak and the steel test capsule, rotation of the test well, and Buttons and 'He-recoil detectors have been found to panning of the hodoscope scanner are all driven by a meet the criteria of gamma insensitivity and energy common stepping-motor controller through a system of discrimination. However, stilbene detectors are more selecter switches and relays that permit only one motor eflicient by a factor of 5 than Hornyak Buttons of the to be operated at a time. Reproducibility of settings is same size and offer the additional advantage of obtained by counting pulses delivered to the stepping pulse-shape discrimination." by which it is possible to motors. The controller has provisions for presetting the obtain neutron and gamma-ray data simultaneously, amount of motion desired. A position readout in the controller is coupled to an automatic data-recording system.

1. Hornyak Buttons. Initial data were taken with Hornyak Buttons that provided a direct comparison with data obtained at TREAT. These detectors are pieces of III. INSTRUMENTATION the standard suspension of zine sulfide in Lucite (purchased from Nuclear Enterprises, Inc.) cut into A. Neutron Detectors rectilinear pieces 12.5 mm square by 6.2 mm thick and mounted edgewise in Lucite light pipes using Nuclear A number of detectors have been used with the Enterprises. Inc. NE-580 optical cement. Finished as-collimator to measure the pattern of radiation emitted by semblies were coated on the sides and one end with the test section. For detection of fast neutrons, Hornyak titanium oxide reflector paint (Nuclear Enterprises, Inc.

Buttons" similar to those used on the TREAT NE-560) and coupled to Amperex Electronic Corp.

hodoscope,' 'lle-recoil proportional counters,"and XP111019-mm-diam photomultipliers.

12 D *

lD 0 ['lh TmMo Ah

PRE AMPLIFIE R DETECTOR HOUSING

~

(e)

(

ORTEC i t3B h

BASE CDVER WPUT CAP = 0 pF r

.5

' _L N Ikam n I

= = ces

,,,g.

y.

y l

auf Tom - -

5 oio) j d

+

ASSEMSLY -

l r

i 180A

0 02 l

D9) !

l t - = = "'

r p

i

,,3

!t r

n e tpgn%p+nyqp g.

gl a

D 7,j g-riMjpt

,.q g

om.o t

m.

sm..m.m.,_,

l i i= -

,2s BUTTON c.

o

, A s.s s' nie wwwee*** os cano*****

x D6 8

s s s s s s s m n '< w gl

'l l

Fig. 16.

os 2

AuPti,ic a Exploded view of a Hornyak-Button detector as-l

- .' *, *o,

'*K sembly.

O

'2" gl g

g im K

$I oiscaium AToa An exploded view of the detector assembly is shown in gl l

onncess

'"K Fig.16. The detector end of the brass photomultiplier I

or

'o l

hcusing screws into a brass plate attached to the I

I ' " '

l SCALER hodoscope collimator. The photomultiplier divider I

g network, shown in Fig. 17 together with the other t

g.

lectronic components used for these detectors, is ac-I I

cording to the recommendations of the tube manufac-

/

wrer except that capacitance values were chosen on the basis of response of the tube with a Nuclear Enterprises, Fis N' Inc. NE-102 scintillator to x-ray pulses from a 5-MeV Electronicsfor Hornyak-Button detectors.

linear accelerator. Values chosen minimized saturation and afterpulsing without imposing too long a recovery were counted using a discriminator setting just higher time on the divider chain after large pulses. It was also than that point at which counts were registered from a felt desirable to hold divider current to less than 1.5 mA, 5-mci '*Co source held adjacent to the detector. The since it was thought that large arrays of detectors might counting efficiency of the detectors as a function of be needed. A negative high-voltage supply was used to energy, derived from the data shown in Fig.18,is shown reduce noise and to eliminate the need for a coupling in Fig.19.

capacitor at the anode, which is at ground potential. This in turn required that detector assemblies attached to the

2. 'He-Recoil Detectors. Two 'lle-recoil proportional photocathode be nonconducting to avoid inducing counters were evaluated as neutron detectors for the spurious noise at the photocathode resulting from elec-present hodoscope studies at the P/o(KA facility and for trical discharge or static efTects in the glass.

an STF hodoscope system. These detectors are of Pulse-height distributions for monoenergetic neutrons interest because of their known low sensitivity to thermal incident upon the flornyak-Button detectors are shown neutrons and gamma rays, their relatively high overall in Fig.18. The detectors were tested at the LASL efficiency for fast-neutron detection, and their 3.75-MeV Van de GraafT accelerator using mono-pulse-height distribution, which permits neutron-energy energetic neutrons from the 'Li(p.n)'Be reaction. For discriminat:on. The detectors have an active size of 12.7 these tests, the neutron energy spread was approximately mm in diameter by 150 mm long. They were operated at 20 kev. Each curve is normalized to a neutron fluence of 1500 V with an ORTEC Inc. model 109 PC preamplifier 2.2 x 10' neutrons /cm, as determined with a 'lle long and model 472 spec.roscopy amplifier with time cons-2 counter" that monitored the fast-neutron flux. For use tants set at 3 ps.

on the hodoscope, amplified signals from the detectors 0**

W 9

13 OO O

HODOSCOPE HORNYAK BUTTON 4

10 I

I I

I i

ll CF-252 w

1.35 MEV g

{f 10 600 KEV d

400 KEV 3

5 MC CO-60

_ g, %,

p.g*

2 10 a

200 KEV N.g

\\

Fig. 18.

h h

h.,,

.}, 't,,

Pulse-height distributions of a g

& IO

? N*,

- "*.}*,f**,*, *,

Hornyak Buttonfor neutrons ofvari-o g

i E. "~ *.

?a.

ous energies andfor an intense gam-Q n..

. maa acca u

ma-ray source.

.. ao 0i i Lt i

e i

i t

10 0

100 200 J00 400 CHANNEL g3 i

i i

i W9g 02 Fig. 19.

gt Absolute counting eficiency of a Hornyak-Button

} On detector as afunction ofneutron energy.

y f

i l

f f

OO O200 0 400 0600 0800 1.0 12 I4 NEUTRON ENERGY (MeV)

These detectors were bombarded axially with mono-ably with the emciency obtainable with the 12.7-by energetic neutrons from the 'Li(p.n)'Be reaction at the 12.7-by 6.4-mm stilbene detectors discussed below.

LASL 3.75-MeV Van de Graaff accelerator.

Ifowever, because of the preponderance of low-voltage Pulse-height distributions for neutrons of various pulses in the 'lle-recoil detectors for neutrons of energies energies are shown in Fig. 20. The asymmetry of the 'He useful for fuel self-imaging, biasing the detector outputs neutron-scattering cross section causes maxima at the at substantial neutron energies results in very high losses low-and high-energy ends of pulse-height distributions in counting emeiency. For instance, for an energy bias of for neutrons of energies between 400 and 1200 kev.

1.4 MeV, the counting rate of these detectors mounted Figure 21 is a plot of the measured counting emciency of on the PARKA hodoscope is about one-twentieth of the these detectors as a functions of neutron energy. The rate available from the stilbene detectors with a similar solid curve is the ' lie total neutron cross section. The bias. A lower-energy bias results in less differentiation fact that detector emeiency falls off with decreasing between test section and driver neutrons.

neutron energy makes it especially easy to bias the output of these detectors against low-energy neutrons.

3. Stilbene Detectors. The use of organic-crystal and The measured emeiency for counting I-MeV neutrons plastic scintillators to detect neutrons is discussed in incident along the detector axis was 3.2% for a some detail by Swartz and Owen." In these media, fast pulse-height discriminator setting equivalent to 75 kev of neutrons produce recoil protons, which in turn give up energy deposited in the counter. This compares favor-their energy in ionization tracks in the detector; the 14

1600 g 7

1500 kev 1200 iogo g,y 700 kev

_E e

2OO kev

^

800

+

Fig. 20.

s

,$"9 #.'+s Pulse-height distributions of a i

'He-reccil proportional counter for s

0

~

A'

~

neutrons ofvarious energies.

i

.g a

._3

1 l

l I

I I

O'-

50 10 0 15 0 200 Channel

_ i I I I i l l

l l

l l,1 1 I i l l

l l I l l

l l _

[

\\

30

[/

\\.

N i

N

'l

~

N

/

x M

/

N

=

f N

U

/

\\

520

/

\\

'N W

/

x 5

/

\\

g

/

o e

U

/

-s3 10

,I 2

a o,/

l i I I I I l I I I I I I I I

I I l ! I I I I I O

O2 04 06 08 to 12 14 16 18 20 2.2 24 Neutron Energy (MeV)

Fig. 21.

Absolute counting eficiency of a "He-recoilproportional counter as afunction of neutron energy.

Discriminator threshold was 75 kev.

15

energy of the ionization tracks are then partially con-

[

'/

verted into light. If the light output of the detector were I

/

~

...o.o/on /

i 7

linear with the energy of the recoil protons, one would

/

expect monoenergetic neutrons to produce a rectilinear l

~

rygy

/

pulse-height distribution corresponding to n-p scattering, y*

/

/ no o.oiro which is isotropic in the center-of-mass system. How-f

/

ever, the light output is not linear with recoil-proton o*

y 3

energy. As a result, pulse-height distributions are dis-U

/

torted, with more pulses at the low-energy end of the

$2-j

/

"8 scale, and the maximum pulse height is not proportional

{

/

to the neutron energy.

8 o,

'j l, ' l, ' i On the other hand, gamma rays deposit their energy in iNctDENT ENERGY (MeV) the detecting medium by Compton scattering of elec-trons in a process that results in a nearly linear U

relationship between the energy of the scattered electron Light output (expressed in equivalent electron and the measured iigin output. Photons of energy E, will

$"#d b#

""# #" [

"# "#"###" ## ##**#### #"N !## ##

produce a continuous distribution of Compton electrons up to an energy E,,,, given by E /E O

1 + (m c /2E )

(4)

The pulse-shape analyzer gives an output pulse of

=

0 O

amplitude proportional to the 10-90% risetime of the 2

where m,c is the rest energy of the &ctron (511 kev).

incoming signal. Figure 24 is a comparison of risetime The distribution is peaked at E,,,

distributions obtained with a "8Pu-Be neutron source Figure 22 compares the measured light output (ex-and with a Co gamma-ray source. For this test, the pressed as equivalent electron energy) of a stilbene input discriminator of the pulse-shape analyzer was set scintillator as a function of incident neutron and gam-to accept pulses with amplitudes greater than one-half ma-ray energy. The fact that gamma rays produce on the the Compton edge from 662-kev "'Cs gamma rays. The average a greater number of scintillations in stilbene than equivalent neutron energy threshold is 1.3 MeV. The do neutrons of the same energy means that, in a mixed independence of pulse-risetime distributions on neutron gamma-ray-neutron field, the output of the sciutillator and gamma-ray energies and the resultant stability of the will be dominated by gamma events unless these are n-y discrimination point (the minimum between the suppressed. Pulses due to gamma rays can be suppressed gamma-ray and neutron distributions) was checked by because the average decay time of pulses generated by comparing the response to reactor radiation (as seen recoil protons from neutron interactions is greater than through the hodoscope) with the response to usPu-Be the decay time of pulses generated by Compton elec-neutrons.

trons.

A block diagram of the electronics used with the stilbene detectors is shown in Fig. 23. The crystals, B. Gamma-Ray Detectors 12.7-mm square by 6.4-mm-thick, were coupled

. geometrically with the hodoscope slots by embedding Not all proposed techniques for STF fuel-motion them edgewise in Lucite light pipes. The scintillator monitoring have involved fast-neutron imaging. Some of assemblics were viewed by Amperex Electronic Corp.

the coded-aperture techniques use gamma radiation as XPil10 photomultiplier tubes, the signals from which their image-forming medium.' Furthermore, it has been were passed through ORTEC Inc. model 113 pre-proposed that clad motion be monitored by imaging amplifiers,. 300 m of RG63/U cable from the high-energy capture gamma rays from iron. The use of critical-assembly building to the control room, and then the hodoscope for gamma-ray imaging, particularly for through ORTEC Inc. model 460 delay-line-shaping imaging capture gamma rays from iron in the face of an amplifiers. The outputs of the amplifiers were analyzed intense but lower-average-energy fission-gamma spec-l for risetime distribution using ORTEC Inc. model 458 trum, requires optimization of the gamma detector.

pulse-shape analyzers.

Ideally, one would like a detector that is small (to L

'16

Er1 Sconner Stepping Motor 300 m RG 63/U Preamp pg Ortec 113 xp rino HV Sislbene Preomp py Ortec 113 XPillO Reactor Fission Ch ber Set-up a Mondor System (Disc ) f' I $

Delay IS'9 I Lineor lPreompl Amp PSA Amp PSA Ortec 427 Gote Ortec Ortec Or tec Ortec

- TC 303 460 458 460 458 l'@-Gote Gen g

Defor Scanner s

control l PHA l Position ii TC 410 b'l

!g U

Controt m

too,

_/\\.

8 o

put)

I

} 2/t/'

't ll Disc H^ mal (BCD Sig)

E

'b-/

)

i

@ @ l@ l5

/

Time Monual Buffer Timer T,

T, y,

y, n,

n, FC Print TI 733 ASR Of Doto Porollel Control Doy M ry Doto Ortec Ortec Terminal TC 565 TC 585 TC587 773 772 Scalers Fig. 23.

Block diagram ofelectronicsfor hodoscope scans with stilbene detectors.

30 minimize sensitivity to ambient backgrounds and to i

i make arrays oflarge numbers of detectors practical), has gg

-[y a high overall detection efficiency for high-energy radi-23e p,. g, ation, and has a response peaked at full-energy absorp-j tion to assure good energy discrimination.

. io-rj

["

5o

. ih I. Nal(TI) and NE-102. Hodoscope gamma-ray im-5.

I aging of LMFBR fuel test assemblies has been studied 7

,f

,,c using stilbene, Nuclear Enterprises, Inc. NE-102 plastic, d

and Nal(TI) scintillators. The plastic and Nal(TI) scintillators, cylinders 12.5 mm in diameter by 12.5 mm

~

long, were mounted on Amperex Electronic Corp.

0" XPil10 photomultipliers. Signals were processed with o',

o'2 the same tube bases, preamplifiers, and amplifiers that i

o 03 o.4 Pulse Rise Tune ( s) were used for the Hornyak Buttons. It was found that the small Nal(TI) detectors have a rather low Fig. 24.

peak-to-Compton ratio for gamma rays at energies Pulse-risetime distributionsfrom stilbene detector greater than I MeV. Moreover, activation ofiodine, with 23:

for a Pu-Be neutron source and for a Co resultant buildup and decay of the 25-min 1 activity, 2:

gamma-ray source.

has been a deterrent to their use. Plastic and stilbene detectors have even less optimum pulse-height responses for the range of gamma-ray energies involved.

m

3

2. Bismuth Germanate. Newly introduced scintilla-16 e,"

662 kev _

tion-grade bismuth germanate (Bi,GeOn) crystals have goo the desired detector characteristics." This material, twice is7 3

as dense as Na!. has an efficiency for gamma absorption 12 such that pulse height distributions from a bismuth germanate (BGO) crystal are equivalent to those from a Nal(TI) crystal with linear dimensions three times 8 ;

p ts.4 % -

larger. Thus a $600, 38-mm BGO detector is nearly equivalent to a 125-mm Nal(TI) detector worth $2000.

4 Three BGO crystals, a 38-mm-diam by 38-mm-high h

g cylinder, a 12.5-mm-diam by 12.5-mm-high cylinder, x

tb t

and a rectangular piece 12.5-mm square by i

_t i

i i

1 1

o -

20 so 80 10 0 12 0 6.2-mm-thick, have been purchased from the Harshaw

]o Chemical Co. for evaluation. The latter two crystals y

were used with the PARKA hodoscope. Since the light I

I r

i i

i I

i output of BGO is only 8% of that of Nal(TI) for incident f4

,.--662hev photons of a given energy, a low-noise photomultiplier is cs needed to obtain optimum pulse-height resolution. We 3

found that an Amperex Electronics Corp. XP2000 SI-mm-diam tube gave satisfactory results. Amperex XP111019-mm-diam photomultipliers, which have been 2

-- e.o +4 used for the Hornyak Buttons and stilbene scintillators in the hodoscope measurements, were found to be quite

-d*

satisfactory for use with the smaller BGO detectors. The BGO crystals were coated on the cylindrical surface and

'g

, h one end with titanium dioxide reflector paint. The rectangular crystal was mounted edgewise on the 19-mm o

50 10 0 150 200 250 photomultiplier. Electronic equipment used for testing Channel Nurrber included an ORTEC Inc. model 266 photomultiplier base, model 113 preamplifier, and model 472 spec.

Fig. 25.

troscopy amplifier. The amplifier shaping time constant Pulse-height responses of 38-mm by 38-mm BGO was 3 s.The photomultipliers were operated at 1300 V.

and NaI(TI) scintillators to "7Cs gamma rays. A In Fig. 25, the pulse-height response of the 38-mm 1-pCl source was countedfor 200 s on contact with BGO crystal for "'Cs gamma rays is compared to that theface ofeach detector.

of a 38-mm Nal(TI) detector. A 1-pCi source was counted for 200 s on contact with the face of each interfere with the detection of small amounts of lower detector.The full width at half-maximum (FWHht) from energy gamma rays and makes spectrum unfolding the full-energy peak of the BGO response is nearly twice dimcult. At 1.33 hieV, the full-energy-peak emciency of that of the Nal(TI) distribution. However, the the BGO detector is 4.5 times that of the Nal(TI) peak-to-Compton ratio of the BGO detector is con-detector.

siderably larger and its backscatter peak is somewhat As photon energy increases, BGO becomes the clear smaller.These characteristics should simplify the unfold-choice over Nal(TI). We illustrate this in Fig. 27, which ing of eontinuous spectra. The full-shows the responses of the two 38-mm detectors to 1.37-2 energy-peak emeiency of BGO is 3.3 times that of and 2.75-hieV gamma rays from the decay of *Na. At Nal(TI) for these detectors.

2.75 hieV, the BGO detector shows a predominance of The pulse-height response for Co gamma rays is the full-energy peak over the annihilation-escape peaks shown in Fig. 26. The resolution of the 38-mm BGO and the Compton shoulders. For the Nal(TI) detector, detector is not quite sumcient to resolve the 1.17-and the full-energy and escape peaks of the 2.75-hieV line are 1.33-AteV photopeaks. However, the Comoton peak in approximately equal and the Compton peaks of both the 38-mm Nal(TI) detector is now prominent enough to lines are quite evident. The emeiency of the BGO 18

F'

,'7 y,'y 8000 BGO

- i,33 yev 60

~

Co 6000-I; 4

e.

t 4000,. g

11. 8 */.

-)

2000

(

~

l m

1 1

I f

f f

f i

i 1

w 1000 i

0 50 10 0 15 0 200 250 300 1500 3

i i

i p i

i i

i 5

N-4.17 MeV j

60 No! (TI)

Co

,/-- 1. 3 3 MeV h,%

h' Fig. 26.

1000 Pulse-height responses of 38-mm by 38-mm BGO and Nal(TI) scin-

-7, tillators to '*Co gamma rays.

(#

,J,.- 6. 6 %

500 o

4 mm i

f I

l*-

I o

50 too 150 200 250 300 Chennel Number detector for the 2.75-hieV full-energy peak is 5.6 times in the "N(p,y)"O reaction. With the detector 2 cm from that of the Nal(TI) detector.

the target, we obtained the pulse-height distribution Figure 28 shows the response of the 12.5-mm BGO shown in Fig. 30 in 84 min with a 2-pA proton beam, 2

cylinder to *Na radiation. In terms of peak-to-total ratio The 8.28. 5.24-and 3.04-hieV lines are from the decay and total counts per photon incident upon the crystal, of the 8.28-hfeV level of "O to the ground state. The this detector is roughly equivalent to the 38-mm Nal(TI) 4.43-AtcV radiation, from the "N(p.ay)"C reaction, is detector.

due to the use of normal isotopic-content nitrogen as a The response of the 38-mm BGO detector to target.

4.43-hieV radiation is plotted in Fig. 29. This radiation, In Fig. 31, the full-energy-peak efliciency of the from the decay of the first excited state of "C, was 38-mm BGO crystal for 0.1-to 8.28-hieV gamma rays is derived from the 'Be(a.n)"C* reaction in a 2nPu-Be compared with that of a 38-mm Nal(TI) crystal. For neutron cource.

energies less than 2.6 hieV, absolute efliciencies were The response of the 38-mm BGO detector to higher obtained by placing calibrated radioisotope sources 10 energy radiation was measured by observing gamma em from the front face of the detector. Relative efficien-rays from the decay of the 8.28-hieV excited level of "O.

cies from 3.04 to 8.28 hieV were obtained from the To produce this radiation, a thin film of tantalum nitride knowi. branching ratios" of the decay of the 8.28-hieV on a platinum backing was bombarded by 1.06-htcV level of "O and were normalized to the radioisotope data protons at the LASL 3.75-hieV accelerator. The oh-by matching the slopes of the two parts of the curve.

served gamma rays result from a 5-kev-wide resons ice 19

60 4

/ **

_1.369 MeV BGO 48 24 g

36

--+. * -- 10. 9 %

2.754 MeV 24

//

j/

o

/

g 12 j

7.9 %

2 e

E

'( '

20-u 50 10 0 15 0 200 250 300

{. 20

, g, i

1.369 MeV

=

c Not (TI)

=

N*

I5 Fig. 27.

s s

Pulse-height responses of 38-mm by 6

-- 6.4 %

,7,754 Mev 38-mm BGO and Nal(TI) scin-2

-~

10 <

tillators to "Na gamma rays.

/

\\

=

/

5 1 =.(

.f.

D.

p 5.8 %

t

'i.

I f

f i

1 1

i 0

50 10 0 15 0 200 250 300 Channel Number 20

3500 Y

e 3000 24 No r I.*69 MeV 2500

+

i c

2 2000 -

S u

s' j4500-

.- 10. 0 %

a a

1000-b *.

4*

-, \\

.754 MeV -

2 pfg, yg, Pulse-height response of a 12.5-mm s

\\

by 12.5-mm BGO scintillator to '*Na O*

a\\

gamma rays.

500-

fs s

N 7.4 */.

I l

I I

O 50 10 0 150 200 250 Channel Number i

i i

I i

i i

~

BGO 9

l2 #

B e (o,nl C

r 4 4 3 - 0.511

\\

r 4.43

\\

4

.e*

o

=

ia f x k *,

3 V

4 43-i02-d T.

[*

x

&. W. 7.3 % ~

g 0-Fig. 29.

Ti Pulse-height response of a 38-mm by e

b ~*

S

~

38-mm BGO scintillator to 4.43-MeV gamma rays.

g e

i l

I f

f I

f l

0 50 10 0 150 200 250 Channel Number l

21

I I I I I I I i i I i i i i i i i i i i I 800

'* N ( p, y) '80 8

g 3.04 (8.28 ~ 5.24) e

.'8

'O 4.43 N(p,ay)12C N

-*6

/

8.28 - 0 f

5.24 - 0 l

2 l

/

/ J,

}4

(*b 4./ i j

s-4 weV p.v./.

i a

~

zus

\\../ el.28 2

V L

g s

I iI i e I

i i iI I I

I I I I

I I

O 20 40 60 80 10 0 12 0 14 0 16 0 I8O 200 220 Channel Number Fig. 30.

Pulse-height response of a 38-mm by 38-mm BGO scintillator In gamma raysfrom the 1.038-AfeV resonance of the "N(p,y)"O reaction.

10 0-1 I I I Iilli l

i i I illL 38mm BGO

\\

N 40 Co l

l

No

'8 57 0

'33 l

- No /

!'o

' I I 80 j N//is ll e 20

=>d s j;j i

/ \\l, i l iN N

li l

x s

10-13 7 - /

hll i

1 i

/-

c c

8 :

15 "o

544/

l 0

l l 6

/

i h

~

38 mm No!(TI)/ 60 p,

g

~

Co I

E

~

l

~

gg y

Fig. 31.

4 No

\\

full ener81-Peak efficiencies of

}

soCo toe,

38-mm by 38-mm BGO and Nal(TI) 2 -

7 scintillators as functions of gam-ma-ray energy.

I I I I Ilill l

I I I I I 11

,01 02 04 06081 2

4 6 8 10 Photon Energy ( MeV) 22

.~

60

, y,,,,

- S7

,/c, 40 ear 20-Cs g

soCo

/

Co 24 No e

22 54,,7 f2 4 No

-f io eg C

p j8 [

No

/

is *[

E

- No 2*No 18 l*

\\~

Fig. 32.

6 0

880' Resolution of the 38-mm by 38-mm BGO scintillator as a function of I

f f I l l tll f

I l l l l fl 4Oi 0.2 04 06 08 1 2

4 6

8 IO Gommo.Roy Energy (Mov)

The resolution of the 38-mm BGO detector is shown It was necessary to determine whether BGO detectors in Fig. 32. Although this resolution is only about half would show effects of activation or radiation damage in that of equivalent Nal(TI) crystals. the higher etliciency the environment in which the PARKA hodoscope detec-and the near absence of Compton continua make BGO tors operate. Accordingly, the 12.5 mm by 12.5-mm the detector of choice for high energy measurements detector was mounted in front of one of the collimator when high resolution is not needed. Its ruggedness, slots while PARKA was operated at 5 mW/g *U for comparable to Pyrex, and its nonhygroscopicity make it 4 h with a 37-pin test assembly in the test well. The an ideal candidate for field applications such an environ-collimator slot was held fixed on the center of the test mental and health monitoring, nuclear safeguards as-assembly and the count rate from the detector, biased at says, and uranium exploration. It would be an ideal 1 MeV was taken every 5 min, starting at the time the detector for downhole assay by capture gamma-ray reactor reached full operating power. The resulting data measarements.

are shown in Fig. 33, together with those obtained simultaneously from a stilbene detector and the results of a previous run using a Nal(TI) detector. The ratio of the 200,000 BGO

, e e r :

0 ; ;

0 ;

E >l.0 100,000 7

80,000-

- .rene

~

e 60,000 - - ^ ^,.......

'Ey >O 33 T

_ _ _ _ _ Nal(TI)

Ey > O 66

40,000 8

}

Fig. 33.

m Count-rate b:didup in g gn,nnn B GO, stilbene, and 8

Nal(Tl) detectors used with the PA R KA hodoscope,

' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' i iO,000O 10 20 30 40 SO 60 70 80 r

240 j

Time From Stortup (min) 23

BGO and stilbene count rates remained constant over IV. HODOSCOPE SCANS the entire 4 h run, suggesting that the 13% increase in count rate for both detectors was related to the radiation A. Single Pin Scans environment and is probably associated with fis-sion-product buildup in the test section. The departure of Figure 35 shows the results of a hodoscope scan the Nal(TI) data from this behavior is obvious.

across a single fuel pin with two llornyak-Button A pt'Ise-height distribution from the 38-mm BGO detectors in the same horizontal plane. The reactor was 2

detector mounted in front of one of the collimator slots is operated at 550 W (4.6 mW/g "U). The count time for shown in Fig. 34. For this measurement, the collimator each point was 500 s. The S/B ratio is low, only 0.2 was pointed at the edge of the test assembly with the compared to a typical S/B ratio of 3 for the TREAT 21 mm-thick steel test capsule in place. In this position, hodoscope.' The low S/B ratio of our hodoscope is due the detector was "seeing" a total thickness of 154 mm of to the epithermal neutron spectrum of PARKA. The steel. We identified the observed high-energy radiation as driver neutrons of the TREAT reactor, a thermal capture gamma rays from iron. The indicated peak reactor, are on the average pasier to separate from the energies correspond to available data." The 7.64-MeV fast neutrons emitted by the test region. The harder peak is due to decay of "Fe to the ground state after driver spectrum will be needed, however, to produce thermal-neutron capture by "Fe. liigher energy peaks, uniform excitation across large fuel-pin arrays; the at 9.30 and 10.16 MeV from neutren capture by "Fe capability of the hodoscope technique must therefore be and "Fe, were too weak to be useful for these studies, verified before incorporating it into the design of STFs The spectrum displayed was accumulated in 3000 s with for large-bundle tests.

PARKA again operating at a power level of 5 mW/g The scan shown in Fig. 35 was made with the viewing 2" U.

slot extending only halfway through PARKA. The Energy (MeV)

O I

2 3

4 5

6 7

8 I

I I

l 82 -

343 lJ

'"o iO I

k Fession g8 8

_* fi56%

Gaminos

$ 92+ t>O3 l

764 r

.Qis l

(p

.S w

4 - -

W S.

2 3 - s.

4 9e 2

\\

7.2g I

I i

i e

i i

i e

i O

SO 10 0 150 200 250 Channel Number Fig. 34.

Gamma-raypulse-height distributionfrom the PARKA hodoscope using a 38-mm by 38-mm BGO scintillator. Identified transitionsfrom thermal-neutron capture in fron arefrom Ref.19.

24

. _ _ ~.

~

l I I I I I I l I I I i i 1 I 64-62-

- Fuel Pin Projection -

60-g 0 $g_

Channel # 1

---Chonnef #2 x 56-( 54-6 52*

B 50-M'A

,E_ ' 4 eh flg. 33*

O 4g_

Y iko-channelscan with Hornyak But-E

\\

A

~

N A- --4

[

tons of a singlefuelpin in PARKA, y_

s with the viewing slot extending only to 49_

~Hodoscope slot widtn

-~

the center of the core.

3g_

I i 1 1 I I t l l l 1 1

!-I l

' f 2 8 4 - 2 0 2 4 6 8 10 12 14 Position At Fuel Pin Plane (mm) single-pin S/B ratio has since been improved by extend-slot to support the fuel elements above the slot. This ing the slot through the reactor so that the collimator channel has since been replaced with a lighter, sym.

slots do not "see" any PARKA fuel. As shown in Fig.

metrical support.

36, the slot extension has improved the S/B ratio by a To study factors afTecting the S/B ratio of the test-fuel factor of 3. This indicates the possible need for through image, single-pin neutron and gamma-ray scans were I

slots for either hodoscope or coded-aperture imaging in m Je using stilbene, NE-102, and Nal(TI) scintillators safety test facilities, llowever, the half slot may be and 'lle recoil proportional counters. Simultaneous neu-

. adequate for full-subassembly or larger tests, where the tron and gamma-ray data were obtained from each of signal from the test bundle its > imay dominate the driver two stilbene detectors. The output of one of these background.

detectors was biased with' a disedminator setting cor-The high counts on the right side of the through-slot responding to gamma-ray events with E > 0.33 MeV y

scan shown in Fig. 36 are due to scattering from i. piece and neutron events with E,, > l.3 MeV. The other of aluminum channel placed in the extended PARKA stilbene detector was biased for E > 0.66 MeV and E,, >

y 2.2 MeV.

'600 i i i i i l l i i i i i i i :

1400-1200-3

~

Singte pen scon y 1000-I E

/,- ~,~N~~_

.l 800~

rig, 36, i

~

Comparison of single-pin scans with

~

Hornyak Buttons before and after

=

Througn-slot weosurements

~ _ __ _ 9,,,.,,,,,,,,,,,,,n,,gna,ma,,,,ay extending the viewing slot through the f f I f f i i f I i t I t I 1 i t i

+12

+9

+6

+3 0

-3

-6

-9

-12

-15 i

Deflection (mm) 25 t

I I

I i

l I

e i

i e Single Pin $ con A Background j

10000 En > l.3 MeV 9000 -

g 8000 e

0 7000 -

A$aa4a a e a a S I 8-6000 -

I

-8 N

j 3000 g

E n > 2.2 M eV e

e j-j 2500 S

Fig. 37, 2000

~

t Neutron scan with stilbene ofa single

~

y e

e e

fuelpin.

O

' O A

ae 1500 I

I 1000 +12

+9

+6

+3 0

-3

-6

-9

-12 Deflection (mm)

Results of the stilbene neutron scanning of a single pin The behavior of the single-pin neutron image was in PARKA are shown in Fig. 37, together with back-studied as a function of threshold energy using a ground scans of the empty PARKA test section. The

'lle-recoil detector with energy biases of 0.35,0.7 and asymmetry in the background radiation is probably 1.4 MeV. The results are shown in Fig. 39. Because of caused by scattering from temporary shielding between the lower energy biases used for the ' lie-recoil detectors, the hodoscope and the reactor. As would be expected, the S/B ratios for a single pin are poorer than those the higher discriminator setting resulted in a better S/B obtained with stilbene detectors and Hornyak Buttons.

ratio: 0.75 for E,, > 2.2 MeV compared with about 0.5 In Fig. 40, the ratio of the single-pin count rate to the for E,, > 1.3 MeV. The S/B ratio for the lower driver-neutron count rate is plotted as a function of discrimnator setting is only slightly better than that discriminator setting (neutron energy bias) for *He-recoil obtained with llornyak Buttons biased for E,, > 1 MeV.

and stilbene detectors. Higher energy biasing than that Ilowever, the stilbene detectors are five times more shown in Fig. 40 for the *lle-recoil detectors was not etlicient than the Hornyak Buttons, so that statistically feas;ble, because of the resulting low count rates.

better data could be taken with lower critical-assembly Plots of single-pin scans taken with 12.5-mm-diam by power and less operating time.

12.5-mm long Nal(TI) and NE-102 scintillators are The corresponding single pin gamma scans shown in shown in Figs. 4 I and 42. These results are similar to the i

Fig. 38 exhibited S/B ratios of ~1, somewhat better than gamma scans with stilbene detectors. The figures are I

that obtained for neutrons. This is contrary to our superpositions of unretouched computer printouts from a experience with larger test bundles.

program written to correct for buildup of fission-and 26 l

~ -., - -

.-~

I t

32 i

i i

lee I

28 a

i E7 >O.33 MeV r

8 24 l

20

~

e e

I

4; 16

e1A A A A A A A A i4 ee 4

4 e9 e

4 x

{12-Ey > O.66 MeV j

3 e

I Fig. 38.

0 l0 l

l Gamma-ray scan with stilbene of a e

singlefuelpin.

g_

e i

~

g*g81A e 8A_

A A A &A A A A 6

i 4

i e Single Pin Scon l

2 a Background O

+12

+9

+6

+3 0

-3

-6

-9

-12 i

Deflection (mm) 7 f

1 4

5 i-4 27 i

l

5400 5000-Single pin scan

_ e Background scan eg_g y

[g 4600-En > 0 35 MeV f

\\

t t

{/

4200

\\

e

-tJ' 8 bf]-

8 ' 8 3800 m

e 9

9

,2400 f'f g

3 I 2200 En> O 7 MeV 5

h

}

Fig. 39.

2000 h

h Neutron scan of a single pin with a 0

h-hla

'}

"He-recoil proportional counter at l800 E E n

Various threshold energies.

4 4

600

~

En > i 4 MeV

~

sf 400

'f f

j 300

'I l

+9

+6

+3 0

-3

-6

-9 Deflection (mm) i i iiiii1 I i FT7 i i i i I I I I i 1 08 A

07 06 o' 05 g

s S 04--

He Detector 3

L

Stiioene Fig. 40.

93 Effect of neutron

~

energy bias on the S/B ?

=O2 ratio of hodoscope 5 scans of a single fuel

~

O'

~

pin.

l ! ! I I ! I I I ! 1 ! I i l l l I I I ! I I I O

O2 04 06 08 to 12 14 16 18 20 22 24 En Discriminator setting (Mev) 28

$tt 34*20 JLr ? 77-SthCLE P!w Sta%. ho LwCtfr PLuc ha I (TLD E

O 66 **e v e.

Se r st**uLLt3 FLtb eth. 20 JwLe

17. No Lucitt PowG 4eede

'3 V

/

35ede-

-/

\\

j 30eM -

l b -l

/.. - - -

w

,A 3

1 7+

,s.

Sff 14*F0 Jute 77 St%CLE o g ee SCah. NO LLCtfE PLuc tea I (fL) E* t.30 me, e.

set s t**wLt 3 PutL PIN. 20 JwLe 7?. h3 Luc t tt PLuo 88000 16408

\\

,/

A,.

p4.e..

f

/

\\

O IIdde

,}

-~

M d

x-

_./

,4%.-

O

%. 4

.i J

v t3 0

o SFT 14*10 JLt v 77. $1NCLE PIN SCA%. NC L'A l 'C

Lu".

Na i ETLD E

2 60 New set e s *=wtED Fut. P:N. 23 JuLe 77

%Q Luct'L *Lu3

.y.

87800-

Fig. 41.

Single-pin gamma-ray scans with a 12.5-mm by 12.5-mm Nal(TI) scintillator at various energy biases.

l 29

e SEf 3 4

20 Jta.T 77-SINGLE Pt= Scam. No Luciff PLLC P4 402 E
O 66 Mew e
sef I S **wLED PtKL Pim. 20 Juke
11. No Lucitt PLuo taeee.
f
\\
tesse
/
N ltese.
t hw.
34eee.
__s y
=
=
Ite**
e SET 14*20 JULv 77 5*NCLE PIN SCAN. NO Luct?E PLuc P4 102 E* 1.30 New e
ter es**utLED autw pth. 20 Juke 17. No LLA: s tt PLuo
+e98 C
b 6
O tese.
t3 6000.
E
(
/
/
late.
HT i WV%A
/
./
H seee e
SET t esto JAf 17 SINGLE 8!9 SCAN. hO LUC:ft PLUG hE 802 E
2.60 New e
st? asepuwttD Fuf. Pl%-
20 JuL' 11 NO luc 1't P.60 1864 T
.L.
~~
1600
~~
'T kY yzq y4 1
+ y -
1 6
-4
.g.
'g,
'g,
'4,
'6.
's.
p.
it.
13-
10.
-5.
PCSittom suma i
l Fig. 42.
Single-pin gamma-ray scans with an NE-102 plastic scintillator at various energy biases.
30
r I
I I
I l
l 1
1 I
fg
,8400-37 pin seon g 8200-k
. {k n
k f-
$8000-f y
y y
g h
k
~
~
k y f { 36 pin scan g
z 7600-y 7400-p 1.10-n g
1.08-4 g Robo S
1.06-104 -
n g
L 5 102-i, o
o o
..lN u 300_
ir-i i
I I
I I
I 1
+12
+9
+6
+3 0
-3
-6
-9
-12
-15 Deflection (mm)
Fig. 43.
Across-J1ats scans with Hornyak Buttons of a 37-pin assembly, intact and with the centerfuelpin withdrawn.
activation-product activity during the scans, normalize substitute for sodium. Th: result (Fig. 44) was little for integrated reactor power, calculate statistical errors, difTerent from that obtained previously. There.was a and plot the results as a function of scanner position, drcp of ~10% in total count rate, but no apparent loss in resolution.
Stilbene detectors biased for 0.33-and 0.66-MeV B.- 37-Pin Assembly Scans gamma rays (1.3-and 2.2-MeV neutrons), were used for across-flats scans of a 37-pin assembly, intact and with The scanning hodoscope was used with llornyak the center pin missing. The neutron and gamma-ray Buttons to determine whether one missing pin in a 37-pin count data (Fig. 45) were taken during a 3-h run starting assembly could be detected. The results of this test are immediately after reaching operating power. The i
shown in Fig. 43. The count time for each data point was count-rate increase with time for the gamma-ray and 200 s.The missing pin produces a 6% decrease in count lower-energy neutron scans is caused by buildup of rate over a distance consistent with the pin diameter and fission and activation products. It should be noted that the collimator resolution. It should be noted that the both scans were taken from left to right and that the absolute count rate difference between the 37-pin scan 36-pin scan followed the 37-pin scan. We subsequently and the 36-pin scan at the center of the missing pin is found that most of this problem may be eliminated by about 40% of the net count rate from a single pin (Fig.
running PARKA at the experimental power level for I h 36). PARKA power levels were identical for the sin-before starting to take data. One may, if desired, also gle-pin and 37-pin scans. This indicates the degree to measure this time-dependent buildup and correct for it.
which neutron scattering and absorption may be ex-This clTect is, of course, less important to the per-pected to influence imaging results, formance of multichannel hodoscopes, with which all l
The 37-pin scan was repeated after increasing the data are taken simultaneously.
' thickness of steel around the test assembly to 12.7 mm The time-related buildup for the scan with E,, > 1.3 and filling the space between the fuel pins with drilled MeV is probably duc to incomplete separation of neutron aluminum discs. The aluminum is a reasonable neutronic and gamma-ray counts caused by slight overlap of 31
a
(
l I
I 1
I I
l l
l l
l l
l l
l l
l l
7600 37 pin scan ff f
4, il dl I
7400 - fep 4,
<> T o
g 7200 g
fg 9
g 7000-ff f 36 pin scan
~
6800 With entro steel and aluminum mockup of sodium 1.10 1.00 -
Rotio
-f 1.06 m
I.04 p
s 0
1.02
.t T
l00 O-1 l
l l
l l
l l
l l
l l
l y
I
+12
+9
+6
+3 0
-3
-6
-9
-12
-15 Deflection (mm)
Fig. 44.
Across-lats scans with Hornyak Buttons ofa 37-pin assembly, intact and with the centerpin f
withdrawn, with aluminum mockup ofsodium and 12.7 mm-thick steel casing.
32
(
66
, i 64 g,
4
^
62 A
A 60
^
'L 8 = 4.6 %
A A A 58 A A A
A 56 -
EY > O.33 MeV A
54 34
. - 36 Pin scan ( enter pin missing) 2 33
- 37 Rn scan 32
- 8 = 5.9 %
I v
, 3i 1*
9 30 A
A A
A l
A A A A
^
{ 29 A
{
A 3 28 o 27 y>0.66 MeV o
A E
4 9
25
~
-8 = 5.4 %
24 p
IA Fig. 45, 23 Aye $ g
\\
Across-flats scans with stilbene detec-22 A A A A A O A A
tors of a 37-pin assembly, intact and ay _A En >l.3 MeV with the center pin withdrawn.
4
-8 = 7.0 %
8 En >2.2 MeV f
,gA g-
,n 7.5 14"*I*8* - "
Ie i e e i i 1 -) en7 7 I I I I I l
7 l
+ 12
+9
+6
+3 0
-3
-6
-9
- 12 Deflection (mm) i l
I I
I l
r l
l 33
)
pulse-risetime distributions as measured by the The data of Table Il show that the absolute neutron pulse-shape analyzer. One might also expect contamina-count loss caused by withdrawing one pin from the tion of the gamma-ray scan data by neutrons that are 37-pin assembly is less than one-half of the net count captured or inelastically scattered in the vicinity of the (with background subtracted) from a single pin in the test detectors and are then indirectly "seen" as gamma ray section. For gamma radiation, the situation is much events. To test for intermixing of gamma-and neu-worse: the count loss due to withdrawal of one pin from tron-induced counts in the detectors, 37,36,1, and 37-pin bundle is las than one-fourth of the net count 0-pin scans were repeated with 20-cm-long Lucite plugs from a single pin in the test section. This raises questions inserted into the reactor end of the hodoscope collimator concerning the hodoscope's response linearity and its holes. The efTect of the Lucite plugs on gamma-ray and sentitivity to the position of a perturbation in fuel density neutron counting rates is listed in Table II. Crude within the test bundle.
estimates would indicate that the attenuation should be We invectigated the position sensitivity by scanning a
~3 fa reactor gamma rays and 2 to 3 orders of 37-pin bundle with a pin withdrawn from the center, the magnitude for fast neutrons. We also note that most of front (nearest the collimator), and the back. The results the neutron counts for the 0-pin, Lucite-plugged case are (Fig. 46) show a general reduction in sensitivity from due to room background rather than to neutrons coming front to back of 20-30% for fast-neutron scans and down the collimator, which explains the apparent lower 35-50% for gamma scans. The center (20-mm) points attenuation of these neutrons by the Lucite plugs.
have been corrected for a known 15% power distribution depression in the center of the test bundle. The indicated errors are from counting statistics only.
TABLE II RESULTS OF TEST FOR INTERMIXING OF DETECTOR COUNTS INDUCED BY GAMMA RAYS AND NEUTRONS Counts per 200 s at 5 mW/g U
Collimator Collimator Plugged Scan Open with Lucite Attenuation 37-pin, E, > 2.2 MeV 7 720 450 17.0 37-pin, E, > l.3 MeV 23 400 1300 18.0 37-pin, E > 0.66 MeV 30 700 10 600 2.9 y
37 pin, E > 0.33 MeV 63 400 19 300 3.3 r
36-pin, E, > 2.2 MeV 7 180 400 18.0 36-pin, E, > 1.3 MeV 22 050 1250 18.0 36-pin, E > 0.66 MeV 28 900 10 000 2.9 y
36-pin, Ey > 0.33 MeV 60 400 18 200 3.3 1-pin, E, > 2.2 MeV 2 800 175 16.0
)
1-pin, E, > 1.3 MeV 9 100 600 15.0 1-pin, Ey > 0.66 MeV 14 000 5400 2,6 1-pin, E > 0.33 MeV 30 400 10 000 3.0 y
0-pin, E, > 2.2 MeV 1600 110 14.5 l
0-pin, E, > 1.3 MeV 6 000 450 13.3 0-pin, Ey > 0.66 MeV 6 500 3 200 2.0 0-pin, Ey > 0.33 MeV 16 000 6 200 2.6 34 l
k
I I
I i-i I
I i
i En >2 2 MeV En > l 3 MeV.
8 Ey >0 33 MeV 7
Ey >O 66 MeV 8
s
N
[o6 s
N N
o f5
~.
s
~.
Fig. 46.
7h-e N ~%
. Sensitivity of hodoscope response i N
with stilbene detectors tofuel-pin re ) 3 K
moralfrom a 37-pin assembly as a Q function of the position of the void in - y g the assembly.
y a:
1 l
l I
l l
I l
l I
0 O
10 20 30 40 Distance From Frora Foce (mm)
Response linearity was tested by scanning with one to I or 2% of the counts from the *lle-recoil detectors were four pins withdrawn in line from the assembly. The due to room background.
results are shown in Fig. 47 for neutrons and in Fig. 48 The *lle-recoil detectors have too low a counting for gamma rays. The neutron scans (Fig. 47) exhibit efficiency for fission-spectrum neutrons to be of regular linear response within the indica:ed statistics but the use in PARKA hodoscope experiments, but their in-i gamma scans (Fig. 48) show a slight (5-10%) increase in sensitivity to gamma rays and thermal neutrons may response over linearity.
make them candidates for use at STF power levels. They The 37-pin assembly with zero to four pins withdrawn do, however, have a longer pulse risetime (1-2ps) than do was scanned across flats with
lie-recoil detectors at an plastic scintillators, which may cause nonlinearity at high energy threshold of 0.35 MeV (Fig. 49). The response count rates. Furthermore *lle-recoil detectors seem to linearity with these detectors at various energy have less long-term stability than do stilbene and thresholds is shown in Fig. 50. The
lie recoil detectors llornyak-Button scintillators.
are far less sensitive to background gamma rays and Figures 51 and 52 show computer-generated plots of thermal neutrons than either the stilbene detectors or the across-flats scans with Nal(TI) and NE-102 scintillators flornyak Buttons. With 20-cm-long Lucite plugs block-of the 37-pin assembly, intact and with the center pin ing the hodoscope channels, we found the ' lie-recoil withdrawn. These data have been corrected for detector count rates attenuated by factors of from 40 to time-dependent buildup of fission-product background 64, compared to 15 and 18 for stilbene detectors.
and activation of the Nal(TI) detector. Results are little
~
Doubling the length of the Lucite plug showed that only difTerent from those obtained with the stilbene detectors.
MO 90W3h D
D n d d al Mrb i
j
32 I
i i
I l
E 28 8o
, Y-
{ 24
/
En > 2.2 MeV
=/,
o s
y 20-g 3
,/
$l6
. )s,'/
Fig. 47.
~
En>I.3 MeV
~
Linearity of hodoscope neutron re-
'?z sponse with stilbene detectors to i
i~
~
~
fuel-pin removalfrom a 37-pin as.
b8
~
/
~
sembly.
4 1
i 1
I O
I 2
3 4
Number Of Pins Removed 32 9
28 s'7 2
/
g
/
4-f 8 24
/
/
-o 4
/
o
~ 20-
/
Ey > O.66 MeV j/
4 e
/
3 16
/
)/
Fig. 43.
l2
/
a
/
Linearity of hodoscope gamma-ray 3
/
Ey>0.33 MeV a 8
/
response with stilbene detectors to fuel-pin removal from a 37-pin as-
/
s#
o sembly.
34
/
I I
I l
O I
2 3
4 Number Of Pins Removed l
l 36-l l
i
HOOO g
i g
l
[
i i
i i
e,==e e
-= e a;s e=a y
o
=
\\
d th
/ ? ~}, s -
o,'.
10000 -
hl,I
\\
q {y,, e i
fis.
't{
- i 9 00o _
, f \\[ / !
\\t /
i y/i 8
\\
Neutron scans with *He-recoil detec-E 8000-
}'
Fig. 49.
?
tors at an energy threshold of OJS f num,,,,,,,,,,,,,,,e fourpins removed in line.
f
~
}
MeV of a 37-pin assembly with up to a none l
{
2 7 000-
{
3 4
{
4 4
1 6000 I
I I
I I
I I
I I
+12
+9
+6
+3 0
-3
-6
-9
-12 Deflection (mm) 32 28
{
En > O.35 MeV
'/
{ --- En > 0.70 MeV 24
{ --- En > l.40 MeV g
/
20
,/
~
/
3 I6
,/
Y c.
5 f
/
a 2 l2
/
Fig. 30.
/
Linearity of hodoscope neutron re-unearity Test sponse with *He-recoil detectors at 8-
He Recoil Oefector various energy thresholds lofuel-pin s
4 _.,
removalfrom a 37-pin assembly.
1 I
I I
O I
2 3
4 Number of Pins Withdrawn 37
l e
Stf 5
REP 80VED LUCi f E PLUO5.#E 8L ACED CENTEN FvrL P1 9.a I (TL) E
O 6(, New e
set 6 Jut
ss 8 9 7 7. pt maa v t 3 C t te f t m *LEL PIN.L CtrE S****
MA e5000
]
e e
0 e" {
%'s eesee.
e SEf somEMOVED LUCitt PLUC5.4f PLaCED CENfE4 FvfL P t ha I (TL) E o 3 30 esev g
e 127 6 JuLv to 1977.ReacwtD CreuttM PLEL Pl%.LVC(fE C
IW' 3
V
t V,
.v.
84800
~
h V
120**
4 o
SFt 6*RE WWED LUCt'E PLucs.mFPLatt3 CrNtre ruCL PI ha I (TL) E
2.60 esew e
Ett 6 JA*
as 8 9 77 #E*Cw t D C E as t f * 'LEL PIN.Lecitt
gese, m
T u
-u u
W II*
10.
-b.
a.
5.
Id.
SE.
POSI'80% i see n Fig. 51.
Across-flats scans with Nal(TI) ofa 37-pin assembly, intact with the center pin removed.
1
\\
38 mm m
D wo o
. 1-o
SET 6*ag ogro LucifE PLycs eEPLaCE3 CENTER E LE L p t NE 802 E
0 66 Mew
o. -
set e Jav se a9e' newceto Ct=fte FLEL *ts.LLct'e 24000 h-f MA,+;
p y
w i
24004 o
5ff 5*RE40wED LLC t f E PLttC5.uf *L ACEO CENite FLFL P!
hE 802 E
t.30 Mew e
sE F e Jwt le 19'? *EuOwfD C E *e f f R FLEL * ! **. L'.C I ' E 8080 pose
-i 1
~
SEf E*REa= owe 3 LUCtff Pt.ucs.stetaCEO CEN'E4 FvEL PI hE 802 E
2 60 Mew
.e o.
set a sw.? to s977.=tasoweo CE**ff* *LEL *t%.Lucs't C
l'oe 3
O O
120e 4
_ 4 3
qN r r
_x rf
.2.
L- +
sp _y r
p:+
1 I a
mh t
- s
,,44 1
pg__
1 See e*31=
'le.
-5.
g.
.5.
10.
gs.
.oss se Fig. 32.
Across-J1ats scans with NE-102 ofa 37-pin assembly, intact and with the center pin removed.
9m o
' q' 39 D
e
. 1..
gW a
m.
f
/,.-u - w..
~
g;g*
.;f-A+ Tt * '
cg
,,cf"I,,.,.
,m.,3.e ef'"
y&eM a
Fig. 53.
Mockup of the 91-pin EBR-ilfuel assembly.
C. 91-Pin Assembly Scans Twenty-one scans of the EBR-Il fuel assembly were made at various neutron and gamma-ray thresholds, 2e in a joint expaiment with Argonne National Labo-looking through one or three flawed pins from various ratory, a 91 pin EBR-Il fuel assembly was scanned. This angles. One across-flats scan with stilbene detectors is assembly (Fig. 53) consisted of an array of wire wrapped shown in Fig. 55. The data points are the ratios of the fuel pins in a hexagonal stainless-steel can with an output of one detector scanning across a perfect section outside dimension of 5.82 cm across flats. The fuel pins, to the output of the other detector scanning across a each consisting of 3.56-mm-diam by 69-mm-long pellets section flawed in the center with a void.The three curves of 52% enriched UO fuel loaded into 0.38-mm-thick are the neutron, the gamma-ray, and the " total"(n + y) 2 cladding, were assembled into a hexagonal bundle with a scans for Ep 0.3 MeV.The neutron scan shows a clear spacing of 5.6 mm between centers. A void and a steel 3.5% difference in ratios over the diameter of the missing dummy replaced two pellets in each of three " flawed" pellet (detector sensitivities were not normalized, so the pins. Figure 54 shows the positions of the flawed fuel base-line ratio is not unity). It is also clear that the pins in the bundle and the layout of the void and steel gamma-ray scan does not reveal the missing pellet. As dummy in a flawed fuel pin. The flaws could be aligned could be expected, summing the gamma ray and neu-with the field of view of the hodoscope. The assembly tron data serves only to dilute the neutron data, therby was mounted on remotely operated precision rotation degrading the sensitivity of the hodoscope.
and vertical positioners. By rotating the bundle, scans Other results of these experiments have been reported could be made through the three flawed pins aligned on a H detail elsewhere.20 22 The tests showed that it is major diameter of the hexagon or through only the
,oss. ole to detect a pellet-sized void in an assembly of center defect either across flats or across corners. By this size by means of a fast-neutron hodoscope, that using two vertical channels of the hodoscope, scans hodoscope image resolution improves with increasing could be made simultaneously through a perfect and a neutron energy (from I to 3 MeV), and that gamma-ray l
faulted section of the assembly. Stilbene detectors and self-imaging of an assembly of this size and complexity is I
llornyak buttons were both used.
inferior to neutron self-imaging.
40
CROSS SECTlON OF SPECI AL E 8R ll 91 F UE L PINS 91 PIN FUE L ASSE MRLY 88 MAPK 1 A NO E8138682 C
[PLUS 3 SPECIAL FLAWED PINS
'. x x x v3hh ~ 6' t
n m
F L AWED FUE L PINS LOCATED C.
IN POSITIONS 1. 4. AND 6 OF t
91 PIN ASSE MfslY Fig. St.
Locations and details offawed pins in the 91-pin EBR-iltest assembly.
L
)I
,3 y,,..
I FUEL ZONE
,..

L n - r, _
' /// '
p'f
E L :4 a
^
O,*.'. F U E L '
VOID
,:. F UE L; :.*j SST F
's $,0,U H
s
" ' :.4 %. +
,/////,
i*
\\
CLADDING 27 in 30 m,,
E ACH SE CTION WAL L THICKNESS TUBE 304 SST t
0 93
-(c) o y /72 o"
o',l',o
}
0 91 g-g o,
_g....-.
,y_o..
.o_o....__
q g
.o
-o,,
o o.
o o-n o
089-
.,o 0 87-9 0 89,-
(n + y),/(n + y)2 J
a
[9 0 87 o~
("2'u
]
FiK SS-on_
o g
-vo
" """~"
.o o..
o Across-pats neutron, gamma, and to-o
" ~ o "G "f tal scans with stilbene detectors of a g O 85-o T
g 1
~1 pawed 91-pin EBR-il assembly. Data 8 083
5 mm==
4 Psn d<0m b points are the ratios of the output of n,/ng = 0 BI6
,/ p n,ing =O 793~
one detector scanning across a perfect O 84-n,/ng l'
'\\
1 o
'o,a iI section to the output of the other i
t detector scanning across a section O82-L~
t
'Q >/
o")
pawed in the center with a void.
o o
o o o "I..
0 80 -
1 g-
.gg..
Peo _, o. o
_o.._;
.n o...
no
,4>
,,41.
,il l
0 79 -
o o
]
j' o"
0 76 -12
-9
-6
-3 0
3 6
9
-12 Scanner Position (mm) 41 D**D DJ0303 M omI o A\\UJ)UJ\\lb
D.127 Pin Assembly Scans pin, we calculate that the observed count reduction was caused by. tb removal of 2.94 g of anU from the With stilbene detectors biased as in the 37-pin as-hodoscope field of view, which is comparable to the sembly scans, a 127-pin assembly with FFTF-sized pin established requirements for mass resolution for sin-dimensions and spacing has been scanned across flats gle-subassembly tests.2 and across corners of the hexagon, intact and with pins Figure 58 shows a scan across r. corner of the removed at various depths in the assembly. Shown in assembly. Such a scan exhibits definite maxima and Fig. 55 are across-flats scans of the assembly intact and minima because the scan direction is perpendicular to the with the center pin withdrawn.The count time was 200 s fuel-pin rows. Since the distance between fuel rows (6.29 for each data point at 5 mW/g "U operating power.
mm) is less than the horizontal resolution (to 2
(Note that counts /200 s at 5 mW/g "U is equivalent to half-maximum intensity) of the hodoscope slot (7.14 2
counts /(J/g "U))--a fortuitous choice!) To better show mm), the hodoscope slot actually " secs" more fuel when 2
the effect of withdrawing the pin, the ratios of pin-out to the slot is centered between two rows than when the slot pin in counting rates are plotted in Fig. 57. The is centered on a row. As result. a count minimum occurs fast neutron signal from the fuel in the center pin, when the hodoscope slot is pointed directly at a row of averaged over the diameter of the pin,is approximately pins.
2% of the total count rate. Using the known vertical Figure 58 also shows the efTects of withdrawing the resolution (to half maximum intensity) of the hodoscope central pin or the corner pin nearest the hodoscope. As 2
slot (19.05 mm) and the known "_U content of the fuel discussed _ above, some of the dnTerence in response i
I i
i
[
l i
16578 72 a Center pin in 71%
i Center sin out 69 -
Il 70-)
e II gig
{ I I ggg se E7 > 0 33 VeV
+
4 42yy y
48 I't t ; i i i I I e
I g3 40 -
E7 > 0 66 VeV 39-38-
_ 37 -
[,
4 9
3 29 o**
Fig. 56.
n ' e * '
e *IE I llt a'
Across-flats scans with stilbene detec-e
? 27 y
o 26 E > l 3 MeV tors ofa 127-pin assembly, intact and n
2s with the centerpin withdrawn.
'o 4
x i4 f
.,,J 3f_'tiii;i,;
f.
i En > 2.2 MeV r
l l
1 I
i l
i I
l g
+15
+12
+9
+6
+3 0
-3
-6
-9
-12 i
Deflection (mm)
F
I e
I i
l 8 = 1.0 t 0.4 %
l 1
i i
1.02 -
i.OO-
}
} } { { -g
,T
~I ff 5~
O.98-3lt I 0 96-E7 > O.66 MeV 094 -
1.02-
-S = 1.010.3 %
II-100 3 _ _ _ _ _ g_7 _,
I T taf
}
O.98 -
_g j O.96 Ey.> 0.33 MeV
, 0.94 c
c
_3 2.020.7%
02 n -
3 0
9 g
1.00-
,rn n
0 98-is n
Fl 57-En > 2.2 MeV
~
f i
Data of Fig. 36, expressed as ratios 0 94 ofpin-out to pin-in count rates.
O 92 -
B = l.810.5 %
{-
102 -
h 1.00 -
I I
I.
t 0.98 f
AI TT fI
{f A
f
' I 0 96-En > I.3 MeV 094 1
I I
f l
I I
I I
+12
+9
+6
+3 0
-3
-6
-9
-12 Deflection (mm) 33400 i
i i
i i
l i
i i
i 3I,000
~
10
Nij i
f 4 6 a,=
,1 29,000-I g6
,0
},, {,'
Mg g,g#g+ r_
4
' Le,
A,4 o
8 o 27000-(#
e-
)
3 25,000-t 127 pins, none withdrawn O
_ A Center pin withdrawn
_ = Near corner pin withdrawn I
I I
l 1
I I
2 l'O00 -15
-12
-9
-6
-3 0
3 6
9 12 15 Deflection (mm)
Fig. 38 Across-corners scans with stilbene detectors biased at E, > 13 MeV of a 127-pin assembly, intact and with the center or near-corner pin withdrawn.
43 S {\\
D**D
D o
oc o
.I00 g
i i
g l
l l
1 i
.E G
T R 099-O y3-f
i T
I O
h 090 f
^
o 5-Il1 a: 0.97 - I g
{ Aeross flots i
E Across corners s 0.96 8
Collimator I
I I
I l
I I-I I
0.95 -40
-30
-20
-10 o
10 20 30 40 50 Distance From Center Of Miss ng Pin (mm)
Fig. 39.
Ratio ofpin-out to pin-in count rates ofa stilbene detectors biased at E, >
IJ MeV as afunction of roidposition in a 127-pin assembly.
t between the two pin locations is due to the power inside wall of the 127-pin test chamber. This 20-cm-long distribution within the assembly, it is evident, however, casing can be remotely raised from and lowered into the that the response of the hodoscope to a void is not field of view of the hodoscope.
independent of the position of the void within the test Table III lists the results of across-flats scans with assembly. Figure 59 shows pin-out to pin-in count-rate stilbene detectors of the 37-pin assembly, with and ratios for across-flats and across-corners scans of the without the casing and with and without a center void, assembly with a single missing pin at various depths The steel casing causes a loss in total signal and within the bundle. The data, taken with a stilbene background of about 30% for fast neutrons and nearly detector biased for E, > 1.3 MeV, show that the total 50% for gamma rays. More important is the fact that the count rate reduction for a single-pin void in a 127-pin casing decreases the mass resolution by ~33% for fast assembly varies from 3% for a void at the near edge of neutrons and by ~48% for gamma rays. Also listed in the assembly to 1% at the far edge. These data, which Table III are the mass-resolution decreases for 25-and have been corrected for the measured power distribution 50-mm-thick steel capsules calculated by assuming that within the assembly, show the need for a detailed static the effect varies exponentially with wall thickness.
hodoscope study of every large test assembly before the Figure 60 displays across-corners fast-neutron scans destructive experiment is run. The need for of the 37-pin assembly, with and without the casing and three-dimensional data, as from crossed hodoscopes, is with and without voids at the center and far corner. As also evident.
discussed above, a count-rate minimum occurs for this orientation when the collimator is pointed directly at a row of fuel. The degree to which the details of the scan E. Effect of Test-Capsule Wallihickness are degraded by the casing is obvious. Similar gam.
ma-ray scans with energy thresholds of 0.33 and 0.66 Of considerable importance to experiment planning is MeV are shown in Figs 61 and 62. As expected, the the effect on the fuel image of the wall thickness of the degrading efTect of the steel on these gamma-ray images, capsule used to enclose the test assembly. A tradeoff particularly for the lower-energy threshold, is more may be necessary between safety of the experiment and pronounced than for the neutron images.
amount of image degradation that can be tolerated -
Because the effective thickness of intervening steel is within experimental objectives. To study this problem, greater for scans away from the center line of the the 127-pin assembly was replaced with a 37-pin as-cylindrical casing, it was felt desirable to repeat the sembly surrounded by a 21 mm-thick steel casing. The above experiment with pins removed from the side of the steel occupies the space between the 37-pin grid and the assembly, rather than from the center. The neutron and
- 44 t i1 lf$
k m
-w q
TABLE III EFFECT OF A 21-mm-THICK STEEL CASING ON HODOSCOPE IMAGES OF A 37-PIN ASSEMBLY E > 0.66 MeV Ey > 0.33 MeV E, > 2.2 MeV E, > 1.3 MeV y
Count rate' ofintact assembly, without easing 38 150 63 200 9 500 19 000 Count rate' of assembly with center void and 35 800 59 600 8 900 17 900 without casing Fractional count-rate decrease caused by 0.061 1 0.003 0.056 0.002 0.065 i 0.006 0.058 i 0.005 center void without easing Count rate' ofintact assembly with 21 mm-thick 20 250 34 400 6 300 13 900 casing Count rate' of assembly with center void and with 19 600 33 400 6 070 13 300 D
21-mm-thick casing Fractional count-rate decrease caused by center 0.032 1 0.005 0.029 i 0.003 0.037 1 0.007 0.042 i 0.005 void with 21-mm thick casing Mass-resolution decrease caused by 21-mm thick 48 i8%
48i5%
38i 8%
28 i 4 %
casing Calculated mass-resolution decrease caused 54 %
54 %
43 %
32%
by 25-mm-thick casing Calculated mass-resolution decrease caused 78 %
78 %
68 %
54 %
by 50-mm-thick casing
Count rate is in units of counts per 200 s at 5 mW/g "U.
2 8
1 i
i i
i i
i i
i i
i I
i lA 20
ns**
4 i
18-og.,$ 8g4 4,
8%
[
4 e
ee
g o
s
?ip I
....e g 14 4.tapp,f,.......q *j a " '
h e.r ** 3 3
./
v
.c "......
u
12m,,,,,7 r
r e
,s
$10 g
p k
56 E
a En > l 3 MeV m
[
+ 37 Pin bundle, no steel o6 x
o e Center and f or-corner pin rnissing. no steel f4 S
37 Pin bundle, 21mm steel jocket E
Center and f or -corner pin missing 21 mm steel jocket-1 g
o2 0
i i
e i
i i
i i
i l
i i
i i
30 27 24 21 18 15 12 9
6 3
0
-3
-6
-9 15 Scanner Position (mm)
Fig. 60.
Across. corners neutron scans with stilbene detectors (E, > 1J MeV) ofa 37-pin assembly, with and without steelcasing and with and without volds at the center and at thefar corner.
72 A
A 2 64-
,1... ***'. * *
w n
se e
o
,,,,....,,a.,=,.*
7..
g Se g48 g'
I b
~42 1
1403..*
4 o$
i b 32 II 93,3 14
'o 24-k *' / ",,.....o..&.seensy9 8
_ N 3
C s_
f E7 > 0 33 MeV f 16
+ 37 Pin bundle no steel g
Center and for-corner pin missing, no steel U 8 37 Pin bundle,21 mm steel jocket a Center and for-corner pin missing,2imm steel jocket I
0 30 27 24 21 18 15 12 9
6 3
0
-3
-6
-9 15 Scanner Position (mm)
Fig. 61.
Across-corners gamma-ray scans with stilbene detectors (E, > 033 MeV) of a 37-pin assembly, with and without steel casing and with and without voids at the center and at thefar corner.
44 i
i i
i i
g i
i i
i 40-
~
.A
, 36-og 4p,, j.e.gV. * ** kp *.s f
3 3
A5 32 8
5 f'
j 5 28-
[ o.
q t
e
)
" 24
/
h
'.e
n m*4,,cQs%,..m.,+, -
90- -
o.
E
? 16 8
f
-\\
12
'd Ey > 0 66 MeV g
h
~
+ 37 Pin bundle, no steel
~
v 8
Cent y ond for corner pin missing, no steet
[
4 e 37 Pin bundle,21 mm steel jocket m Center and for-corner pin missing,21mm steet jocket_
O I
I i
1 1
I I
I I
I I
I I
I 27-24 21 18
-15 12 9
6 3
0
-3
-6
-9 15 Sconner Posit.on (nm)
Fig. 62.
Across-corners gamma-ray scans with stilbene detectors (Er > 0.66 MeV) of a 37-pin assembly, with and without steel casing and with and without voids at the center and at thefar corner, gamma-ray backgrounds, as sensed by the scanning count rate from the fuelis reduced by more than 50%
hodoscope when there is no fuel in the test region, are not while the background is undiminished. These efTects, for uniform across the ent. ire field of view. At the edges, both neutron and gamma-ray imaging, are not ap-3 i
. background levels are nearly equal to signal levels from preciably dilTerent from those for scans of the assembly he center of the 37-pin assembly. For this reason and
center, i
because the major diameter of the 37-pin assembly is Figure 63 also shows that the test assembly causes a greater than the field of view of the hodoscope, it was 1015% increase in the neutron background from the test necessary to displace the test section 19 mm to one side region. This is attributed to a pow:r-distribution shift of the axis of the through-slot in PARKA. This was relative to the flux at the ion chamber being used as a accomplished by removing the test section and its liner power monitor and controller. The monitoring chamber from the reactor core, shifting seven PARKA fuel rods, is external to the reactor core.
and replacing the test section at its new position.
Figures 63 and 64 show across-corners neutron and gamma-ray scans with stilbene detectors of the ofi-center F. Studies of Clad Motion Detection 37 pin assen.bly with and without the steel casing and with and without voids.
As discuswd in Sec. III.B.2, we had identified the The neutron scans of the empty test region show that high-energy rad!Mion in the gamma-ray spectrum ob-the casing does not appreciably affect the fast-neutron tained with a BGO detector as capture gamma rays from background in the center of the test region, although it iron. This spectrum (Fig. 34) was obtained with the does somewhat flatten the background across the region.
38-mm BGO detector mounted in front of a collimator At the side edge of the assembly, the steel reduces the slot pointed at the edge of the 37-pin assembly and with neutron signal from the fuel by 40% and thereby reduces the simulated test capsule in place. The use of these
. the S/B ratio and the mass resolution. The etTect is more capture gamma rays to detect clad motion is hindered by pronounced for gamma rays. We find the net gamms-ray the presence in the experimental setup of various steel 47
'F,
' ' U i
5.
r%J,
+-
30,,,,,,,,,,,,,,,,,,,,,,,,,,,,
~
28 - ( A) Bare Fuel Pin Assemb+y
,(8) Assembly With 22mm
- Neutrons, En a 1.3 MeV Steel Casing 26 e Full ossembly
[ o Pin I removed
[
[
24
- e Pin 2 removed 22
'Y f f [A
~
_ A Empty test section,-
i 20 s'
Q "'
n
/
go '
w/
pv
's-18
/
x_.-
E 16 f
v
O
' ( ' " -
N y
e j
Cpv
~
Opg~
h I4 l
e.fr - y p(_
UC
p 12 t
(r.p(uo-y p.
tn p,3 p, -,
s..
(,
z^,.
o 10
\\o 7
e,
[8 - **
V f.
A **e g
C d ] p'((3 v-g e
t
(
m-g j
6 N
^^^^^^-
.y j
4 I
.~ U
'\\\\
2 Collimator resolution (FWHM)
'I''''
_' ' ' ' ' ' ' ' ' I ' ' ' ' ' ' ' ' '_
O '24 18 12 6 0 6 12 18 24 24 18 12 6 0 6 12 18 24 Scanner Deflection.' rom Center Of Slot (mm)
Fig. 63.
Across-corners neutron scans with stilbene detectors of the ofcenter 37-pin assembly.
120
,,;,t
- ( A) Bore Fuel Pin Assembly
- (B) Assembly with 22 mm 5 omma rays, Ey > 0.33 MeV G
Steel Casing 10 4 e Full assembly 96-Pin i removed
~ + Pin 2 removed
/'-
/
[ A Empty test section /~_
[
/
[
7 80
[
~
)
O ~'
g
/
Q 48 l Cg'@n ox o' o s:
a2 [
[
fz@gO g:
/@
p C
f64 :
g*
%'C
~
fO OfO-
~
~
g
(; p (] p
'h.O '- O "' O
~
~ - -
, wg7" Q u p ' p-
/e u-
' Man y a
$ s p-p-
g 32 -s,1 y-3
\\.
-m s o)(J u
24 g
s-4 (a
N
C,A,
n6 bp,,, j'
- Q...e.............,,, g'-
-s N
./
B
I''''
''''I''''
O ' ' ' '8 24 1 12 6 0 6 12 18 24 24 18 12 6 0 6 12 18 24 Scanner Deflection From Center Of Slot (mm)
Fig. 64.
Across-corners gamma-ray scans with stilbene detectors of the ofcenter 37-pin assembly.
48 0 *
D) " ~D)l]Ju Ninto
%M o e in o Juu
objects (other than the cladding itself). These objects where 4(x,E)is the neutron flux of energy E at position x, include the collimator and the test capsule.
oJE) is the radiative capture cross section of iron for By moving the detector to one side of the collimator neutrons of energy E, and p(x)is the density ofiron at slot, it was determined that ~75% of the observed iron position x. If one makes the simplifying assumptions that capture gamma rays originated in steel near the detector, s(x)is constant and equal to S throughout the steel and that is, from capture of thermal neutrons in the steel of that no other material is contributing to removal of the the collimator near the detector end. By using the capture gamma rays, one may integrate Eq. S to obtain 12.5-mm BGO detector, this high-energy background was reduced to about 23% of the total signal. It is n - (s/I)(1 - e # )
(7) apparent that,if one is to use a hodoscope collimator to obtain high-energy gamma-ray self-images, either the which approaches a saturation value R. = S/I for an entire collimator should be shielded from thermal neu-infinite slab. Furthermore, the signal AR from a small trons to reduce the production of iron capture gamma additional thickness AT of steel in line with the col-rays in the collimator, or the 20 cm or so of collimator limator, as for instance in the test region, is closest to the detectors should be made of a high-density IR AT exp(-IT). For a capsule with 4.5-cm-thick material with a low neutron-capture cross section, such walls (total steel thickness of 9.0 cm), R = 0.88 R and as bismuth.
AR/AT = 0.028 R cm-'. A l-cm-thick clad blockage Test capsules for use in the upgraded TREAT reactor in the test assembly would therefore increase the number are expected to have steel walls with thicknesses of up to of 7.641eV gamma rays reaching the detector by only 45 mm. Thus, a collimator slot viewing a test assembly
~3%. Neutron flux depression through the test-section enclosed in such a capsule will also be viewing radiation and capsule walls would be expected to reduce Py and from at least 90 mm of steel. The mass-removal coefli-hence the differential sensitivity even further, cient for 7.641eV gamma rays in steel ' is 0.030 cm /g, To investigate the effect of neutron spectral shifts and 2
2 which is equivalent to a linear cross section of 0.236 flux depression within the test section, hodoscope iron cm-'. The signal from steel in the test region will be capture gamma-ray intensities were calculated from attenuated by passage through the capsule to 16-group ONETRAN calculations of the 37-pin as-exp(-0.236 X 4.5) or 35% ofits original strength.
sembly and steel capsule in PARKA. The calculated The steel capsule, however, not only attenuates the intensities t re compared in Table IV with measurements 7.64feV gamma-ray count rate from steel in the test of high-energy gamma rays from the test region with the region, but also contributes to the total 7.64feV count collimator viewing various thicknesses of steel. The rate. If the steel capsule is regarded as a uniformly collimator viewed the center and the edge of the test radiating source of 7.64feV gamma rays (and if absorp-assembly, first with only the test-section walls in place tion in the test assembly is ignored), the count rate R at (total steel thicknesses of i1.2 and 12.7 mm) and then the detector from gamma rays originating in the capsule with the 21-mm-thick capsule in place (total steel thick-is nesses of 54 and 98 mm). Pulses from the 12.5-mm BGO y
detector corresponding to a gamma-ray energy interval R=
s(x) e
dx (5) f 6.2-8.1 MeV (with noncollimated background sub-tracted) were counted for 1000 s at a power level of 5
<0 235 mW/g U. The sariation with steel thickness of the where T is the length of the path through the steeljacket, ratio of measured counts to the ONETRAN prediction s(x) is the 7.64teV gamma-ray source strength at shows that the nonlinearity of detected capture gamma position x, and I is the linear removal cross section for rays is even greater than predicted by Eq. 5 with s(x) the gamma rays. He source strength term s(x)is of the derived from the ONETRAN calculations. With the test form assembly enclosed in a 45-mm-thick steel capsule, 1-cm-thick clad blockage would cause an increase in I,
4(x,E) o (E) dE (6) count rate of capture gamma rays of the order of only c
1 %.
0 49
TABLE IV detector of a 37-pin test assembly and of the empty test region, both with and without the 21-mm steel capsule.
OBSERVED AND CALCULATED Photons with energies >3 MeV produce images with 7.6-MeV GAMMA RAYS FROM A 37-PIN ASSEMBLY p rer S/B ratios than those with energies of I-3 MeV. It AND STEEL JACKET IN PARKA is therefore desirable to use an upper-level, as well as a threshold, discriminator for imaging fuel motion with Steel Observed Calculated Ratio of gamma ray detectors. We also conclude that Thickness Net Counts in Relative Observed to non energy-dispersive detectors, such as gam-ma-sensitive fluors used for direct imaging of (mm) 1000 s Net Counts Net Counts self-radiation from a test assembly, should, if possible, have an energy-dependent response tailored to maximum I 1.2 2507 446 5.62 sensitivity in this energy range.
12.7 2817 665 4.24 54 3443 897 3.84 98 4162 1267 3.28 H. Summary Table V is a summary of results obtained from G. EITect of Energy Selection on Gamma-Ray Imaging hodoscope scans of various-sized test assemblies with a of Fuel missing pin. The ability to detect one missing pin in a 127-pin assembly is established, but appears question-We have shown that selfimages of fissioning fuel may able for larger assemblies. It should be emphasized, be obtained with gamn a-ray detectors. We have also moreover, that this static test of hodoscope resolution by found that the S/B ratio of such gamma-ray images can no means establishes the feasibility of detecting the loss be enhanced by appropriate selection of the gamma-ray of this quantity of material from a test assembly under energy interval used to form the image. This is indicated dynamic conditions, but rather establishes an upper in Fig. 65, which is derised from pulse-height distribu-performance limit for self-radiation-imaging systems.
tions from actms flats scans with the 12.5-mm BGO I
I I
I I
I I
I 7
Casing Thickness 6
4.5 mm O
26 mm g
5 2 8
_ E#
~
7
~
G
-h :;
%E 3
Q J
f Gamma-ray S/B ratio for a 3" oin Ry Q[f
.}
assembly as a function of detected
[
/
energy.
f, g
4
! ' g a
j:g
. g s
?A A h im n
1 O
I 2
3 4
5 6
7 8
Photon Energy (MeV) 50 i
TABLE V FRACTIONAL DECREASE IN HODOSCOPE IMAGE INTENSITY CAUSED BY CENTER VOID IN VARIOUS-SIZED TEST ASSEMBLIES Fractional Decrease Caused by Center Void Number of Pins in Assembly E, > 1.3 MeV E, > 2.2 MeV Ey > 0.33 MeV Ey > 0.66 MeV 1
0.33 0.43 0.45 0.54 37 0.054 0.070 0.046 0.059 9l' O.032 0.00 0.00 127 0.018 0.020 0.010 0.010 91 pin EBR-II assembly.
V. CODED-APERTURE IMAGING STUDIES at a power level of 1500 W. Inferior images, useful for setup purposes, could bc obtained by exposing the Coded-aperture gamma-ray imaging of LMFBR fuel Cronex-2DC film for 10-20 min with type TI-2 screens.
assemblies under test is being developed by SLA in Lack of funding has caused suspension of the experi-conjunction with experiments to be performed at the ments, but the following conclusions were reached.
ACRR. In support of this program, a coded-aperture o PARKA can be used effectively as a driver reactor imaging system designed and fabricated by SLA was for static coded-aperture imaging experiments.
installed in the PARKA shielded instrument-room wall e A gamma-ray pinhole clearly images a 37-pin in place of the hodoscope collimator. Top and side views assembly in the presence of reactor background.
of the installation are shown in Figs. 66 and 67. Images e Clear images of a 37-pin assembly have been of a test assembly in PARKA are projected through the reconstructed from shadowgrams made with an viewing slot and coded aperture onto x-ray film at the oft-axis Fresnel zone-plate coded aperture.
back of the aperture shield assembly.
An exposure with a sheet of NE-102 plastic scin-Exposures of the 37-pin assembly were made using tillator and Eastman Kodak Co. Tri-X film pro-4 Eastman Kodak Co. type AA and XR-5 and E.I.
duced ca image; this indicates that the bulk of DuPont De Nemours & Co. Cronex-2DC x-ray Clm radiation forming the coded-aperture images is with tantalum sheet and type TI-2 organic intensifiers.
gamma radiation, rather than neutrons.
Films were machine-developed in the LASL Group M-1 Examples of pinhole im ges, Fresnel zone-plate shad-x-ray la5 oratory, owgrams, and reconstructed images of the 37-pin test The best quality photographs of Fresnel zone-plate assembly were exhibited at the 1979 American Nuclear images were obtained by exposing type AA film with Society-European Nuclear Society Meeting on Fast i
type TI 2 screens for 10 000 s with the reactor operating Reactor Safety Technology in Seattle, Washington.24 51
N CONCRETE
'UN 8 LOCK WORTAREO SLOCK E XISTING INNER LEAD AND EDGE POLY (DIMS UNCERT AIN '
/ UNCERT I CONCRETE REAR STEEL SHIE LD COLLIMATOR PARKA j
\\
ACTfVE g
'y 7_
.j POLY
~ -j FRONT CORE i
l
' St r 3
M
~
POtv N
s Sy T 2, CODED SHIE LD BLOCK g
APE RTURE
[
ALTE RN ATING COO RETE LOCATION LEAO. POLY AND,
ALTERNATING STE E L PLATES LE AD. POLY AND CONCRETE UN STE EL PLATES BLOCK MORTARE0 BLOCK m-Fig. 66.
Top view of the SLA coded-aperture imaging system.
CORE
(\\
CONCRETE UNMOR BLOCK TARED INNER BLOCK EDGE PARKA REAR
.UNCE R TAIN ACTIVE SHIE LD CORE E XIST.
CONCRETE E X 85 7. Pb,
, COLL ATOR
~ LY,Fe POty FRomy l
l g
P g
Sp r
_,_ ] i
.I
)-----'
L Y
[ SE T y -=O,
poty EL$'
BLOCK 8
K CONCRETE ALTE RN ATING I
LE AD, POLY CONCRETE URE AND STE E L BLOCK LOCATION PLATES ALTERNATING
^
STEE A ES Fig. 67.
Side view of the SLA coded-aperture imaging sprem.
52
Exhoust Tube Kovar Seal 9" '
ased Coatmg h
/
N S
Y I
i
- t n
___-v---
r p
Plotinum Electrode
- 5 mm dio Fig. 68.
Cross-sectional view ofa '*Ufission chamber used as an in-core detector, A detector of this type has also been used as an auxiliary power-level monitorfor all hodoscope tests.
VI. IN-CORE DETECTOR STUDIES cable to a RIDL model 27501 discriminator in the control room.
PAR'KA has been used for studies of in-core detec-A pulse-height distribution from one of the U fission 23:
tors,in particular "U and 23sU fission chambers. These chambers is shown in Fig. 69. The two-humped fis-2 chambers, ~S mm in diameter, were fit into the stain-sion-fragment portion of the distribution provides a less-steel fuel-pin cladding. Figure 68 is a cross-sectional convenient check on the performance of the detector and view of one of thcsc detectors. The 2"U detector is associated electronics. For counting, the discriminator sensitive to neutrons with energies above 1 MeV and was set at the minimum between the fission distribution thus. responds primarily to neutrons from the test and the low-energy alpha and noise pulses, at the point assembly. In contrast, the 2"U detector weights the indicated by the cursor. This particular chamber has lower-energy neutrons and is therefore more sensitive to been used for three years as a monitor of reactor power neutrons leaking into the test assembly from the driver.
without noticeable deterioration in performance.
23:
As a result, removal of a fuel pin adjacent to a 0 and a Measurements on a 37-pin assembly were made with 2"U fission chamber produces an ~2% decrease in the the fis' i chambers located in an empty fuel-pin casing count rate of the za:U fission chamber and an increase of that replaced the center fuel pin. Detector response was similar size in the count rate of the "U fission chamber.
determined as fuel pins were removed from each of the 2
The chambers, which were fabricated at LASL, were three hexagonal rings of pins surrounding the detectors.
found to be extremely reliable, stable, and easy to make.
The results summarized in Table VI show that counts Uranium-metal coatings of ~l mg were evaporated onto from the 2nU detector decrease as the fuel pins are the platinum anodes.* Most of the joints were removed, whereas counts' from the 2"U detector in-soft-soldered, after the interior of the chamber was crease. The "U/2nU count rate ratio is a more sensitive 2
cleaned with abrasive and solvents. After assembly, the measure of fuel loss than is the count rate of either chambers were pumped out through a liquid-nitrogen detector separately. For these experiments, entire fuel trap to forepressure only, backfilled to 100 psig with a pins were removed; it is expected, however, that similar 90% argon-10% methane mixture, and sealed off. One results would be obtained if short lengths were removed chamber was prepared with an node bearing n Cf to in the vicinity of the 10-mm-active-length fission detec-facilitate testing the characteristics of the fission cham-tors.
bers and associated electronics.
Figure 70 shows the response of fission chambers The chambers are operated at 200 V. Signals are fed located at the center af the 37-pin assembly as fuel pins at the reactor to a preamplifier of LASL design (model are removed progressively from the inside toward the 224 QN) and then to a Radiation Instrument Develop-outside. As pins are removed, both detector responses ment Laboratories (RIDL) model 27001 amplifier. The change by almost a factor of 2. For a true fast driver, output of the amplifier is fed through 300 m of RG63/U however,it should be noted that the diagnostic capability of in-core fission detectors would be limited because
We are grateful to John Povehtes of LAsl Group CNC Il for preparation there would be little change.m the neutron spectrum, and or idene mms.
53 5
l
.;e
~
ls,$.R.O
?y p
Q Fig. 69.
/
. x Pulse-height distribution from the *'*U fission
'n chamber.
W l
l i
TABLE VI IN-CORE DETECTOR RESPONSE TO FUEL-PIN REMOVAL Relative Response l
Pins Removed 23'U Detector 23sU Detector 235U/2'8U Ratio' Center pin 1.000 1.000 1.000 Center pin and six pins from inner ring 1.098 0.892 1.23 Center pin and six pins from middle ring 1.093 0.928 1.18 Center pin and six pins from outer ring 1.058 0.970 6.9
'The absolute magnitudes of the U and 23sU detector responses are accurate to only 120% since the 2
masses of the fission-chamber coatings are not accurately known.
20 i
r80 ie-d60
., 16-E 14 0 g l 4 g(
El 3 83'U Detecter a:
s
-N
-f20 a
12-
{o 3
N
-ico g Fig. 70.
g Io-
\\
g In-core detector response tofuelf n remoralfrom j
N
- eo l
[
i a 37-pin test assembly.
oe-y q
Q j
'4
-60 g
[O6-23'U 0etector
%'s -
of M
O 10 20 30 Numter of Fuel Pins in the Assett,fy I
i j
l l
l b
TABLE VII IN-CORE DETECTOR RESPONSE TO REMOVAL OF COMPLETE FUEL-PIN RINGS Relative Response Pins Removed 2"U Detector 23:0 Detector 2nU/2 sU Ratio' Center pin i
1 39 Center pin and inner 6-pin ring 1.12 0.88 50 Center pin and inner 12-pin ring 1.17 0.86 53 Center pin and inner 18-pin ring 1.12 0.89 49 2
23:
The absolute magnitudes of the "U and U detector responses are accurate to only i20%
since the masses of the fission-chamber coatings are not accurately known.
i hence little difference between the 2"U and 23sU re-that a single burst of Godiva, a prompt-burst reactor sponse, as fuel pins are removed.
housed in the same experimental area as PARKA, Table VII shows the response of the centrally located induced radiation levels in PARKA comparable to those fission chambers to removal of complete rings of fuel induced by a calibration run of PAhKA for 10 min at 55 pins. The first ring contains 6 pins, the second 12, and W. As a consequence, we became interested in the the outside ring 18 pins. The fact that the fission possibility ofintentionally pulsing a suberitical PARKA chamber responses are relatively constant for these three with Godiva to induce short, inte ise transients in a test changes indicates that the filtering efTect of a complete assembly without operating PARKA in a supercritical ring of pins is independent of the diameter of the ring.
mode. Experiments were initiated to investigate the Our results indicate that in-core detectors can provide potential of this technique.
diagnostic information in multipin STF tests. It has been PARKA power had been calibrated by measuring the proposed that one or more fuel pins be replaced with core fission distribution with uranium-aluminum flux 2
" hardened" channels enclosing a line of alternate "U wires and then determining an absolute value of fis-2 and "U fission chambers cooled by 11owing argon gas.
sions/g 2"U at one point in the core for some desired The "U detectors would monitor the local neutron flux operating level and time. Standard procedures exist at 2
driving the fuel pins and the 2nU detectors would sense the facility for this type 'of calibration. From these the proximity of fuel. For the high power levels encoun-measurements it was determined that normal operr4ng tered in STF tests, it would be preferable to measure power for hodoscope scans wtih llornyak Buttons is 550 ion-current signals as a function of time rather than W and that a typical Godiva burst resulted in an energy count rates.
release of 30 kJ in PARKA.
Bursts with Godiva were made at two distances from PARKA and at two PARKA reactivities. Flux-wire VII. TRANSIENT OPERATION OF PARKA activations at the center of Godiva were compared with those at the test-assembly center and in the PARKA The hodoscope scans and in-core detector tests were core for each burst. Table Villlists the relative fissions /g characterized by long count times with PARKA operat-2"U for these locations.
ing at low power. Such an operating mode, which may be As expected, the variation of PARKA yield with adequate for testing concepts involving nuclear counting distance from Godiva is seen to be less than the instruments, is not satisfactory for imaging experiments inverse-square law, because of the contribution from involving such concepts as the use of coded apertures.
room-scattered neutrons. The yield varies inversely with Almost by accident, we learned that PARKA might be the subcritical reactivity. Presumably the prompt yield used as a high-level, short-duration source. It was found during the short part of the transient varies inversely with 55
. TABLE VIII PARKA YlELDS FROM GODIVA BURSTS Relative Yield 2
- (fissions /g "U)
PARKA reactivity = -5.7 $ and PARKA reactivity = -5.7 $ and PARKA reactivity = -3.6 $ and Location PARKA-Godiva separation = 5.5 m ' PARKA-Godiva separation = 2.3 m PARKA-Godiva separation = 2.3 m Godiva center I.00 1.00 1.00 Test-assembly center
' O.0043 0.014 0.023 PARK A core 0.0078 0.026 0.045 the reactivity interval from prompt-critical. This behav-30 i
i i
i ior is consistent with a PARKA pulse having a sharp leading edge followed by the equilibrium-mode 20-prompt-neutron decay. Pulse shapes plotted in Fig. 71 show this to be the case except for an initial transient.
These curves were obtained with a fission chamber
{
\\
located in the PARKA core and operated in the current 0 10 {'\\
T mode.
$ 8-N PARKA kinetics are well known from measurements
{7
'N and calculations on PARKA and on similar systems
g 4
\\
during the Rover program. Prompt-neutron decay I
\\3.6 8 suberiticar measurements on PARKA with the Rossi-a technique j
'~g g
give a value of 40 ps for neutron' mean lifetime t. The 3
3 N s 5.7 8 suberiticci
\\
computed effective delayed-neutron fraction yp is
-2 0.0074. Hus, the prompt-neutron decay constant at
~
s N
delayed critical coc is given by
'N[
DC = -(1 - K )/T = V8/T = -185 s-1 (8) a p
f I
I
,o o5 io L5 20 25 where K,is the prompt-neutron multiplication factor. At Time Following Godivo Burst (ms) other suben..tical reactivities the decay constant is given by Fig. 71.
PARKA response to Godiva bursts at in dWerent a=a EE (9)
PARKA reactivities.
DC p where AK,is the reactivity interval from prompt critical.
2"U is under the prompt-neutron pulse and 5 X 10
For the PARKA reactivities at which the measurements fissions /g "U occurs in the first millisecond, correspon-2 were made, -5.7 and -3.6 $, we obtain values for a of ding to an average power density during this interval of
-1240 and -850 s-', respectively. Following initial 1.5 kW/g 2"U at the center of the test assembly.
transients, the equilibrium decay of PARKA chen varies increasing the PARKA reactivity increases the total as c*'.
energy under the prompt spike as its width increases but The actual magnitude G fissions /g 2"U in the test has little effect on the peak value of the power. In any assembly at a separation of 2.3 m and a reactivity of event,it is feasible to stretch pulses to widths of ~10 ms
-3.6 $ is 1.4 X 10. This yield can be increased to 10" by increasing PARKA reactivity. The peak power in fissions /g zuU by operating Godisa closer to PARKA small test assemblies can be increased by a f tctor of 10 and at higher intensity. Of this yield,8 X 10 fissions /g 56 L
_f-235
- to 15 kW/g U by.eaclosing the assembly in a delayed neutrons. Counts resulting from the average polyethylene flux trap at the center of PARKA.
power level are subtracted as background.
To obtain more information aoout the response of '
According to Eqs. 8 and 9, the prompt neutron decay
- PARKA to external pulses, we substituted a portable constant a should vary linearly with the reactivity.
pulsed neutron source for Godiva. IIere, a single Godiva interval from prompt critical Ak,, provided the neutron pulse of near-fission-spectrum neutrons is replaced by
. lifetime remains constant at the different reactivities. For many smaller pulses, shorter by an order of magnitude, PARKA, changing the position of boron vanes in the of 14 MeV neutrons. The 14-McV neutrons produce a beryllium reflector does change the neutron lifetime somewhat different intial fission distribution in PARKA significantly; therefore, a does not vary linearly with Ak, and hence, in principle, a different initial transient. The Values for a observed here should be more precise than reactivity of PARKA was changed by adjusting the those predicted above. It is not expected that the decay control rods rather than the number of graphite shim of a PARKA pulse induced by a Godiva burst will difTer rods as in the measurements with Godiva bursts. In from the decays shown in Fig. 72 except for the initial
. proposed diagnostic experimech, PARY % reactivity transient.
would be adjusted with the control rods.
These pulsing experiments have established that (1)
The prompt-neutron-chain decay was measured by the initial PARKA pulse amplitude is independent of
~
collecting counts in sequential time channels following PARKA reactivity, (2) the total yield under the each pulse from the nturon generator. The neutron prompt-neutron pulse varies as 1/a except for a small detector, a BF proportional counter, was located at the -
initial transient efTect, (3) peak power levels in the test 3
-center of an array of 30 fuel pins (a 37-pin assembly with assembly as high as 10 kW/g 28'U 'can be attained by 7 center pins removed). Counting data are plotted in Fig.
using maximum Godiva pulses and a flux trap around 72 for PARKA at delayed critical and at 1.8 $ and 5.2 $
the assembly, and (4) the width of the PARKA pulses is below critical. To avoid a. continuous increase in in the range 1-10 ms. These pulses would be useful for PARKA power during the measurements at delayed transient testing of hodoscopes and for evaluation of critical, pulsing was interrupted periodically and coded-aperture techniques under realistic STF condi-PARKA was taken subcritical to allow decay of the tions.
i 10 i
i i
Z 05-d ak=0.
s 5
ok = -18 8 a=-580s" 8 01
~~
FI 72-C
~L
}
[
Decay of PARKA prompt neutrons
~
5 s~%s induced by pulses from a 14-MeV '
neutron generator.
y\\
d ok = -5.2 8 a = - 1530 s u
i I
O.Ot I
2 3
4 Time Followmg Pulse (ms) 57
. ' ACKNOWL.EDGMENTS
5. G. J. Berzins and K. S. llan," Pinhole Imaging of a Test Fuel Element at the Transient Reactor Test The authors wish to thank the members of the LASL Facility," Nucl. Sci. Eng. 65,28-40 (1978).
Critical Experiments and Diagnostics Group who con-tributed to this project, included A. R. Brown, C. C.
6. J. G. Kelly and K. T. Stalker, "ACPR Upgrade Byers, M. B. Diaz, E. O. Ferdinand, G. E. flansen, B.
Fuel-Motion Detection System" in " Transactions of PeHa, E. A. Plassman, R. L. White, R. E. Malenfant, and the Second Technical Exchange Meeting on Fuel-W. L. Talbert, Jr. We also wish to thank A. DeVolpi, C.
and Clad-Motion Diapostics for LMFBR Safety L Fink, and E. Rhodes of Argonne National Laboratory Test Facilities," Argonne National Laboratory re-for many helpful discussions. The coded-aperture experi-port ANURAS 76-34 (1976).
ment at PARKA was planned by and executed under the direction of David A. McArthur of Sandia Laboratories,
7. S.
A.
Wright _ and S.
A.
Dupree, "In-Core Albuquerque. Computer programming was ac-Fuel-Motion Detection for Large Scale Tests," in complished by II. M. Forehand and B. Wilson. Test fuel
" Transactions of the Second 7.:chnical Exchange pins were fabricated by LASL Group CMB-6. under Meeting on Fuel-and Clad-Motion Diagnostics for supervision of Keith Davidson. We are also grateful to LMFBR Safety Test Facilities," Argonne National Donald G. Simons of the US Naval Surface Weapons Laboratory report ANURAS 76 34 (1976).
Facility for the loan of the tantalum nitride target used for detector testing.
8. A. DeVolpi, R. J. Pecina, R. T. Daly, D. J. Travis, R. R. Stewart, and E. A. Rhodes, " Fast-Neutron flodoscope at TREAT: Deselopment and Opera-REFERENCES tion," Nucl. Technol. 27,449 (1975).
1. G.1. Bell, J. E. Boudreau, T. McLaughlin, R. G.
9. "Results of Fast Reactor Fuel Pin Inspection by
Palmer, V.
Starkovich.
W.
E. Stein, M. G.
Acoustical liolography," Final Report (March 15, Stevenson, and J. L Yarnell, " Preliminary Report.
1977) LASL Contract LG7-42560-1, Holosonics, Study of Fast Reactor Safety Test Facilities," Los Inc., 2400 Stevens Drive, Richland, WA 95352 Alamos Scientific Laboratory report LA-5978-MS (unpublithed).
(May 1975).
10. T. R. Ilill, "ONETRAN: A Discrete Ordinates
2. R. Avery et al," Report on Experiment Needs and Finite Element Code for the Solution of the Facilities Study," Argonne National Laboratory One-Dimensional Multigroup Transport Equation,"
report ANURAS 76-22 (September 1976).
Los Alamos Scientific Laboratory report LA-5990-MS (June 1975).
3. II. U. Wider, M. G. Stevenson, and D. A.
McArthur, " Material Diagnostic Requirements for
11. W. F. IIornyak, "A Fast Neutron Detector " Rev.
STF," presentation at the Nuclear Energy Agency Sci. Instrum., 23, No. 6, 264 (1952).
Specialists' Meeting on Fuel-and Clad-Motion Diagnostics for Fast Reactor Safety Test Facilities,
12. II. O. Menlove, R. A. Forster, R. H. Augustson, A.
Los Alamos, New Mexico, December 5-7, 1977 E. Evans. and R. B. Walton," Characteristics ordlie (unpublished).
Gas Tubes for Fast-Neutron Detection,"in " Nucle-ar Safeguards Research and Development Program
4. W. E. Stein, V. Starkovich, and J. D. OrndolT, Status Report, September-December 1970," Los "X-Ray Monitoring of Fuel Motion," in "Trans.
Alamos Scientific Laboratory report LA-4605.MS actions of the Second Technical Exchange Meeting (January 1971) p.13.
on Fuel-and Clad-Motion Diagnostics for LMFBR Safety Test Facilities," Argonne National Labora.
13. F. D. Brocks. "A Scintillation Counter With Neu-tory report ANURAS 76-34 (1976).
tron and Gamma-Ray Discriminators," Nucl. In-strum. Methods,4,151 (1959).
58
14. L V. East and R. B. Walton," Polyethylene Mod-
20. C. L Fink, R. P. Ilosteny, A. DeVolpi, E. A.
' erated 'lle Neutron Detectors," Nucl. Instrum.
Rhodes, and A. E. Evans," Analysis of the 91-Pin -
- Methods, 72, 161 (1969).
Subassembly Tests at PARKA," presentation' at the Nuclear Energy Agency Specialists' Meeting on
15. C. D. Swartz and G. E. Owen," Recoil Detection in Fuel-and Clad-Motion Diagnostics for Fast Reac-Scintillators," in Fast Neutron Physics, Part I, J. B.
tor Safety Test Facilities, Los Alamos, New Mexico, Marion and J. L Fowler, Eds. (Interscience Publish-December 5-7,1977 (unpublished). ANL CONF.
ers, New York,1960).
91 A (100,100) 10:49:12 (26 September 1977).
16. C. L. Fink, A. DeVolpi, and G. Stanford, "Ad-
21. A. DeVolpi, C. L. Fink, E. A. Rhodes, R. IIosteny, vances in Clad Blockage Detection," in "Trans-
11. V. Rhude, R. E. Boyar, L J. Duncan, and A. E.
actions of the Second Technical Exchange Meeting Evans," Fuel Displacement Diagnostics for a 91-Pin on Fuel-and Clad-Motion Diagnostics for LMFBR Subassembly," Trans. Am. Nucl. Soc.,28,499 (June Safety Test Facilities," Argonne National Labora-1978).
tory report ANL/ RAS 76-3 t (1976).
22. C. L. Fink, A. DeVolpi, E. A. Rhodes, and A. E.
17.' O. II. Nestor and C. Y. Iluang," Bismuth German-Evans,"liodoscope Performance Tests on a 91-Pin ate: A liigh-Z Gamma-Ray and Charged-Particle Fuel Bundle at PARKA," IEEE Trans. Nucl. Sci.,
Detector," IEEE Trans. Nucl. Sci., NS-22, 68-71 NS-26, No.1,827 (February 1979).
(February 1975).
23. G. W. Grodstein,"X-Ray Attenuation Coefficients
18. A. E. Evans, B. Brown, and J. B. Marion," Study of from 10 kev to 10 MeV," National _ Bureau of the "N(p,y)"O Reaction," Phys. Rev., 149,863-879 Standards circular $83 (1957), p. 37.
(1966).
24. J. G. Kelly, K. T. Stalker, D. A. McArthur, K. W.
19. L. V. Groshev, A. M. Demidov, L N. Lutsenko, and Cha, and J. E. Powell," Theory and Application of V.1. Petekhov, Atlas of Gamma-Ray Spectrafrom the Coded-Aperture Fuel-Motion Detection Sys-Radiative Capture of Thermal Neutrons (Pergamon '
tem," in " Proceedings of the International Meeting Press, Inc., London,1959), p. 82.
on Fast R,: actor Safety Technology," Seattle, Wash-ington, August 19-23,1979 (American Nuclear Society, LaGrange Park, Illinois,1979) Vol. V, p.
2302.
APPENDIX PUBLICATIONS AND PRESENTATIONS OF WORK COVERED BY THIS REPORT
1. John Orndoff," Simulation of LMFBR Test Facility
2. John Orndoff and A. E. Evans, "STF Simulation Conditions with LASL Critical Assemblies," presen-with PARKA and Application to Diagnostic In-tation at the Information Exchange Meeting on strumentation Evaluation," in " Transactions of the Fuel-and-Clad Motion Diagnostics in LMFBR Second Technical Exchange Meeting on Fuel-and Safety Test Facilities, sponsored by Sandia Labora-Clad-Motion Diagnostics in LMFBR Safety Test tories, Albuquerque, New Mexico, November Il-12, Facilities," Argonne National Laboratory report 1975 (unpublished). LA-UR-75-2212.
ANL/ RAS 76-34 (1976).
59
' 3. A. E. Evans, J. D. OrndolT, and W. L Talbert, Jr.,
7. A. E. Evans, J. D. Orndoff, and W. L. Talbert, "STF Diagnostic Instrumentation Evaluation _ with
" Evaluation of Hodoscopes for Fuel-Motion Meas-PARKA." presentation at the Nuclear Energy urement in Multipin Bundles," Trans. Am. Nucl.
Agency- ~ Specialists' Meeting on Fuel-and Soc. 30,465 (November 1978).
Clad-Motion Diagnostics for Fast Reactor Safety Test Facilities, Los Alamos, New Mexico, Decem-
8. A. E. Evans and J. D. Orndoff, " Progress in ber 5-7,1977 (unpublished). LA UR-77-2712.
Fuel-Motion Diegnostics Instrumentation at PARKA," in " Proceedings of the International
4. C. L Fink, R. P. liosteny, A. DeVolpi, E. A.
Meeting on Fast Reactor Safety Technology," Seat-Rhodes, and A. E. Evans, " Analysis of the 91-Pin tie, Washington, August 19-23,1979 (American Subassembly at PARKA," presentation at the Nu-Nuclear Society, LaGrange Park, Illinois,1979) Vol.
clear Energy Agency Specialists' Meeting of Fuel-V, p. 2225.
and Clad-Motion Diagnostics for Fast Reactor Safety Test Facilities, Los Alamos, New Mexico,
9. A. E. Evans, " Gamma-Ray Response of a 38-mm December 5-7,1977 (unpublished), ANL CONF.
Bismuth Germanate Scintillator," IEEE Trans.
91 A [100,100] 10:49:12 (26 September 1977).
Nucl. Sci. NS-27. No.1,172 (February 1980).
5. A. DeVolpi, C. L. Fink, E. A. Rhodes, R. Hosteny,
10. A. E. Evans, " Application of Bismuth Germanate II. V. Rhude, R. E. Boyar, L. J. Duncan, A. E.
Scintillators to Detection of liigh-Energy Gamma Evans, and J. Barton, " Fuel Displacement Radiation," Trans. Am. Nucl. Sci. 33, 693 (1979).
Diagnostics for a 91-Pin Subassembly,"Trans. Am.
' Nucl. Soc. 28,499 (June 1978).
I1. J. G. Kelly, K. T. Stalker, D. A. McArthur, K. W.
Cha, and J. E. Powell," Theory and Application of
' 6. A. E. Evans, J. D. Orndoff, and W. L. Talbert, the Coded-Aperture Fuel-Motion Detection Sys-
" Evaluation of LMFBR Fuel-Motion Diagnostics tem," in " Proceedings of the International Meeting with PARK A," IEEE Trans. Nucl. Sci. NS-26, No.
on Fast Reactor Safety Technology," Seattle, Wash-1,815 (February 1979).
ington, August 19-23,1979 (American Nuclear Society, LaGrange Park, Illinois,1979) Vol. V, p.
2302.
60 i
1 DISTRIBUTION Nuclear Regulatory Commission, Bethe'sda, hf aryland, R7
.380 2
Technical Information Center. Oak Ridge, Tennessee
. 50 Los Alamos Scientific Laboratory, Los Alamos, New hiexico
.480 Special Distribution Special Distribution List Argonne National Laboratory US Naval Surface Weapons Center IRT Corporation Argonne, Illinois White Oak, Silver Spring h1D 20910 7650 Convoy Court Attn: Dasid Rose. RAS Attn: Donald G. Simons, R-41 San Diego, California Alex DeVolpi, RAS Attn: John Barton Charles Fink, RAS Nuclear Enterprises, Ire.
Richard L. Bramblett Ed Rhodes, RAS 935 Terminal Way G. Stanford, RAS San Carlos CA 9J'J70 Kernforschungszentrum Karlsruhe R. Avery Postfach 3640, D 7500 R. J. Pecina Unisersity of Washington Karlsruhe, GER A1ANY R. T. Daly Department of Nuclear Engineering Attn: Kalus B6hnel D. J. Travis Seattle, Washington
11. Blum R. R. Stewart Attn: Gene L. WoodrulT R. P. Ilosteny R. W. Albrecht General Electric Co.
S. A. Cox P. O. Box 780 hl/C K-51 L W. Person Electric Power Research Institute Wilmington, NC 28401 A. Klickman 3412 Ilillgrose Ave.
Attn:
F. C. Schoenig, Jr.
P. Bertoncini Palo Alto, CA 94303 N. Lapinski Attn: Walter B. Loewenstein Science App'ications,Inc.
2680 llanaver Street University of hlaryland II;ttman Nuclear & Development Corp.
Palo Alto. CA 94303 Department of Physics 9190 Red Branch Rd Attn:
T. Gozani College Park, Afaryland Columbia, AtD 21045 Attn:
W. F. llornyak Attn: Fred liittman University of New hiexico Jerry D. h!arion Department of Nuclear Engineering Canberra Industries, Inc.
Albuquerque. Nh! 87131 liarshaw Chemical Co.
45 Gracey Avenue Attn:
n. A. Knief Crystal & Electronic Products Dept. hieriden, CT 06450 6801 Cochran Road Attn: L V. East Solon, Oli 44139 Attn: Al. R. Farukhi Reuter-Stokes, Inc.
O. II. Nestor 18530 South h1iles Parkway C. Y. Iluang Warrenssille licights Cleveland. Oli 44128 Sandia Laboratories Attn: Thomas Flemming Albuquerque, New hiesico Joseph Skarupa Attn: James E. Powell David A. NiacArthur Science Applications, lac.
J. G. Kelly 121I W. 22nd Street S. A. Wright Oak Broo. IL 60521 K. T. Stalker Attn:
R. S. hlay S. A. Dupree K. W. Cha L Al. Choote L D. Posey P. J. hfcDamal
u S GOVERNMENT PRINTING OFFICE: 1980-777-022'194 61
-

ML20002E023

Text

Navigation menu

Search