MARTIN MARIETTA ENERGY SYSTEMS LIBR

AT

3 445k 0349855 1

 

 

.@m : oz R
T T UV A

o N

 
 
  

e e

 

 

 

   

    

 

 

. ORNL-1947 f
? o This docun;rnent consists of 235 pages. iL
; e - ' _ | _ . Copy /Qé of 192 copies. Series A. '
Ifi
IE Contract No. W-7405-eng-26
AlRCRAFT NUCLEAR PROPULSION PROJECT
QUARTERLY PROGRESS REPORT
For Period Ending September 10, 1955
W. H. Jordan, Director
S. J. Cromer, Co-Director :
. ' ~ R. L. Strough, Associate Director
A. J. Miller, Assistant Director ‘
| ’ A. W. Savolainen, Editor
El "‘:. - E
! DATE RECEIVED BY INFORMATION AND REPORTS DIVISION
; (SEPTEMBER 30, 1955)
| b
DATE ISSUED
i 'E
OCT 26 1955
b
1
OAK RIDGE 'NATIONAL LABORATORY i-.
Operated by
i UNlON CARBIDE NUCLEAR COMP ANY ‘
q A Dw:sron of Umon Carbide and Corbon Corporation g
j’ _ ‘ k Post Offlce Box P v
i ST Q»uAch Rldge,‘Telnnessee ;
sfl ‘s. - %
3 445L 0349855 ]
 

 

 

 

 

VWRONOSOL AWM —

-b-.b.&AhhkwwwwwgwwwwswMMMMMMMM_.-—_..._._a_....._._..._..
CUNEWN O OVXNOUNAEWON O OVONOTUNEWOMNOCO0ONNRARWON—~O

47-48.

£
0

. N. F.Lansing

. Affel

. Baldock
Barton
Billington

. Blankenship

. Blizard
Borkowski

. Boyd

. Bredig

. Bruce

. Callihan

. Cardwell
Cathcart

. Center (K-25)

. Charpie

. Clewett

. Clifford
Coobs

. B. Cottrell

. D. Cowen

ITmI>MS=00>PM-TTOV-TO®

. Cromer
. Crouse
. Culler
Dee
. DeVan
. Douglas
. Dytko
. Emlet (K-25)
Feldman
. Ferguson
. Fraas
Frye
. Furgerson
. Gray
. Grimes
. E. Hoffman
. Hollaender
. S. Householder
. T. Howe
. W. Johnson
. H. Jordan
. W. Keilholtz
. P. Keim
. T. Kelley
. Kertesz
. M. King
- A Lane;

MO OoOmMm-—map» TP rrn

R
C
C
D
F
E
C
G
M
F
A
D
J
Cc
R
G
C
J
W
D
S
R
F.
J.
J
D
E
L
M
D
A
J
W
H
W
E
A
A
J
R
w
G
C
M
F
E
J

 

 

INTERNAL DISTRIBUTION

ORNL-1947

Progress

50. R. S. Livingston
51. R. N. Lyon
52. F. C. Maienschein
53. W. D. Manly
54. E. R. Mann
55. L. A, Mann
56. W. B. McDonald
57. F. W. McQuilkin
58. R. V. Meghreblian
59. A. J. Miller
60. K. Z. Morgan
61. E. J. Murphy
62. J. P. Murray (Y-12)
63. G. J. Nessle
64. R. B. Oliver
65. P. Patriarca
66. R. W. Peelle
67. A. M. Perry
68. W. G. Piper
69. H. F. Poppendiek
70. P. M. Reyling
71. H. W. Savage
72. A. W. Savolainen
73. R. B. Schultheiss
74. E. D. Shipley
75. A. Simon
76. 0. Sisman
77. G. P. Smith
78. A. H. Snell
79. C. D. Susano
80. J. A. Swartout
81. E. H. Taylor
82. D. B. Trauger
83. E. R. Van Artsdalen
84. F. C. Vonderlage
85. J. M. Warde
86. G. M. Watson
87. A. M. Weinberg
88. J. C. White
89. G. D. Whitman
90. E. P. Wigner {consultant)
91. G. C. Williams
92. J. C. Wilson

C. E. Winters

  
   
  

o e L

 

 

X-10 Document Reference Library (Y-12)
Laboratory Records Department

Ft-ords, ORNL R.C.

)

T TN L T T —

-,

-

€

 

 

o e e e g woTTETT T

E
b

 
 

 

 

 

128,
129.
130.
131
132
133,
134,
135-137.
138.
139.
M40, B
141,

- 142,
143-145.
- 146.
147-149.
150.
151
152.
153.
154.
155.158.
159
- 160.
161.
162,
163.
164.
165.
166-167.
168.
169.
170.
171.
172.
173..

e '-‘-_174-176;__
- 177181

" EXTERNAL DISTRIBUTION

AFDRD Jones

AFDRQ

AFSWC

Aircraft Lab WADC (WCLS)
Argonne National Laboratory

yASSISfan Secretary — Air Force, R & b
ATIC

Afomlc Energy Commission, Washington
BAGR - WADC

Battelle Memorial Institute

Boeing — Seattle

BuAer — Mueller

Chief of Naval Research

Col. Gasser (WCST\)

Convair - San Diego

CVAC - Fort Worth

Director of Laboratorles (WCL)
Directorate of Weapons Systems, ARDC
Douglas

East Hartford Area Office

Eciolpmént Ldbbi’qtory — WADC (WCLE)

'GE ~ ANPD

Glenn L. Martin

lowa State College

KAPL

Lockheed - Burbank

Lockland Area Office

Los Alamos Scientific Laboratory

Maintenance Engineering Services Division — AMC (MCMTA)
Materials Lab (WCRTO)

Mound Laboratory

NACA ~ Cleveland o

NACA - Washmgton o

  
 

North Amerlcan"'-/Aerophy5|C5 - RS
Potem‘ Branch I‘lgton S

'_Powe. Plant Laboratovry ~ WADC (WCLPU)

Pra’rf &’:_Whl’rney (] copy to R. | Strough)

- 182 R

     
   

Sl

‘.“ifiDlwsmn of Reééarch and Medicme AEC ORO

 

.o

 
Ko e

i, i

 

 

 

Reports previously issued in this series are as foliows:

ORNL-528
ORNL.-629
ORNL.-768
ORNL-858
ORNL-919

~ ANP-60

ANP-65

ORNL-1154
ORNL-1170
ORNL-1227
ORNL-1294
ORNL-1375
ORNL-1439
ORNL-1515
ORNL-1556
ORNL- 1609
ORNL-1649
ORNL-1692
ORNL-1729
ORNL-1771
ORNL-1816
ORNL.- 1864
ORNL-1896

Period Ending November 30, 1949
Period Ending February 28, 1950
Period Ending May 31, 1950
Period Ending August 31, 1950
Period .En‘ding December 10, 1950
Period Ending March 10, 1951
Period Ending June 10, 1951
Period Endin)g September ]0', 1951
Period Ending December 10, 1951
Period Ending March 10, 1952
Period Ending June 10, 1952
Period Ending September 10, 1952
Period Ending December 10, 1952
Period Ending March 10, 1953
Period Ending June 10, 1953
Period Ending September ]0, 1953
Period Ending December 10, 1953
Period Ending March 10, 1954
Period Ending June 10, 1954
Period Ending September 10, 1954
Period Ending December 10, 1954
Period Ending March 10, 1955
Period Endifig June 10, 1955

 

€

 

 

 

 
 

 

 

FOREWORD

This quarterly progress ‘i'eport of the Aircrof'f Nuclear Propuisionr Project at ORNL records
the technical progress of the research on Icirc_ulati'ng-.fue( reactors and all other ANP research at
the Laboratory under its Contract W-7405-eng-26. The report is divided into three major parts:
l.  Reactor Theory, Component Develbbhént, and Construction, [l. Materials Research, and
tH. Shielding Research. |

The ANP Project is coifipfised of about 510 technical and scienfi,fi‘c personnel engaged in
many phases of research directed toward fherqc‘rhrievemem‘ of nuclear pfopulsion of aircraft. A
considerable portion of this research is performed in support of the work of other organizations
participating in the national ANP effort.  However, the bulk of the ANP research at ORNL is
directed toward the de\felopmenf of a circulcfing-fuel type of react.or.

The design, construction,i and operatibn of the Air'crdft Reactér Test (ART), with the coopera-
tion of the Pratt & Whitney Aircraft Division, are the specific objectives of the project. The
ART is to be a power plant sYs'remr that will include a 60-Mw circulating-fuel reflector-moderated
reactor and adequfife means for heat disposal. Operation of the system will be for the purpose

of determining the feasibility, and the problems associated with the design, construction, and

“operation, of a high-power, circulating-fuel, reflector-moderated aircraft reactor system.

 

 

-

 
B iy e e i,

 

 

 

 

 

b kil

 

 

.,

 

 
i

 

 

FOREWORD

 

CONTENTS

........................................................................................................................................................

SUMMARY ......oooovomenncnnn e |
REACTOR THEORY, COMPONENT DEVELOPMENT, AND CONSTRUCTION

PART 1.
1. REFLECTOR-MODERATED REACTOR

...................................................................................................

..................................................................................................

Aircraft Reactor Test Facility
Shielding Experiment Facility
Aircraft Reactor Test Design

....................................................................................................................

......................................................................................................................

Reactor Design

...........................................................................................................................................

Main and Auxiliary Radiator Design.. ..ot
" Fuel-to-NaK Heat Exchanger Design ...,
Core FIoW STUGI@S ..oiiiiiieiiie ettt et eb e et b et et e e e st a s st et e e re st e sae et e s
Core Design ARGl Ses ....co.ooiiiiiii ettt e st er s e ee s e
Fuel Pump Performance .........cooiiiieieiet ettt ettt et m e ebe et nes et s

Controls and Instrumentation

....................................................................................................................

Engineering Test UNit .ottt st st b sttt rseses

Reactor Physics.............. OSSO UOO RSO SOPOORPOOY e ettt ses
. Control Rod Heating and BUFIIUD ..ot ettt et as st s s st as s e s e s s et s s s s en e
ART BOron Layer oottt ettt bbb a b bbb n e b s s
High-Temperature Critical Experiment vs ART ..o
Multigroup, Mulhreglon Reactor Caleulation ..o s

2. EXPERIMENTAL REACTOR ENGINEER!NG....7...........'... ..... e

In-Pile Loop Development and Tests et ee ettt a e e et A oA R ae b A b bt a ettt ene s es e e b,
L 00P IR STAIIAEION ... ottt ettt e
Loop NO. T OPEIation ....ccoiuiriceerieieree ettt sttt et e e s s ettt s ene s e st eaeneeen.
Horizontal-Shaft SUmp PuUmp ...t
Ol TrradiGion ...

Deve!opment and Operahon of F'orced Circulcmon Corros;on and Mass Trcnsfer Tests oo
Operation of Fused Salt—Inconel Loops ..o st
Liquid Metals in Mul’rlmetcl Loops..._ ..... e s s

Pump Development .....iov....s/vooovveoeiioreeeee, et tbetatet et eeeb et ettt et b sbebereres e b eae e raatete e r e et
Mechanical Shakedown and Be-armg and Secl Tesfs ..... bbb

Short-Circuit Pump Test STANd oo, et eteteiesheitteeeaeat e e ntare et e aren e aree et erne e eteaneeee

High- Temperature Pump Performcnce Test Stcmd ................................................................................

 
 
 
 

Heat Exchcnger Tests. . 7 7
Intermediate Heat Exchanger Tesfs e -'";"'”‘""t"'t ........ e e e
Small Heat Exchqnger Tests ....... - '

.............................................................................

 

Str u;tqg’ql Tes’rs

........................................................................

 

 

  

15
15
15

15
15
19
19
24
28
29
30
31

33
33
33
33
35

36

36
36
36
37
38

38

38

42
42
44
44

45
45
49

51
51
52
54

54

vii

 
 

 

 

 

3. CRITICAL EXPERIMENTS

Room-Temperature Reflector-Moderated-Reactor Critical Experiments .........c.ccovorriiiinnrininnins S

High-Temperature Reflector-Moderated-Reactor Critical Experiments......cccccoviiiinnnns e s

PART 1. MATERIALS RESEARCH

4. CHEMISTRY OF REACTOR MATERIALS .ooooooeeoeee oo eeeseses s ess e o
Phase Equilibrium Studies of Systems Containing ZrF ; and/or UF ; .cooooviiiniiiiiii, -

The System LiF-ZrF .. oo s
The System UF (-ZrF | oo
The System NaF-LiF-ZrF ;.o
The System NaF-LiF-UF (e
The System NaF-RBF-ZrF ~UF ;oo s
The System KF-ZrF ; oo e
The System NaF~KF-ZrF4 .......................................................................................................................

Phase Equilibrium Studies of Systems Containing Bl s
The System NaF-LiF-BeF , ..o
The System NaF-LiF-BeF ,-UF ;.o
The System NaF-BeF j-UF ;..o
The System KF-BeF oo
The System NaF-KF-BeF 5 ..o s

Solubility of UF; in BeF,-Bearing Compositions ...

Chemical Reactions in Molten Salts ..o s s
Equilibrium Reduction of FeF, by H, in NaZrF o oo
Reduction of UF ; by Structural Metals ..o
Stability of Chromium and Iron Fluorides in Various Molten Fluorides .........c.ocooooivieviniiiin.
Experimental Preparation of Pure FIuorides. ..ot
Reaction of UF , with Uranium in Alkali Fluorides ...
Reaction of Uranium Metal with Alkali Fluoride Mixtures .........ccccocoviviiieiiiiiee e
Metal-Salt EqQUilTBrium oottt ettt et ettt en et et eaene

Production of Purified FIluorides ...t e e s s
Recovery of Contaminated Fuel for Re-Use.. ..o e
Removal of CrF, from NaF-ZrF ,-UF , Mixtures ..o
Effect of ZrO, in Fuel Preparation ...
Pilot-Scale Purification Operations .........coceuvieioiiiicriveieiereesssseesessresssnsesieeesess s nesns e
Production-Scale Operations ... et ettt ettt e
Preparation of ZEF | ..ottt bt sttt esr b et ne st e paen
Batching and Dispensing Operations ...ttt et asn e e s e seene
Loading and Draining Operations ...
Enriched Fuel Preparations ..ottt sttt '

 Fundamental Chemistry of FUSed SAIES ..ot es s essees st eeeee e eeeseeee s

Solubility of Xenon in Molten Salts ... e
EME Measurements in FuUsed Salts .ot ee et eee et e st e e eee et eeeseeee et et es s eerasesesesnesseies

Activity of Chromium in Inconel .......c.oooiiiiiiiiic s et .

Viscosities of Molten NIrates ...ttt e
Optical Properties and X-Ray Patterns for Recently Discovered k
Compounds in Fluoride Systems ...t e s e
High-Temperature X-Ray Spectrometer Studies .....cccvvivcininiiiiiiientncece e et nes
i Physical Chemistry of Fused Salts ..o v
7 Diffraction Studies of Liquids.....cco.ooiuriioiiiicee oot e

viii “

¥} £

 

 

 
RGN i e+ R

 

Ton

S.

F OrCed-CHrCUlaHON SHUGIES omvrooooeoseeeee e eseoee oot ee e oot es et
Fluorides in Inconel ....... e e bR AL b e AR
Sodium in Inconel and in Stainless Steel ..o

Thermal-Convection Studies ... e

- Effect of Yarious Loop Cleaning Methods.......ccocooccvvieiriesrierne. ettt sttt et e et est s s aeanan
Effect of Heating Method .......oooi oottt et et

- Effect of Temperatures............. e et ettt eeh ettt Ao e At bt et R ekttt e b ettt
Effect of Applied Electromotive Forces .......coovoiiiiiiiciiiiii et

' Effecf of Oxide Additions 1o SOdiUm ...t

General Corrosion SHUAIES ....o...eoooveeeereoeeeeeeeeeeseeeeese s e s eeeseesess st sess et e
Hot-Pressed Metal-Bonded Tungsten Carbide in NaF- ZrF UF
Solid-Phase Bonding OF COIMEES ..ccveeeeeeeeeee sttt e s et st es e s s
Effects of Ruthenium on Physical Properties of Inconel ...

‘Brazing Alloys on Inconel and on Nickel in Sodium and in NaF-ZrF -UF .o,
Brazing Alloys on Inconel and on Stainless Steel in Lithium ..o
Hastelloy B=Inconel in NaF-ZrF -UF .ol et — et r e s ea et s st n e e n st s e en
Boiling Sodium in Inconel ... et
Inconel Exposed to a Sodium-Potassium-Lithium Mixture. ...

- Molybdenum, Vanadium, and Niobium in Static Lithium ..o e
Ana|y5|s of Metal Crystals from Inconel-Sodium Thermal Convecflon Loop .iovvveirinrieeii e,

Fundamental Corrosion Research ... e et et eteeet e tetee et eee ettt tebe et st es e teates e te e anent st etesers s s
Film Formation on Metals .........c.......... eeeeerteeraetastattesuiteesisie et ee st es s e bR bR A ba e b SRt e s bbbt ket
High-Temperature Spectrophotometry in Fused Hydroxides .......c.coovoiioieiiniiiriinieieseienicieeeceecenes
Mass Transfer and Corrosion in Fused Hydroxides ...,
Thermol Dlssoc:qhon of Sodium Hydroxide................ ettt eeatehs et aebehe s b et et eht bt ettt et et et e b et eab s e

~ Chemical Studies of COEFOSTON e vereeee e eeereseeesseessseesessssseeseessesemeseessesesesssesoessseeeees s oeereee e
" Inconel in NaF-LiF- ZrF4 U
Stability of UF, in NaF-KF- L|F I INCONEI oo e
Effect of Chromlum on the Mass Transfer of Nrckei in NaOH ............................................................

METALLURGY AND CERAMICS .................................. ". .......... ) ......... s s .

Mechanical Property Tests of Inconel ................. e ettt reea—eit et et ot e tbeanseeaeae e s e e anseeteeas
Stress-Rupture Tests e s ST OSSOSO OSSO POV OT U TERTUOPEPTPIOROPIOOY
BEllOWS TSt irvooaersee oottt s est et aees st s ettt et
Interaction Between inconel crnd Berylllum....;..;.Q.;.:....; ......... e et

Development of Nickel Molybdenum A!loys ettt ettt ettt s et '

© OXidation SHUAIES .o e ettt
Fabrlcatlon Studles...'_’./.;.‘._‘..'.'.'.'..;'..4...&." ...... e e s it .

~ Physical Properhes and M:crosfructure Sfud:es of Hastelloy B

' Brazmg of Boron Car_blde...:...f.'.',..f...'.'.:.'..'....:.'....

o Fabncohon of TestComponents
"~ Components for High-Temperature Critical Experiment.........cooiiiiiiniinnaincnes s e '

Stress- Rupture Tesfs of Hasfelloy B

CORROSION RESEARCH ..o ees oo oo eees oo e et er oo ere s ereee s oo

 

 

Brazing Alloy Developmem‘ and Teshng
Deveiopmenf Tests .

Brozmg of Cermet Valve Seats fo‘ inconel Componem‘s o 7' o

   
 

NcK-tg-Air Radiator

 

 

95

95
95
97

100
100
101
101
102
102

104
104
104
106
106
108
109
109
111
111
113

113
113
117
117
118

119
119
120
120

122

122
122
123
124

125
125
126

127

127

130
130
130
131

132
132
133

 
 

 

S ff'.:_:f.jfl-"'f "ANP Servnce Laborcfory

 

 

 

134

Intermediate Heat EXChanger NO. 3 ... eseeersee e eeeeee s s ereeesesesoeees s ee s er e es e ee s
- Cornell Radiator No. 2 ... e s 135
o NondestrUCf-ive"Tesfing ................................................................................................................................ 136
1 Ceramtc RESEATCR et eeseee s eeereseeseesesee s eeeraessees s s emeseee e e et oo 140
* Graphite-Hydrogen Corrosion-Erosion INVESHGation ........coovcovvereevieeessersessesseseesssssssesseseeseseereme 140
Rare-Earth-Oxide Control Rods..........ccc.oiiiiiiiiieiin e s e ssse s 140
Calcium Fluoride and Aluming Detector SPacers ...........cococioiiininiiniininne e e 140
Fluoride Fuel Pellets. ..ottt ettt e 140
Synthesis of Boron Compounds .........coiiiviriiieiiriietcee ettt er bttt bt 140
Fabrication of Dysprosium Oxide Disks ...ttt 143
Fabrlcuhon of Europlum Oxtide Wafers......coooiiieceeee ettt 143
ik Specm| Materials Studies ....... SEURUROTPURRORRN ettt ettt e 144
R Cglfiumblym/Rexserqrch .................................................................................................................................. 144
Composite Tubing Fabrication ..ottt e e 145
Oxidation Tests of Aluminum Bronze.......ccccceeienuen, et eh et r ettt een et s et eas et s er e bt e ereeeete e e etae s eneos 146
C Lead-Calcium AHOYS ..o ettt 146
Neutron Shielding Material ..........ccooiiii et 147
HEAT TRANSFER AND PHYSICAL PROPERTIES .oooiovooveveroereeseeeeees oo e eooesno 149
Fused Salt Heat Transfer ... et nens 149
ART Fuel-to-Nak Heat Exchanger..........cccocoeveviinenn, ettt ettt r e 150
ART Core Hydrodynamics........o..oooioieioeeeteee oo eoe e . 151
Reactor Core Heat Tramsfer ... e 151
Heat Capacity ottt et e es e et enees 154
VESCOSITY oottt ettt ettt e e e et ee et 155
Thermal CondUCivity ..ottt eee et st ee e s er e 157
Electrical Conductivity oottt ee e e 157
RADIATION DAMAGE ...ttt e 158
LITR Horizontal-Beam-Hole Fluoride-Fuel Loop ......iocoioiiiieieieeeieeeeeeeteeeeeeeeee e 158
Miniature TR-Pile Loop oottt ee e er et es et 164
Creep and Stress-Corrosion T@StS. ..ottt eee et eer e ee e e res st s e es e ee s es s 165
Flux Measurements in the MTR .. oo e, 166
Analysis of Reactor-Grade Beryllium ..o 169
ANALYTICAL CHEMISTRY OF REACTOR MATERIALS oo, - 172
Determination of Oxygen in Sodium........co.co..iviiiiioceieee e ees e 172
7-Butyl Bromide Method ........ccoooiiiiiiiiii et e ee e 172
Vacuum-Distillation Method................... OO UOU PO O SOOI 173

Volumetric Determination of Zirconium in Fluoride Salts with , _
Disodium Dihydrogen Ethylenediaminetetraacetate (EDTA) e, 173
Determlnatlon of Uranium Metal in Fluoride Salt M:xtures et b ettt n e, 175
Ehmmahon of Air from @ Dry Box.........cooiiiiiii s | 175
o 'Defermlnatlon of Oxygen in Metallic Oxides by Bromination ..o, — 175
Lt Determmahon of CO by Means of a Solution of PACl, o 176
177

 

 

hn:x;!;_

 

 
 

 

  

10. RECOVERY AND REPROCESSING OF REACTOR FUEL ...coooovviiiiiierce i 179
 Pilot Plant Design.. oo ooveeieee e et s e s s 179
Engineering Developments ...........c.cocoiiiiiiiiiiie ettt 179
Process DevelOPmMENnt ..........cccoivivoieee oottt ek 180
o PART Ill. SHIELDING RESEARCH

11. SHIELD DESIGN st SRR -
- Welghts of Spherlcolly Symmetric Unit Shields for C:rculahng-Fuel Reactors.......cceeeeevieeeivceennnennnn, 185
Sources of Radiation in a 300 Mw Circulating-Fuel Reactor ... 185

~ Shield Weight Dependence on the Dimensions of a 300-Mw Clrcula’rlng-FueI Reactor ....cccocvveeunnn. 189

- Neutron Shield ... ettt eeateteettteteoatteteertaeearse e e ehe et s et b e e et e et eet e et e e et te et e san e 190
Gamma-Ray Shield ... 191
We|ght D et e MINGEIONS o oottt ettt ettt s e rs e e et e en e e et e et e et e ent e e aan e eeenbeeneas 195
12. LID TANK SHIELDING FACILITY oo oo .97
Reflector-Moderated-Reactor and Shield _Mockup TSt S oottt e e 197
Gamma-Ray and Neutron Measurements Beyond the Mockups.....oooociiie, 197
~ Sodium Activation in the Heat Exchanger Region of the MockUpS oo 197
13. BULK SH!ELDING FACILITY s 199
_Angular Distribution of Fast Neutrons Through Water oo e e 199

| Decay of Flssmn-Produc’r Gamma Radmflon ............................................................................................ 201
14. TOWER SHIELDING FACiL[TY ................................................................................................................ 205
' Procedurerfor Using TSF Data for the Optimization of a Divided Neutron Shield.........coooeiee. 205
- Calculation of Shield Weight ....................... OSSOSO PO OO OO TS POT SO UORR OO 205
 CalcuUlation of DoSe Rat ....c.ooioviioeiieieie oottt bttt 206
Calculation of Minimum Shield Weight for Specified DoSe .o, 208
Appl:caflon OF TSF DIGHG ..ot 211
- Results of a Sample Shield Opflmizution Calculation ..o s 215
Measuremenfs in fhe GE ANP Crew Compcrtmenf Mockup e I e 217
ORGAN!ZATION CHARTS e 221

xi

 

 

T

 
 

 
 
 

 

 
 

 

 

IR

 
1

 

Cewl

""ro gef q high degree of mIXIng T
orie’ system “the vortex axes paro”el ‘the |s!and ‘and

ANP PROJECT QUARTERLY PROGRESS REPORT

SUMMARY

'PART I. REACTOR THEORY, COMPONENT
DEVELOPMENT, AND CONSTRUCTION
- 17. Reflector-Moderoted Reactor -

A confract was oworded on Augusi 19 ]955 to the
V. L. Nicholson Cornpany, ‘Knoxville, Tennessee,
_'rhe Iow bldder for pockoge ] of the Aircraft Reactor
‘Test (ART) focdrfy. Therr low ‘b:d of $765,835
included $264,373 for the reactor cell, with the
Chrcogo Bridge & lron Co. engaged to serve as

“the subcom‘ractor to des:gn, fabricate, and erect

the cell, Package 1 includes, in addition to the

“cell, the major modifications of Building 7503.

Recent experiments at the Tower Shielding
Focr]rfy (TSF) emphasized the need for shleldlng
data from the ART, and therefore a layout has been
comple‘red to provrde for the measurement of the
gommo-roy ‘spectrum of the ART as a function of
the angle of emission from the reacfor shleld

; surfoce.

- Layout drow:ngs have been completed for the
prmcrpol features of the ART reactor—pump-—heat
exchanger-—-pressure shell assembly, and the de-
tails are considered to be sufflcrenfly firm for
procurement to be started. Fabrication of the

4‘she'Hs, such as the pressure shell, reflecfor |

shell, etc., will be particularly hme-consumlng
because of die problems, and fherefore procurement

' ’work began with the shells,

A one-bclf-scole plcshc model of fhe pump ‘and

expansion tank reg'°“ at the *OP of the reactor hos

  
  
 
   

been completed to facilitate exommcmon of fob -
cation, stress, and flmd-flow problems, ond wo k

has started on a‘full scale olummum'model
sign condmons a e stab

 

for the main and auxiliary radlorors,fand rhe de-
sign for the fuel-to-NaK heat ‘exchanger has been
selected Design colculotlons ond loyout drowungs N

   
 
  
 

IR L

mveshgoted bofh mqke us

 

 

 

in the other they spiral helically downward around

it. In both systems the two fuel pump volutes dis-

 

 

_date Mfor ini
The ob|ec’r|ves of 'fhe ETU ore to devefop fechmques_

  
  
 
 
  
 

chorge 'rongenholly into the core to give a system
that is insensitive to the one-pump-out condition,
Tests performed to date were planned to assist in
the systematic development of an inlet-guide-vane
and turbulence-generator design that will produce
radial velocities of sufficient magnitude to keep
the boundary-layer fluid mixed with the free stream.
High-speed photographs of dye injections and con-
ductivity measurements of salt injections are being
used fo examine radial flow, circumferential dis-
tribution, and transients, Core designs with lower
degrees of divergence than that of the present
21-in. design are being studied to determine the
effect of inlet annulus dimensions on the adverse
pressure gradients encountered by the fuel in
flowing from the core inlet to the equator.

Performance tests of the fuel pump with water as
the circulated fluid were completed. The noise
present in the initial testing was found to result
from a local flow condition that existed at the
tongue of the pump volute. A modification of the
volute design has eliminated the noise, and other
impeller modifications have brought about a con-
dition in which the hydraulic force balance on the
impeller occurs near design speed and flow., The
tests have demonstrated that the design point lies
in the region of maximum efficiency.

A flow sheet has been prepared that shows,
schematically, the locations of the principal
instrumentation ond com‘rol components of the

ART _ond control pcmel loyout studies are being
,_mcde. Excepf for special sensmg eqmpmem‘, it

appears fhat no development ‘work will be requrred

LA construchon program fo_r the Engmeermg Test

(.ETU) hqs been esfoblushed with a _target
[ operohon f September 1, 1956.

    

     

e e B S T

b]y, fo obtain some

on the -rodlotors ond the NaK 'ro fuel ond fhe sodlum~

urfs, cmd to fest some of the mstrumen-

be used_ 'on ‘the ART

 

o Calculations 'ofwcontrol rod heatmg and burnup to

" be expected in the ART were completed, and the
relative merits of various materials were studied,
For example, for a 1/‘!-in. annular control rod of

 

 

 

 

T B TR

 
B W
I

" ANP PROJECT PROGRESS REPORT ¥

 

 

 

  

: BIOC in arcep'per matrix (30 vol % B 4C) thébdrhub“
= penetrahon would be one-sixth as great as that in

a europ:um oxide cermet, but the heat produced
would be one-third greater. The boron layer be-

'f:_'rween ’rhe reflector and the fuel-to-NaK' heat ex-
_ changer was exammed with respect to heat genera-
':”.':tlon, flux afienuahon, helium gas evoluhon, and

burnup.  Also, the differences between the ART

and the “hi'grh ‘ferfi'perdtdre critical experiment were

. ‘evaluated. It is estimated that the critical con- |
. j'_'f'cenfrahon of the ART WI“ be between 4 6 and 5.4
“,'Ti_'mole % UF '

compufe fluxes as a funchon of one space variable
in slcb cyhndflccl or spherical geometry, and it

'wull allow 125 groups and 125 regnons.

2 Experlmental Reucfor Engineering

The frrsf loop for” c:rcu!cmng fluoride fuel inb

hole HB-3 of the MTR was shipped to NRTS on

June 20. It successfully passed the preoperational
checks, and startup of the loop proceeded satis-
factorily to the final step, that is, melting of the
freeze line and filling of the loop with the fuel
mixture. The heater for the freeze line was found
to be inoperable, and therefore the experiment was
terminated. The loop has been returned to ORNL
and disassembled, The modifications required to
overcome the difficulties encountered have been
incorporated in a second loop that is nearly ready
for shipment to NRTS, A third loop is also being
fabricated. The auxiliary facilities required ot the
MTR for operation of the in-pile loop were com-
pleted, including a loop retractor mechanism for
adjustment of the loop nose position during reactor
operation.

Twenty-two fused-salt—Inconel forced-circulation

| loops were operafed in the test program for studying
corrosion and mass-transfer in high-temperature-

differential, htgh velocity systems under conditions
that simulate reactor flow rates and temperatures.
Nine similar loops were operated with sodium in
Inconel or in srcnnless sfee! tubmg.

A new test loop has been des:gned w:th whlch
it will be possnbie to obtain accurate information
~ ... on 'rhe oxygen content of sodium or NaK during

pera’r:on of the loop. The main loop of the system

is to include a bypass cold trap and a sampling

 

en.d analyzing device, and cnauxfl:arypluggmg-

indicator loop will be attached to the main loop.
Also, a loop has been designed, in which NaK will

be circulated, that has the same ratio of surface

area to volume as that of the primary NcK cnrcuufs

of the ART,

The design layouts for two loops for high-
temperature tests of full-scale models of the ART
fuel and sodium pumps were completed, and fabri-
cation and assembly were started, The first group
of bearing-and-seal and mechanical tests on the
rotary elements of the pumps was completed. In

A new muihgroup,mulhreglon reactor caleulation general, the test results indicate that the bearing-

Cis bemg programed for the Oracle. The code will

and-seal designs for the upper and lower units can

‘be made to function satisfactorily. One short-

circuit pump-test stand was completed and shake-
down tests with water were started.

Intermediate heat exchanger test stand A was

'operc:'red for 690 hr in a series of furnace and"‘_:

diffusion cold-trap tests of the NaK circuit and a
2-hr cleaning cycle of the fluoride-fuel circuit.
The gas-fired furnace for heating the NaK proved
to be copable of transferring 1.13 Mev of heat to
the NaK aofter minor modifications, including in-
stallation of a larger burner, had been made,
After the fuel mixture had been circulated and
dumped, a leak occurred in one tube bundle of the
heat exchanger, and subsequently one of the
NaK-to-air radiators developed a leak. The leak
in the heat exchanger was found to be a radial
crack on the inside of a tube bend, It was dlso
found that severe distortion had occurred as a
result of the thermal cycling created by operating
the heat exchanger with NaK in the tubes but with-
out fuel around the tubes. The temperature dif-
ferences thus created ranged from 200 to 1000°F
The failure of the radiator is also thought to be the
result of extreme thermal cyclmg, but the analyses
of the difficulties have not yet been completed.

‘Additional test assemblies are being fabricated.

The first series of tests were completed on a
small (20-tube bundle) heat exchanger that was
operated nearly continuously for 1560 hr. Heat
transfer, pressure drop, and corrosion and mass-
transfer data were obtained. No appreciable

amount of mass fransfer could be detected by
visual inspection, and metallurgical examination

is under way. The oxide content of the NaK' was

found to be high, and therefore cold 'rrcps ore ‘ro be
lnstolled in subsequent test !oops.

 

 

 

i

.;tl;":‘ I - A .

 

 
 

 

-,

 

4 "cenhmeter of fuel region, and a

-\ one-fourfh-scale model of the lower half of the
21-in. reactor core shell was fobrlccted for thermal-

‘stability tests, and a design was completed of an

Inconel strain-cycling rig. A third thermal-cycling
test of a sodium-béfyl!iquEnconel system was

'sfarted A systemcmc study was made of devices
" for the measurement of oxides in circulating

sodium or NaK and the removal of the oxides

during operahon. Designs of cold traps and plug
) mdlcators were prepared '

' 3 Crmcal Expenmenfs

c The fuel concentrcmon in the room-iempera’rure'
’ crmcal assembiy of the reflector-moderated circu-

latmg-fuel reactor was decreased from 0.416 (3%
excess react:vnfy) to 0345 g of y23s per cubic
“‘clean’’ critical
mass of 19.9 kg of U235 was obtained. Two
s'rrucfurol chqnges in the assembly were also
studied. ln one of the modlflcchcms, the average

TWIdfh of one of the end ducts was mcreased from

'(29 to 2 80 m., which increased its volume al-

'most 2.5 hmes. The corresponding critical mass
‘was’ about 24 kg of U235. In the other alteration,

the radius at the center of the berylllum island
was increased from 5.18 to 7.19 in. The critical
mass of th:s assembly was 18 4 kg of U235

Another crmcal assembly of this reactor is bemg‘ |

opercfed at fempercfures between 1200 and 1350°F,
The mmlmum cnhcal concenfrcn‘lcm has been de-

termmed as 6.29 wt % (2 96 mole % urcmlum ina

mixture of sodium, zrrconlum, and enriched uranium

'; __fluor:des. The Over-aH temperature coefficient of
redctivity is -2 >< ]0-5 (Ak/k)/oF and the control
. h :

ifielfmg pomfs were accompamed\:by exceechngly‘a':"T
" high vcpor pressures of ZrF ;. In the NaF-LiF-UF

system ‘it was found that mixtures in the reglons

  
  
  
 
  
    

PERIOD ENDING SEPTEMBER 10, 1955

of low meltlng pom’r contained too much UF, for
use with circulating-fuel reactors,

Preliminary studies were made of the NaF-RbF-
ZrF ,-UF, system fo determine whether it could
provide a low-viscosity fuel. Previous work on the
RbF-ZrF, system had demonstrated the need for a
fourth component in order to obtdin a low-melting

fuel with 20 to 40 mole % ZrF4. All the work was

carried out with RbF which contained about 20
mole % KF as an impurity. The KF caused un-
expected results in the phase diagram work, which
has therefore been postponed until better RbF can
be procured Viscosity measurements in the sys-
tem RbF-ZrF -UF (48-48-4 mole %) below 600°C
were hmdered by particulate matter, but the
measurements made above 600°C gave viscosity
values which were somewhat lower than those in
the NaF-ZrF -UF, system. In order to obtain
further msrght into ’rhe mixtures containing rubidium,
sfudles were made in the KF- -ZrF , and NaF-KF-
ZrF4 systems, The latter system proved fo be
much more complex than either the NaF-ZrF ,-UF

or NaF-LiF-UF, systems, which show no 'rerncry
compounds. The NaF-KF-ZrF, system was shown
to have at least five ternary compounds.

In the system NaF-LlF-BeF2 it was found that
acceptably low melting points can be obtained with
low LiF concentration by moving into the ternary
system along the drainage path leading from the
NaF-Na,BeF, eutectic toward the LiF-Na,BeF,
eutectic. In order to obtain-a melt with a kinematic
viscosity as low as that of the ARE-type fuel, it
was established that the BeF, content cannot be
greater than 30 mole % to assure complexing of
BeF as BeF ™" The LiF content will probably
have 1o _be. less than 10 mole %. The melting
“point of NaF LIF Ber (63.5-7.5-29" mole %) was
found to be 4A525°C -and a defermma’non of its

  

mixtul es the concen-

ot 800°C is about

   

il i

trotioh"of Urcimum (in g/cm

 

the same as the concentration of NaF-ZrF4-UF4
(50-46-4 mole %).

'____f’uture. ) The_ |

P T

o

 

TR
L g driiac, o peah

 

 

R

'ANP PROJECT PROGRESS REPORT

“io 0 Liquidus temperatures were determined for mix-
" tures of NaF-BeF, (70-30 mole %) with about 2.5
 ‘mole % UF4 because of the reported low viscosity
: *’_flig_of the NaF- BeF2 mixture., The liquidus tempera-

fures were in the _region of 560°C, which is un-

3 Jf’sahsfactorlly high. A more complete investigation

< of melting points in the NaF-KF- BeF system

"ff'}_f{f;:,-conf:rmed prevrous observahons of hrgh hqurdus
~ temperatures. ’

   

Analyses were carried out on mixtures of UF3 -U

,ikf-and UF -U dissolved in LrF—BeF and NaF-BeF

:-;»‘.‘,me]fs cmd filtered in either copper or nickel equ:p-
5 _;'_imenf. The results showed that the solublhty and |
;“}srcblhty of UF in these melts were not solely a
_‘func'rlon of the form of the uranium addition and

 

the confomer mefcl The resulfs were errahc, ‘and

‘_further experlmentoflon |s necessary To defermrne

the controihng foctors.
By means of an equrllbrcmon technique pre-

- .viously described, a more precise value for the
“equilibrium constant of the reaction FeF2 +

H,=— Fe® + 2HF in NaZrF, was found to be K
5 2 at a concentration of FeF2 corresponding to
6490 ppm of iron. A study was made of the re-
duction of UF , in NaF-LiF-UF, (22-55-23 mole %)
by Cr® and Fe® at 600 and 800°C. The equilibrium

 

 

210,
Iy
4(@9) || NaF-ZrF, ()

M| MF NaF ZrF

2(satd)’ (a )'

i

chromium concentrations were found to be lower
than those of fuels with greater-concentrations of
ZrF,. The equilibrium iron value remained about
the same as that of fuels made with NaF-ZrF,
mixtures and the NaF-LiF- KF eutectic, but lts

temperature coefficient was reversed to give

slightly higher values ot 800°C than at 600°C,
Data were obtained on the stability of iron fluorides
cnd chromlum fluorides in various fused matena!s.

Reducflons of UF4 wr'rh excess uranium were

carried out in several melts. In the NaF-LiF

eutectic the UF, was more stable in copper than

in n:ckel equipment, In the NaF-KF-LiF eu-

tectic the stability of UF, was ‘independent of
the container, and, with copper containers, a

uranium- copper alloy was found. This alloying

7_wnrh copper did not occur with either NaF-LiF
"/euiechc or BeF, melts,

A sfudy of the reaction of metallic uranium with

""_'_several fused mm‘erlcls at elevated temperctures

 

 

dlsclosed reduchon and volaflhzohon of olkah'
metals.  Equilibrium studies on mixtures of

potassium metal and NaF gave results that led to
an equilibrium constant at 800°C of

o K= (Na/K)KF/NaF) = 0.2 .

Workable methods have been devised for re-
processing fuel used in component testing at
ORNL and Pratt & Whitney. It is expected that
2000 to 3000 Ib will scon be available each month
for reprocessing and re-use. Since a satisfactory
commercial source of ZrF , has not yet been found,
a unit is being fabricated which is capable of pro-

‘ducing 1000 Ib of ZrF, per week by direct low-

temperature hydrofluorination of ZrCi ,.
A shortage of ZrF, reduced operation of the

large-scale (250-1b bcn‘ch) fluoride-processing facil-

ity to one-half the normal capacity during this
quarter. A total of 6960 Ib of purified fluoride
compositions was produced, including the fuel
carrier for the ART high-temperature critical ex-
periment, Pilot-scale equipment was used to pro-
duce 53 batches of purified fluorides. Enriched
fuels were prepared for the high-temperature
critical and in-pile loop experiments.

Examinations have been made of cells of fhe

type

1 1
NaF ZrF4(a;), M Fz(sq*d) M,

where M and M! are Cre, Fe°,'crnd Ni®, These
cells appear to be reversible and are reproducible.

Interpretation of the data is complicated by the

solid fluorides in equilibrium with the melt not
being simple metal fluorides and their compositions

being dependent on the amount of fluoride present

in excess of the saturation value, Reproducible
emf’s have also been obtained in the case where

M and M! are Cr° and where fhe (':.rF2 content is
" less than the saturation value at one of the elec-

trodes. In the cell Cr|NaF, ZrF,, CrF, | Inconel,
with no barrier between the Inconel and chromium,
the potential eventually dropped to zero because

of the Inconel being converted effecflvely to a
chromium electrode. The mechanism for tronsportr

of chromium might include dlsproporhonahon of

CrF, to CrF, and Cr® at each electrode, with the

solution being carried from one electrode to the
other by means of convection currents,
Ophcal properhes and x-rcxy patferns were de-

. el e
o T

 

 
 

 

T

termined for several newly encountered compounds
in the fluoride systems. High-temperature x-ray
diffraction was used to study the polymorphs of
Li,ZrF,.

Progress is reported in the efforts to obtain
high-temperature x-ray and neutron-diffraction data
on fused materials to serve as an aid in de’rermmmg
the molecular structure of the melts. A summary is
presented of the work on the electrical conductance,
density, and freezing-point depression measure-
ments with various alkali and heavy metal halides.
Self-diffusion coefficients are reported for sodium
and nitrate ion in fused sodium nitrate.

5. Corrosion Research

Examinations were made of several Inconel
forced-circulation loops operated with NaF-ZrF -
UF, (50-46-4 mole %) as the circulated fluid for
various times under otherwise identical conditions.
These loops, which were direct-resistance heated,
had o temperature differential of 200°F, a Reynolds
number of 10,000, a maximum fluid temperature of
1500°F, and a maximum wall temperature of 1600
to 1625°F. A curve obtained by plotting depth of
attack vs operating time exhibited the two-stage
type of attack found previously in thermal-convection
loops; that is, the initial rapid attack that occurred
while chemical equilibrium was being established
and the impurities were reacting was followed by
the slower mass-transfer type of attack. An at-
tack rate of 3 to 4 mils per 1000 hr was found for
the second stage of attack. In two similar loops
operated under the same conditions, except that
one was heated in a gas furnace and the other was

heated by electrical resistance of the tubing, the

depth of attack was not found to be affected by the
heating method, Another similar loop with a maxi-
mum wall temperature approximately 100°F higher

fluoride mixture temperature.

Additional data have been accumuluted on mcss::':
“force ’irculatlon syss
‘ ‘perature “dif-
ferential and mass fransfer "was‘hfound but “in-
creqsmg ‘the oxnde contenf “of the “sodium did

transfer in sodwm-lncone! f ced—
tems. No correlcn‘lon befw '

     
 

  

increase the mass transfer. In an all-stainless-
steel loop in which sodium was circulated, there
was a mass-transferred layer that was 0.8 mil

PERIOD ENDING SEPTEMBER 10, 1955

thick, in contrast to a 9-mil-thick layer found in an
Inconel loop with a stainless steel cold leg.

In an effort to ascertain the cause of the erratic
results being obtained with Inconel thermal-
convection loops, several loops were cleaned by
different methods and then operated with NaF-
ZtF -UF, as the circulated fluid. No effects
attributable to the cleaning method could be
found. In other tests, heat was applied by direct
electrical resistance of the wall rather than by the
usual “‘clamshell”’ electric heaters, The resulis
confirmed a previous conclusion that the depth of
attack was not affected by the method of heating.
Since the forced-circulation loops had indicated
that the maximum loop wall temperature was a
significant variable, thermocouples were installed
on the hot legs of two standard lnconel thermal-
convection loops to study this effect. Preliminary
results indicate that the wall temperatures may
have been as much as 1670°F, in contrast to wall
temperatures of about 1600°F in the forced-circu-
lation loops. The higher temperature difference
between the wall and the fluid may account for the
attack in the thermal-convection loops being deeper
than in the forced-circulation loops.

A series of thermal-convection loops have also
been operated with small applied potentials., The
loop which circulated NaF-ZrF ,-UF, for 2000 hr
with a positive charge applied to the hot leg showed
only dbout one-half the depth of attack found in
the loop operated with a negative charge applied
to the hot leg. With a negative hot leg the attack
was about the same as that with no applied po-
tential. Inconel thermal-convection loops operated
with sodium gave results which confirmed those
obtained with forced-circulation loops.

Hot-pressed metal-bonded tungsten carbide

~ cermets were fested in NaF-ZrF -UF, and in
than the usual 1600 to 1625°F showed heavy sub-
surface-void attack to a depth of 18 mils. Thus
additional evidence was obtained that the wall
temperature is a more crmcol varmble fhan 15 the'_" -

sodium in seesaw apparatus; no measurable attack
was found on any specimens, Similar additional

tests were made of the Kentanium cermets. The
__best of the Kentcmlum cermets are being fabricated
into valve dlsks and seats for self-bonding tests

under service condmons.
Ruthenium was plated onto Inconel for tests of

" creep-rupture properties of the plated specimen.
'Prehmmary results indicate an increase in creep

rate and a decrecse in rupture life, m comparlson

‘with standard Inconel. Additional screening tests

were made of Inconel T-joints brazed with various
dlloys and exposed to fluoride mixtures and to

v

R W TS T T

 

 
 

 

 

ANP PROJECT PROGRESS REPORT

- sodium, and several brazing alloys on Inconel and

~ on stainless steel were tested in static lithium.
Al ‘the brczmg alloys tested showed poor corrosion

* resistance to lithium.

A static test of a Hastelloy B specimen in an

Inconel capsule containing a fluoride mixture
showed the Hastelloy B to be unattacked, but the
Inconel was attacked fo a depth of 8 mils, in con-
trast to a normal attack to a depth of 2 mils in a
static Inconel ccpsule without the Hcsfel]oy B
 specimen,
Experiments are under way with a boiling-sodium—
* Inconel system so that the effect of oxide-free
- sodium on mass transfer can be studied. In pre-
liminary experiments no mass transfer could be
~ detected.

Seesaw corrosion tests have been made on
Inconel capsules loaded with sodium-potassium-
lithium mixtures in which the lithium content was
varied from 2 to 30 wt %. The heaviest attack was
found in the hot section of each tube, and it varied
from 0.5 mil in the presence of 10% lithium to 2 to
3 mils in the presence of 5% lithium. Molybdenum,
vanadium, and niobium were tested in static
lithium at 1500°F, Molybdenum was unattacked,
niobium was only slightly attacked, and vanadium
showed grain-boundary penetration by an uniden-
tified phase to o depth of 2 mils.

Measurements of the oxidation rate of metallic
sodium at 25 and 48°C were extended to periods of
2 x 10% min, The data obtained do not fit any of
the postulated oxidation-rate theories, although
highly protective oxide films were formed.
Measurements of the oxidation rate of metallic
columbium confirmed the previously reported change
in rate law with time. Tentative conclusions have
been drawn . ‘concerning the origin of the change.

Improvements in technique have been made for

e studymg fused sodium hydroxide by spectrophoto-

- metric techmques. Measurements of corrosion in
fused sodlum hydroxlde have shown that water
~'vapor is an important inhibitor and that the pres-
ence or absence of a closed electrical circuit be-
V’rween the hot ‘and cold parts of the corroding
~ system has a significant effect on mass transfer.

© Measurements of the self-dissociation of sodium
hydroxude have been made for the first time in the
"~ absence of s:de reactions. Both the expected
decomposmon into water and sodium oxide and
 the postulated decomposition to produce hydrogen
have been confirmed,

A series of studies of thefour-componentfuel
NaF-LiF- ZrF4 -UF , (22-37.5-35.2-5mole %) exposed

for 100 hr in sealed ccpsules of Inconel in the

standard rocking furnace have indicated that this

mixture may be less corrosive than others being
considered, A series of mixtures of the NaF-KF-
LiF eutectic with UF; and UF, added were also
tested in Inconel in the rocking-furnace apparatus.
The data indicate that UF, is quite unstable under
these conditions, regardless of the original UF,-
to-UF, ratio. It also appears that considerable
disproportionation of UF ; must be expected in this
sy stem. ' '

In experiments for determining the effect of
chromium on the mass transfer of nickel in NaOH,
it was not obvious that the chromium was particu-
larly beneficial. However, it was shown that some
mechanism for preventing the loss of hydrogen

from the system ~ perhaps cladding the nickel

with some metal impervious to hydrogen - might
be quite berieficial. :

6. Metallurgy and Ceramics

Mechanical property investigations of Inconel
have continued with stress-rupture tests of Z-in.
tubing in argon and in fused salts at 1300, 1500,
and 1650°F. A comparison of creep-rupture dnd
tube-burst data showed similar rupture times for
0.060-in.-wall tubing and 0.060-in.-thick sheet in
argon and in NaF-ZrF -UF, (50-46-4 mole %) at
1500°F. The data being obtained indicate that
the presence of an axial stress in the tubing does
not appreciably affect the time to failure., Data
obtained for tubing with 0.010-, 0.020-, and
0.040-in. walls show, in comparison with data for
0.060-in.-wall tubing, that the thinner the wall the
shorter the rupture life at comparable stresses.

An evaluation test in fused-salt fuel of an

Inconel bellows with a welded diaphragm showed

only normal corrosion attack in the weld areas
and no cracks that resulted from flexmg of the

bellows. A determination of the extent of inter-

action between beryllium and Inconel in contaci
at high temperatures in an inert environment was
made. The results indicated that an intermetdllic
layer was formed that would be detrimental to the
load-carrying capacity of the Inconel. This in-
formation was needed in the design of the high-
temperature critical experiment, :

Additional tests were made of nickel-molybdenum
alloys containing titanium, aluminum, vanadium,

 

 
 

zirconium, columbium, or chromium which con-

firmed the previously reported embrittlement of
these ternary alloys after a long heat treatment at
elevated temperatures in hydrogen. The ductility
of the alloy can be restored by a high-temperature
anneal in vacuum. In the study of factors affect-
ing the fabricability of Hastelloy B, it was found
that canning the extrusion billets with Inconel

teduced the pressure required during the extrusion

and at the some time clad the alloy tubing with a
heat-resistant alloy.

A series of creep-rupture tests of solution-
annealed, 0.060-in.-thick, Hastelloy B sheet at
1500 and 1650°F in NaF- ZrF ,-UF ; (50-46-4 mole
%) were completed and de5|gn curves were pre-
pared. The influence of aging heat treatments on
the creep-rupture properties of Hastelloy B is

being studied in the range 1300 to 1800°F in

argon. The creep-rupture properties at 1800°F

‘are not substantially affected by aging, since at

this temperature a single-phase alloy exists. How-
ever, at 1500°F a second phase appears to exist

in aged specimens which increases the rupture

life at high stresses. Short-time tensile tests

' conducted after long-time aging of cold-worked
Hastelloy B at high temperatures indicate that
‘tesidual stresses from the cold-working operation
‘are quite detrimental to ductility. Data showed

the ductilities of cold-worked specimens to be
considerably lower than those of specimens an-

nealed before aging. Microstructural studies indi-

¢ate that cold work induces precipitation in larger
quantities and perhaps in smaller particles than
does annealing.

‘Additional oxudanon-resns’rance tests of high-
temperature brazmg alloys were ‘conducted, and

melting-point studies are being made by using

sintered, conical samples. Experiments are under

way in an attempt to find an alloy for brazing a
boroh carbide compact to an Inconel envelope.

The alloys lnveshga’red thus far that wet the

boron carbide also react with it to form brittle
bonds that crack upon coohng. The boron carbide—
Inconel assembly is requu'ed for rodlqhon shield-
ing in fused-salt pumps. A'success ful method ‘was
es’rabhshed for brqzmg Kenmmum cermet vc]ve"
seats to lnconel sfrucfural components. Nnckel |s' N

from the different coeffimenfs of thermal expcnsxon
of Inconel and the cermet,
The lnconel core-shell assembly for the high-

PERIOD ENDING SEPTEMBER 10, 1955

temperature ‘critical experiment was fabricated
after the development of experimental techniques
for minimizing distortion. The fabrication of a
third, 500-kw, NaK-to-air radiator is under way,
The design of the radiator is essentially the same
as that used previously, but improved fabrication
techniques are being utilized, Presintered rings
are to be used for the preplacement of brazing
alloy. One tube bundle of the fuel-to-NaK inter-
mediate heat exchanger No. 3 has been completed,
and construction of the second tube bundle is under
way. Design modifications for this heat exchanger
have included the use of larger diameter fubing
(0.250 in.) and the elimination of right-angle
corners of header components to provide better
stress distribution. A second liquid-metal-to-air
radiator was fabricated for the Cornell Aeroncuflcal
L aboratory,

An eddy-current method for flaw detection in
[ow-conductivity tubing is being investigated.
Studies have shown that an ultrasonic method for
the inspection of small-diameter tubing is suf-
ficiently sensitive to detect the types of flaws
encountered to date.

Corrosion-erosion in graphite-hydrogen systems
at 2400°C, with a hydrogen velocity of Mach 0.15,
is being investigated., Present indications are
that the slight weight losses observed were due to
the small amounts of water vapor in the gas.

Two rare-earth-oxide control rod assemblies were
prepared for critical experiments; flux-detector
spacers of calcium fluoride and alumina were pro-
duced; dysprosium oxide disks were prepared for
use in measuring thermal flux; and a method for
preparing europium oxide wafers was investigated,
A study of the feasibility of synthesizing MozBs
and B,C was started, and the optimum pressing
conditions for pelletizing fluoride fuels are being
determined.

Investigations of diffusion barriers for use be-
tween Inconel and columbium have shown only
tantalum and copper to be useful. Tests for de-
termining 'rhe more saflsfoctory of the two barrier

materials are under way. Mixtures of columbium

and UO,, being considered for use as fuel elements

“were tested at 1500 and 1832°F for 100 hr. A

solid solution of Cb-U and a pinkish phase, as yet

Jumdenhfned formed in the samples.

The deformation patterns obtained in two- and
three-ply extrusions of metal tubing were studied,
and limits were established on the metal ratios

 
 

1

 

 

ANP PROJECT PROGRESS REPORT

and on the configurations of the billets that can be
successfully extruded, The creep performance of

" lead-calcium was found to depend upon the care

with which the master alloys were made. This
implies that an accurate knowledge of the amount

“of alloyed calcium rather than total calcium is

required.  Therefore, chemical analyses, which
give a measure of the total calcium, were found to
be an unreliable index for predicting the creep
properties of the alloy.

The fabrication of boron-containing materials
for use as neutron shielding is being studied. At
present, attempts are being made to thermally bond
boron-containing layers to Inconel, The use of
B;C tiles is also being considered for locations
where therma[ bonds are not necessary,

7. Heut Trunsfer and Phys:cul Properties

Forced-convectlon hec'r transfer medasurements

were made for ‘molten NaF-KF-LiF- UF, (11.2-41-

45,3-2.5 mole %) which was flowing furbulem‘ly

through type 316 stainless steel tubes. The re-
sults were comparable with those obtained pre-
viously in an Inconel tube and were thus 40% below
the general turbulent-flow heat transfer correlation,
A check on the experimental apparatus was ob-
tained by operating it with water as the heat trans-
fer medium, and the data were in good agreement
with the general turbulent-flow correlation, The
low heat transfer valuves obtained with NaF-KF-
LiF-UF, therefore appear to be real.

The heat transfer and friction characteristics of
a full-scale model of the ART heat exchanger
were determined with and without the presence of
tube spacers. These results were compared with
conventional heat transfer and friction relations
for simple duct systems. As was to be expected,
both the heat transfer coefficients and the friction
factors decreased upon removal of the tube spacers,
However,' when the spacers were removed, the

| tubes were not held rigadly and channeling occurred

in the flow pch‘ern.
A summcry of hydrodynamic research on models

.of the 18- and 2l-in. ART cores has been pre-
pared.  Rotational and axial flow patterns, as
‘well as various entrance conditions, were studied,

One core, which had a low ratio of flow cross-

sectional area at the equator to flow cross-sectional
“area at the inlet, was characterized by uniform and

steady flow.
~ The temperature distributions within fluids flow-

ing through converging and diverging channels were

.experimenfcny determined in the volume-heat-

source system, Information on the transient be-
havior of the wall temperatures, as well as on the
asymmetry of wall temperature profiles, was ob-
tained for these uncooled channels. A report has
been prepared which describes applications to
general convection problems of previously de-
veloped mathematical temperature solutions for
forced-convection systems having volume heat
sources within the fluids.

The enthalpies and heat capacities of LiF-KF
{50-50 mole %) were determined in the liquid and
solid states, The viscosities of eight fluoride
mixtures were determined. A mixture of RbF and
LiF (57-43 mole %) yielded a viscosity of 9.0
centipoises at 500°C and 3.4 centipoises at 650°C,
An RbF counterpart of the ART fuel was formu-
lated, This mixture, whose composition is RbF-
ZtF ,-UF, (48-48-4 mole %), had a viscosity of
9.5 centipoises at 550°C and 3.1 centipoises at
850°C, Its kinematic viscosity was found to be
about 20% lower than that of the corresponding
NaF-ZrF ,-UF , mixture.

8. Radiation Damage

The high-temperature forced-circulation fluoride-
fuel loop recently operated in a horizontal beam
hole in the LITR was examined metallographically.
Corrosion of the Inconel tubing by the circulating
fluoride-fuel mixture was found to be fow and to
be substantially the same as that found previously
in capsules exposed in the MTR. No increases in
corrosion attack because of irradiation and no
other unusual effects were found. In the portion of
the loop that showed the maximum corrosion, the
average penetration was 1 mil and the maximum
was 2.5 mils. In general, the changes in the
Inconel were those expected in specimens sub-
jected to the heat treatment imposed by operation
of the loop. The average corrosion for the entire
loop was 0.5 mil, and no deposits of mass-trans-
ferred material were found, The fuel circulated
in this loop was NaF-ZrF --UF4 (62.5-12.5-5
mole %). The fission power generated in the
fuel was calculated to be 2.8 kw, and the maximum
power density was 0.4 kw/cm3,

The miniature in-pile loop was operated in a

“vertical position in the LITR, but the experiment
was terminated after about 30 hr because a faulty’

pump motor prevented the maintenance of steady
fuel flow. The test was incomplete as a corrosion
study, but it was possible to make a fairly thorough

1
”
T T

 

 

 

g%

 

 

 
 

 

 

 

 

study of the in-pile characteristics of the loop.
The necessary design modifications are being
made, and a new loop is being fabricated,

The Reactor Experimental Review Committee
has approved the insertion of the pressurized
stress-corrosion apparatus in HB-3 in the LITR,
and specimen assemblies for a series of tests are
being filled with fuel. Bench tests are in progress
on an apparatus for insertion in the MTR that is
designed to test creep in two nonfuel atmospheres.
An MTR (tensile) creep test apparatus irradiated
during two reactor cycles is being returned to
ORNL for postirradiation measurements., The
bench equivalent of this apparatus has been
assembled and tests have been started.

The maximum high-energy neutron flux in HB-3
of the MTR was measured to be 3.1 x 103 fast
neutrons/cm2.sec, and the thermal-neutron flux
was found to be 2.3 x 104 neutrons/cmZ.sec. A
flux-depression experiment in HB-3 indicated a
lack of sensitivity of flux depression to the kind
of nuclear absorber used. The power to be ex-
pected in the MTR in-pile loop was estimated from
the measurements to be 31 to 36 kw.

Analyses of reactor-grade beryliium obtained from
The Brush Beryllium Co. and the R. D, MacKay
Company showed that the predominant source of
gamma activity after long irradiation followed by a
few days decay is Sc48, The quantity of scandium
present is so small that it cannot be found by
chemical analyses with a limit of detection of

200 ppm.

f9. Analyhcal Chemlstry of Reactor Materials

Modifications of the n-butyl bromide method for
the determination of oxygen in sodium were evalu-
ated. The modifications included the addition of
a column of silica gel and diatomaceous earth

for the rapid purification and desiccation of re-

agents and an lmproved apparatus in which the

‘reaction between sodlum and butyl bromide could

be carried ouf in an afmosphere of argon, Although

7 the oxygen content of the mu|or|'ry of the samples
L which were used in 'reshng these modn‘lcuhons_‘ -
- ‘was in excess of 200 ppm, concem‘ratlons of the =

order of 20 ’ro 40 ppm of ‘oxygen were found m' o

4specro|ly prepared sodium,
:Tests were made of fhe method for the determina-

tion of oxygen in sodium by titration of the Nazo,_'

that remains after vacuum distillation of the sodium
metal. An apparatus was constructed that was
similar to one developed by the Argonne National

PERIOD ENDING SEPTEMBER 10, 1955

L.aboratory. Preliminary results on sodium sampled
at 1200°F showed an oxygen content of the order
of 50 ppm. Since samples of sodium at high tem-
peratures can be obtained with this apparatus and
since fow levels of oxygen can be detected, the
apparatus is to be attached directly to a forced-
circulation high-temperature-differential sodium
loop so that analyses can be carried out during
operation of the loop.

Development of a volumetric method for the de-
termination of zirconium in fluoride salts was
completed, In this method zirconium is converted
to a stable complex by the addition of an excess
of disodium dihydrogen ethylenediaminetetraacetate
(EDTA) to a dilute H2804 solution containing
zirconium, The excess EDTA is ftitrated with
trivalent iron to a disodium-1,2-dihydroxybenzene-
3,4-disulfonate end point at a pH of 4.8, Titration
can be conducted in the presence of as much as
0.1 M of fluoride ion by first complexing the
fluoride ion with beryllium, Slight modifications
in the procedure make the method also applicable
to the determination of zirconium in the presence
of moderate amounts of trivalent iron, divalent
nickel, and trivalent chromium.

The apparatus for the determination of uranium
metal in mixtures of fluoride salts by decomposi-
tion of the hydride in an atmosphere of oxygen at
reduced pressures was modified to include two

combustion tubes so that one sample can be oxi-

dized while a second is being converted to the
hydride.

Analytical assistance was given in a study of
the rate of elimination of atmospheric gases from
a dry box with argon. The most efficient flushing
action was found to be the fairly rapid injection of
argon at the bottom of the dry box, without supple-
mentary agitation. With an argon flow rafe of 25
cfh, the concentration of oxygen in the atmosphere
of a 21-ft3 dry box was reduced by a factor of 100

by flushing with two volumes of argon.

Investigation was continued of the application
of the bromination method to the determination of
oxygen in ZrF ‘and its mixtures with alkali-metal

fluoride sal'rs._ Incomplete removal of oxygen wdas
observed for samples of pure ZrO, after bromina-

tion for 6 hr at 950°C,

“In the bromination method for the determination

of oxygen in metal oxides, the oxygen is converted

to CO and then oxidized to C02 for measurement.
A method for the direct measurement of CO was
studied in which the CO is absorbed in an aqueous

 
 

 

 

ANP PROJECT PROGRESS REPORT

solution of PdC|2 and KCl. The net increase in
- hydrogen ion concentration of this solution is a
function of the CO present. Excellent titration
curves were obtained only when an amount of Kl
in excess of ‘the PdCI present was added to the
_soluhon prIOI’ to h’rroflon with a standard base.

‘IO. Recovery and Reprocessing of Reactor Fuel

A _,rey_rl_ewr of the deslgn and construction prob-
- lems involved in the completion of the pilot plant

" for fhe fe'coikéfy of fused-salt fuels has indicated

~that a construction completion date near the end
of February 1956 will be more realistic than the
December 1955 date given previously. An engi-
neering flow shee'r was issved, and approximately
65% of the process equipment items are on hand
of are in some stage of procurement or fabrication,

The dump tank containing the ARE fuel was
moved, uneventfully, from the ARE building to
the pilot plant building. Methods have been de-

: 4fiV|sed for removing the fused salt from the dump

‘tank and from other types of containers for charg-
ing into the fluorination vessel.

Direct-resistance heating of transfer lines was
found to be satisfactory for preventing plugging,
except at fittings, where supplemental external
heating will be required, A freeze valve was de-
signed for closing the transfer lines to and from
the fluorinator, After 15 cycles of freezing and
thawing, this valve, when frozen, held against a
pressure of 20 psig without leaking.

An improved procedure for decontamination of the
UF6 product of the fluorination step was developed
which involves the absorption of the UF , on NaF
at 100°C and desorption by heating to 400°C, with
the product gas passing through a second bed of
NaF before collection of the UF, in a cold trap.
Since in this two-bed process, in contrast to the
previous process in which a single absorbent bed
was used, the fission products never enter the
product-collection system, decontamination factors
of greater than 105 were obtained in re-used equip-
ment. Preliminary results indicate that nitrogen
may be used as a sweep gas in both the fluorina-
~“tion and the NaF absorption and desorption steps

Sote reduce the amount of fluorine required for
'processmg.

s . ,' PART Hl SHIEL DING RESEARCH
' o, Sh|e|d Design

A survey of fhe welghts of spherically symmetric
qxj_n{ffishl_el_ds,for_ circulating-fuel reactors was made

10

for a range in dose rates of 0.1 to 10 rem/hr at a

distance of 50 ft and a range in reactor power of
100 to 300 Mw. An estimate was obtained for the
added weight of an NaK-to-NaK secondary heat
exchanger and its shielding, The additional weight
was found to vary sharply with the manner in
which the dose rate was divided between the

"secondary heat exchanger and the reactor and with

the absolute value of the sodium activation. The
chief sources of radiation in the 300-Mw circulating-
fuel reactor for the NJ- 1 power plant were determined.

New data recently obtained at the TSF and 'flie
LTSF are being used in a parametric shield weight

study for a 300-Mw circulating-fuel reactor in

which the important reactor dimensions are varied.
Differences between the shield test mockups and
the design reactors are accounted for 'o_h“the' _boé_is
of the present understanding of the sources of
radiation in each. As a result, shield weight
dependence upon reactor and shield d|mens:ons
and materials can now be calculated with greater
certainty, The results obtained to date indicate
that divided shield weights can be significantly
reduced by increased shield-shaping based upon

TSF and LTSF data and analyses,
12. Lid Tank Shielding Facility

The static source tests of the second series of
the circulating-fuel reflector-moderated-reactor and
shield (RMR-shield) mockup experiments have been
completed. The final tests included neutron und
gamma-ray measurements beyond the mockups to
determine the effect of placing an intermediate or
high atomic weight material immediately behind the
beryllium reflector, varying the thickness of the
reflector, and distributing the lead gamma-ray
shield in borated water,
RMR-shield, little, if any, welght saving results
from adding bismuth to the outer region of the
reflector rather than using lead in the shield. The
addition of a 2-in.-thick layer of copper in the
same region was insufficient to effect an appreci-
able weight saving, but there was ev:dence that
there might be enough self-absorption of ccpfure
gamma rays in a 4-in,-thick copper layer to show a
valuable weight saving,
thickness (8, 12, and 16 in.) did not have an
appreciable effect, :

The study of distributing the lead gamma-ray
shield in borated water showed that, for lead
thicknesses up to 5 in., there would probably be

 

For a typical 300-Mw

VYarying the beryfhum.

 

[ .!),‘_i . ’ '

 

 
L)

. . . - ——

An

 

 

 

T

o completed

no weight saving as a result of distributing the
lead rather than placing it in one piece but that
there might be an appreciable weight saving as
a result of distributing the lead beyond the first
5-in. layer, The secondary gamma-ray dose rate
produced in the lead and borated-water shield fell
off at the same rate as the thermal neutron flux

~ and thus was apparently caused by thermal-neutron

captures in the shield.

~Sodium activation tests were performed to de-
termine the activation of the coolant in the heat
exchanger region as a function of the heat ex-
changer thickness, the boron curtain thickness and
distribution, and the reflector thickness, Results
of representative tests showed that the sodium
activation in the heat exchanger was increased in
going from a 4-in. thickness to a 6-in. thickness.
The probability of escape of the resonance neu-
trons was reduced by the increased thickness; the
probability may be increased, however, by dis-
tributing the boron curtain through the heat ex-
changer region. An increase in the reflector thick-
ness from 8 to 12 in. decreased the sodium acti-
vation by 80%. A gamma-ray shield of copper
placed between the beryllium reflector and the
first boron curtain increased the sodium activation
by a factor of about 6.

13. Bulk Shleldmg Fucull'ry

A part of the experiment designed for determining
the gross fission-product gamma-ray spectrum was
completed.  Small samples of enriched uranium
were irradiated in the ORNL Graphite Reactor for
time intervals ranging from 1 to 8 sec, and the
gross fission-product photon spectrum was studied
by using the mulhple-crysfol gammo-ray specfrom-
eter. The decay of six energy groups, covering

PERIOD ENDING SEPTEMBER 10, 1955

from the reactor were integrated. The integrated
dose {0.49 mrep/hrew) compared surprisingly well
with the total dose measured by the Hurst-type
dosimeter (0.52 mrep/hrew),

Measurements were also made through 70 cm of
water at a point 19.6 c¢cm from the center line of
the reactor and through 5 cm of water on the center
line of the reactor. In all cases the measured dose
remained constant, within the statistical devia-
tion, when the collimator was pointed at the active
lattice of the reactor.

14. Tower Shielding Facility
The results of Phase | of the TSF differential

shielding experiments have been incorporated in
the development of a procedure for optimizing the
neutron shield of a divided aircraft shield. In this
optimization procedure the neutron shield at the
reactor is divided into N conical shells. The
thickness of the nth conical shell is then denoted
by T, (n=1,2,...,N). Thecrew shield is assumed
to be cylindrically shaped, with a rear thickness
T, and a front and a side thickness T .. The pro-
cedure then consists in (1) expressing both the
total weight of the neutron shield and the dose
rate at the center of the crew compartment as
functions of the T 's, T, and T_; (2) using the
method of Lagrange multipliers to obtain the equa-
tions which the T ’'s, T, and T must satisfy in
order that the weight be a minimum for a specified
total dose rate at the center of the crew com-
partment; and {3) developing an iterative procedure
for the solution of these equations,

There are still some gaps in the experimental
input data required for the optimization. Where it
" has been feasible, these gaps have been filled by

extendmg the existing data by qualitative theoreti-

the range of 0.28 to 5.0 Mev, was foliowed from

5 to 150 sec after flssnon.

lease per fission in the time interval cnd energy”
range described was found to be about 1.5 Mev.’

The total energy re-

Additional measurements of fast-neutron dose as
a funcflon of qngle in a water shueld have been

Tower
analysis,

Shleldmg FdClhty. It is
which involves simple

~more meaningful than that reported in the pre-
vious ANP Quarterly for which data obtained with

the GE-ANP mockup were used.
The angular-distribution measurements made on
the reactor center line at a distance of 70 cm

| The ‘tesulting data were analyzed
.~ for use with the shleld ophmlzatlon ‘'studies of 'rhe_'___
felt that thus_ .

geometry, is )
 for a crew shleld fh!ckness of less fhon 5 cm of

" cal considerations of the attenuation processes

“involved. An important limitation does exist, how-
‘ever, in the use of this optimization procedure,
In the TSF experiments, scattered dose rate meas-

or T, smaller
Therefore, for T _ and
Tr less than 5 cm, “exirapolations must be used
Indications are that, for such small thicknesses,
the requcn‘lon Iengfhs chcmge apprecmb!y, hence,
"it does not appear ‘advisable to use the procedure

suremenfs were. nof ’raken for T
flmn abouf 5 cm of water.'

wa’rer.

A neutron shield optimization calculation for a
typical reactor and shield configuration was made
by using the above procedure, This calculation

11

 

 

 
~indicated that the procedure is quite satisfactory

and yields results which converge rapidly enough

for the solution to be obtained in a reasonable

length of time by hand calculation. In going from

~ the first to the third iteration, the weight of the

calculated shield, in this sample calculation, was
reduced from 11.3 to 9.6 tons,

 

 

 

 

12

 

A further investigation of the GE-ANP R1 re-
actor and crew shield mockups is under way.
Measurements of gamma-ray doses inside the crew
compartment mockup have been completed, and,
at present, a study of the distribution of gamma-ray
infensities in air around the reactor shield is
being made, '

 

 

 

 

 
 

 

 

 

Part |

REACTOR THEORY, COMPONENT DEVELOPMENT,

AND CONSTRUCTION

 

 
 

 

”

‘system; and all work required on roads, grounds,

* which design could not be completed for inclusion

‘process lines, and process equipment,

. hessee. Their
:fh the Chtcago_ Brlcfge & Iron Cq.‘enga

0 1rerquzlre that

1. REFLECTOR-MODERATED REACTOR

 

E. S. Bettis A. P. Fraas
~ W, G, Piper
Aircraft Reactor Engineering Division
A, M, Perry
Electronuclear Research Division
AIRCRAFT REACTOR TEST FACILITY The package 1 drawings for the southwest area
F. R, McQuulkm are being withheld pending design of the complex
Aircraft Reac’ror Engmeermg Division piping and equipment (package 3 work) that will

be installed within this area. Final release of the

Consfruchon OF‘ the ART Facility has been package 1 drawings for this area is scheduled for
divided into three ‘‘packages’” of work. Package 1 October 15, 1955. Plans and specifications for

includes alterations and additions to Building ) K bei 4
+ ORNL.
7503; construction of the 7503 cell, air duct, package 2 work are being made a

stack, adsorber tank, spectrometer facility, and SHIELDING EXPERIMENT FACILITY
fuel storage tank; a portion of the electrical power , R. D. Schultheiss

. ‘ . Aircraft Reactor Engineering Division
and fencing. Package 2 consists in additional 9 S

mechanical, electrical, and structural work for Recent tests made af the Tower Shielding

Facility indicated that provision should be made
for the measurement of the gamma-ray spectrum
of the ART as a function of the angle of emission
from the reactor shield surface. It was decided
that four collimated beams radiating from an
auxiliary service and utility equipment, and lines equatorial point at the surface of the water shield
up to the cell and to the vent-gas piping system. at angles of from 0 to 70 deg from the radial
Package 3 includes the installation, by ORNL direction would serve to give the essential data.

forces, of the experimental instruments, controls, One additional collimated radial beam will be
available from an equatorial point at the surface

of the reactor pressure shell; this beam will be
used only during low-power operation. The layout
required for providing these beams and the fa-

prime contractors who expressed prior interest cilities for measuring them are shown in Fig. 1.1.
submitted bids, which were opened August 16. In addition to the facilities shown in Fig. 1.1, a
The bids ranged from $765,835 to $869,560, with ~~ gamma-ray dosimeter will be located on the roof
bids for the cell included; the cell bids ranged “bovefhefefidor o

from $229,000 to $280,000. The contract was pEE

awarded on August 19 to the low total bidder, AIRCRAFT REACTOR TEST DES|GN

the V. L. Nlcholson Compcny, Knoxv:lle} Ten- S TR |
i Included $264,373 for» e

in package 1. The work included in packages 1
and 2 is to be performed by an outside contractor.
The work items included in package 2 are the
diesel-generator auxiliary power supply facility,

Drawings and specifications for package 1 were
prepared by the K-25 Engineering Division for
lump-sum prime contracting. All six prospective

Reactor Desngn
A P Fracs

  
  
   

 

      
 

    

constructton N ol C
buddlng cd i |on, ; _ days, shown in Flgs. T, 24 VT 3 “and T 4 7and Table 1.1
commencing August 29 1955, The cell and gives the key dimensional data. The design of

 
  
 

   

southwest corner installation are to be completed | the pump—expansion tank region, which includes
by June 1, 1956. / the sodium-to-NaK heat exchanger, has been
I‘//
AN M :
(\:}5 {;_ ]5

 

 

R IR FTENT  TET

T T

TR

 
R e e e

i

 

 

G o ik et

 

' ANP PROJECT PROGRESS REPORT

 

 

el RS

 

 

 

 

 

 

|

1
L SPECTROMETER l \
ERECIERSGILS ROOM |

 

 

 

 

'COLLIMATOR TUBES ~=—

///—__

 

 

 

 

 

 

REACTOR

\
\\ st e

T

Fig. 1.1. ART Shielding Experiment Facility.

worked out in a fashion which seems to be satis-
factory from all standpoints. A one-half scale
plastic model of that region has been completed

‘to facilitate ex_qmination' of fabrication, stress,
_and flmd-flow problems. Work has started on a
. full-scale aluminum model of this same assembly;
. with aluminum, the procurement and machining
- time will be reduced drastically from the time that
- would be required with Inconel, and yet the
- fabrication pt

ion problems that will arise in the pro-

   

: jif‘i'-f-.f"*"f-'gfil(ui:firfi''o/r': ‘of this model will be the same as those

  

16

L L e i L e L

CELL

 

 

7
7
V

i

to be involved in the fabrlcahonofthe ART]t |
is expected that many welding problems will be

revealed and that modifications can then '_be made
to facilitate fabrication with Inconel. The alu-

minum model will also be used for flow tests with

- . oo
water and stress analyses with strain gages or

stress-coat paint. The remaining design work on

subassemblies is sufficiently well along fhaf'
arrangements for procurement of the Inconel and
other parts for the ART began on August 1.

Fabrication of the various shells, such as the

ORNL-LR-OWG 9178

 

-
3
.
.
oo
. F
B
i
« ¥

“

>}

L C
i ol U

e

 

 

 

 

 
e

    
  
 
      
   
  

e PERIOD

 

B N e L

Y

LINER

FUEL—TO—NaK HEAT!
EXCHANGER |

/—\

  
   
   
 

NaK {COLD) ==

   
   
   
 
 
  

NaK{HOT) mmp

FUEL PUMP

 

REFLECTOR-MODERATOR

 
 

FUEL TO
L PUMP

)

  
    
 

   

i

 
    
  
   
 
  

ANNULUS

 

 

 

 

FUEL EXPANSION TANK

ISLAND

    
 
   

SLEEVE * CONTROL ROD

SCDIUM

   
 

INNER CORE SHELL

       
 

  
 

Ne PUMP -

    
 

 

 
 
 

    

 

 

— B SPACER WIRE ,{
Na RETURN-K :
1 . OUTER CORE SHELL
Na INLET

 

 

Na—TO—NaK -
HEAT EXCHANGER

i i GG

THERMOCOUPLE

  

DING SEPTEMBER 10, 1955

o D

o
5 ,
S e
T PN
T i
7 f‘)\\‘,/"’ i
L ¥ “ !
et N
e
e
Pt
!y
! .
o
L
; i
< : é
il
—
< i ]
£ i
- i
TN P
Y P
- L
R

 

 

T ! , SR A
\"'\.“‘ l ,;‘;‘ - e
vy Fig. 1.2, Vertical Section Through Reactor.
17
I\

 

 

 

T Y RN "TPR N i —— i e TR T TS
 

 

 

- e e et o e e g A S T 8 U T e

 

 

»
-

 

 
 

 

 

 

ORNL-LR-DWG 8835

 

 

 

 

Lo

T T TE VT LR

FUEL TO

\,

S
N <5
TR S

N 77 TR A T L
s
,

 

 

 

 

 
  
    
 
 
 

 

“pressure shell, refls
ticularly time-c
“and therefore procureme

   

 

Thgmcnn and_auxil _
the ART are i .. sign

    

 

E:]AC hh;cfld |E:tor

consists of \'ryj;-)év3fill(‘);:s;'c‘zifil'ess-steel-'cIrqd'-cbpifie“r o

(0.0025 in.-0.005 in,—0.0025 in.) fins spcced_ 15

 

 

  
     

 
  

core Units for

T wE

 

Fuehte ok Hoot Exchongr Desan
R. D. Schultheiss

s Ajperaft Réqctor :Eng'ineering Ddi\'rilsiér‘l

  

'Design calculations and the layout drawings for
the main heat exchanger were completed, and a

19

B T I B T T O T v eer o

 

W T S 7 e ST £

 

TTeT

IR TR " I Wy

 
.

 

 

 

 

 

 

 

 
  
     

0. ORNL-LR-DWG 8837

 

Ne RETURN FROM
MODERATOR

    
    
    
     
  
 
  
    
   
   
 

 

   

s

A

 

_ CONTROL ROD
=% THIMBLE

ISLAND EXTENSION-"|

 

 

FUEL PUMP BARREL

 
  

i
{
|
i
|

i
_——FUEL EXPANSION TANK

 

PERIPHERAL RING

/
/
s

   
 

MODERATOR AND ISLAND \ /

Fi.g. 1.4. Horizontal Section Through Sodium Pumps.

selection of a heat exchanger was made. The 'i'ig. ‘The design data are presented in Table 1.4,

~design selected is similar to the designs previ- The new layout makes it possible to halve the
 ously described except that the tube configuration ~ number of heat exchanger tube bundles without
. in the vicinity of the header sheets has been doubling the number of jigs. This has the ad-
modified so that all tubes in a given layer have vantage of halving the number of NaK pipes
~the “sdm‘e_ shape and can be made with a single penetrating the reactor pressure shell,

  

 

@y,

-

ay

 

 

 

 

 

 

 
 

 

 

 

"

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 1.1. REACTOR DIMENSIONS

 

REACTOR CROSS-SECTION EQUATORIAL RADII (in.)

Control rod thimble

Inside ' 0.750
Thickness ’ . 0.062
Outside - 0.812
Sodium passage
Inside 7 0.812
Thickness 0.094
Outside 7 0.906
Beryllium v V o | |
Inside ' 0.906
Thickness ' 4219
Outside 5,125
Sodium passage
Inside 5.125
‘Thickness 0.]é5 |
Outside - 5.250
Inconel shell (inner core shell) -
Inside 5.250
Thickness 0.125
Outside - 5.375
Fuel
Inside - .5._3'7"5
Thickness 5,125
Outside 10.500
Ovuter Inconel core shell | |
inside 10.500
Thickness o 0.125
Outside ' 10.625
Sodium pessage
Inside - ' 10.625
Thickness Coee e 0,094
Qutside 10.719

   

Beryllium reflector ™~ 7

 

lnside - » . ‘1,/‘( ;;,;'_m.-;. flA-\-\ 10'7_19 - l—
o 10949

 

2

nconel shell

Thickness N » ' ©0.250

Qutside ' 22.0:43

 

 
 

      
   

 

 

B4C tile
inside
Thickness
Qutside

Helium gap
Inside
Thickness
Qutside

Quter reflector shell
Inside
Thickness
Outside

Spacer thickness

Tangent to first heat exchanger fube
Tube radius

Center line of first tube

Eleven 0,2175-in, spaces

Center line of twelfth tube

Tube radius

Spacer

Channel
Inside
Thickness
Qutside

Gap
Inside
Thickness
Outside

Boron- jacket
Inside
Thickness
Outside

BAC tile
Inside
Thickness
Outside

Helium gap
Inside
Thickness
Outside
Pressure shell liner
Inside
Thickness
Outside

 

22.043
0.375
22.418

22.418
0.020
22.438

22,438
0.062
22.500

0.015
22.515
0.094
22,609
2.392
25.001
0.094
0.015

25.110
0.125
25.235

25.235
0.040
25,275

25.275
0.062

25.337

25.337
0.313

125.650

25.650
0.020
25.670

25.670
0.375
26.045

21

T TR

o G S

T T ATy YT T IR T Y

 
" "ANP PROJECT PROGRESS REPORT

TABLE 1.1. (continued)

 

Vsc',%;l.ium passage
, UIAri-sr-i:de
B Th|ckness .
L Ou]f;ide '
Pfessure_ SHeIA-[
. [nslde
. Thickness

  

~ CORE

~ Diameter {inside of outer shell at
“équater), in.
|s land ou t5|d e dla r:hefer ,in,

Core inlet outside diameter, in,

 
   

iametet, in,

 

inlet area, in. 2

 

Core equatorial cross-sectional

areaqa, in.2

MODERATOR REGION

Volume of beryllium plus fuel, £r3
Volume of beryllium, £3

Cooling passage diameter, in.
Number of passages in island

Number of passages in reflector

FUEL SYSTEM

Fue!l volume, £t3
In 26-in.-long core

In inlet and outlet ducts
In expansion tank when 1/2 in. deep
In heat exchanger

‘In pump volutes

Total in main circuit

Fuel expansion tank
Volume, £43
Widfli, in,

Length, in.

SODIUM SYSTEM (f13)

Sodium volume

 

  

© .7 In annular passages

 

7 ""*"In heat exchanger

 

 

26.045
0.125
26.170

26,170
“1.000
27.170

21

10.75
1

- 6.81
58.7

' 256.2

28.2
24.99
0.187
100
288

3.21
1.41
0.08
2.84
0.84
8.38

0.5787
13.625
32.500

1.1
0.165

 

 

In island tubes

In reflector tubes

In inlet and return piping
In pumps and volutes

In first deck

In second deck

In external piping

In expansion tank

Total sedium volume
FUEL.TO-NaK HEAT EXCHANGER

Tube data, in.
Center-line spacing
Qutside diameter
Inside diameter
Wall thickness
Spacer thickness

Mean length
Equatorial crossing angle

Inlet and outlet pipe, in.
Inside diameter

Qutside diameter

Header sheet, in.
Thickness

Inside radius
Fuel volume, ft3
Number of tube bundles
Number of tubes per bundle, 12 X 24
Total number of tubes
Latitude of north header center line

LLatitude of south header center line

PUMP--EXPANSION TANK REGION

Vertical distance above equator, in.
Floor of fuel pump inlet passage
Bottom of lower deck
Top of lower deck
Bottom of upper deck
Top of upper deck
Top of sodium pump volute
Center line of fuel pump discharge
Center line of sodium pump

discharge
Top inside of fuel expansion tank

Inside of dome

0.064
0.233
0.077
0.175
0.260
0.308
0.019
0.042
2.45

0.2175
0.1875
0.13735
0.025
0.030
72.000

26°20°

2.375
2.875

0.375
3.812

2.84
12

288
3456
41°30”
47°

17.500
19.125
19,625
24.000
24.500
27.750
21.500
26.125

29.500
29.875

)

 

™

 

 
 

o

 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 1.1. (continued)

 

Qutside of dome

Top inside of sodium expansion
tank N

Top outside of sodium expansion
tank

Top of fuel pump mounting flange

Top of sodium pump mounting

flange

FUEL PUMPS
Center-line épacing, in.

Volute chamber, in.
Width
Length

"~ Height

Impeller speed, rpm
Estimated impeller weight, |b

Critical speed, rpm

Shaft data, in.
Diameter
Ovethang
Over-all length
Outside diameter between bearings

Qutside diameter below seal
Distance between bearings, in,

Impeller data, in,
Piameter
Discharge height

Inlet diameter
Lower journal bearing
outside diameter, in,
Thrust bearing height from

equator, in,

P

- Number of vanes in impeller o

  
 

- Diameter of bottc

‘ring, in.

 

- Centereline spacing, in.

Volute éhamber, in,
Width

- Diameter of top pos momngrmg,m. o

*Outside diameter of top flange, in,

20.875
24.312

34.812

47,000
50.220

21.000

13.625
32.500
4.375

2750
11
6000

2.250
14.750
31.500

2.375

2.250

12.000

5.750
1.000
3.500

13.400

48.187

    
   
 

 

 

Length

Height
Impeller speed, rpm
Estimated impeller weight, Ib

Critical speed, rpm

Shaft data, in.
Diameter
Over-all length
QOutside diameter between bearings

QOutside diameter below seal

Distance from center line of lower

bearing to center line of impeller, in.
Distance between bearings, in.
impeller data, in.

Diameter

Discharge height

Inlet inside diameter

Lower journal bearing outside

diameter, in.

Thrust bearing height above
equator, in.

Number of impeller vanes

Diameter of top positioning ring, in.

Diameter of bottom positioning
ring, in,.

Outside diameter of top flange, in,

SODIUM-TO-NaK HEAT EXCHANGER

Tube data, in.
Center-line spacing
Qutside diameter
Inside diameter
Wall thickness
Spacer thickness
Mean length

Number of bundles

Number of tubes per bundle, 15 x 20
Total number of tubes

Inlet and outlet pipe, in.

Inside diameter

QOutside diameter

8.687
2.500

2880
10
6000

2,250
31.500
2.375
2.250

13.300

12.060

5.750
0.500
3.500

3.400

51.907

10
6.200
6.190

10.000

0.2175

0.1875

0.1375

0.025

0.030
28

300
600

2.375
2.875

 

23

e e

e

 
 

 

 

ANP PROJECT PROGRESS REPORT

TABLE 1.2, RADIATOR DESIGN CONDITIONS |

 

 

Main | Auxi'iidry
Power, Mw 55 Y
NaK inlet temperature, °F 1500 00
NaK outlet temperature, OF 1070 900
~Total .N(c!'(__frlo\'.v at average temperature, cfs 10.45 g0
- Alr inlet temperature, °F B 100 w00
- ;-'-"Air' o:ufr_‘|“e'r‘_tem.;.:erafure, °F | 1128 N 810
~“Total air flow at inlet conditions to blower, cfm 179,000 22,800

~ Air pressure drop across radiator, in, H,0

5.58 . 5.84

 

TABLE 1.3. RADIATOR DESIGN DATA

 

 

Main | ‘_7:'_AL|'xE|i-c_1ry |
" Face area, ft2 6.25 ' 6.25
 Mean free area, ft2 3.66 3.66
Air mass velocity, Ib/ft2.sec 3.46 3.71
Collar plus tube wall thickness, in. 0.035 0.035
Fin area, f2 922 922
Collar area, ft2 5.2 5.2
Inside tube area, ft2 42.7 42.7
Mean tube area, ft2 44.2 44.2
Number of tubes 360 360
Number of rows 8 8
Air Reynolds number 1321 1590
NaK flow area, in.2 5.34 5.34
NaK Reynolds number 91,400 111,000
NaK mass velocity, Ib/ft2.sec 820.0 1211

 

Core Flow Studies

W. T. Furgerson G. D. Whifmon
E. C. Lindley W. J. Stelzman
A. M. Smith J. M. Trummel

Aircraft Reactor Engineering Division
Two approaches to the core hydrodynamics
problem are being investigated. Both make use
~of a vortex sheet in the annulus between the
“island ‘and reflector in an effort to get a high
o degree of mixing. In the first system the vortex
~~ _axes parallel the island, while in the second they

spiral helically downward around it. In both
systems the two fuel pump volutes discharge
tangentially into the core inlet to give a system
that is insensitive to the one-pump-out condition,

Fight series of tests have been made on the
axial vortex system in the metal core rig. The
tests constituted a systematic development of an
inlet-guide-vane and turbulence-generator design

‘which would produce radial velocities of sufficient

magnitude to keep the boundary-layer fluid mixed
with the free stream, |

 

 

-

 

 

 

 
 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

A EERE TABLE 1.4. FUEL-TO-NaK HEAT EXCHANGER DESIGN DATA

 

Tube diameter
Number of tubes per bundle
Number of bundles
Tube center-line spacing
Tube wal thickness
Tt‘Jbe array, square pitch?
'Meon 1u5e |éngth |
Fuei'?émper_aft.)re range
NaK temperature range
 Fuel pressure drop through heat exchanger
NaK .préssure drop thrbugh heat exchanger
Fuel Reynolds number in heat exchanger
NaK Reynolds number in heat exchanger
Fuel flow rate?
NaK flow rate?
Fuel volume in heat exchanger
NaK volume in heat exchanger tubes (not ihcluding headers)
Heat exchanger thickness (includes 0.015-in. side-wall clearance)

Limiting combined tube stresses® at tube wall temperatures

Log mean temperature difference

Estimated maximum tube wall temperature, neglecting secondary

Eeoting effects
Fuel mixture NaF-ZrF 4UF, {50-46-4 mole %) heat transfer coefficient
NaK (56% Na—44% K) heat transfer coefficuent
Capacity at design operating condlhons A

Inconel surface (tubes and channel) in contact wrth fuel

    

Inconel surface in _contact wflh NoK o

   

|ncone| volume of tubes in heot exchanger o

0.1875 in.

288

12

0._2]75 in.
0.025 in.

12 x 24

5.95 ft
1250 to 1600°F
1070 to 1500°F
43 psi

41 psi

4135
144,000
2.96 cfs
10.45 cfs
2.84 #t3
2.14 #3
2.61 in.

1125 psi at 1135°F
220 psi at 1535°F

136°F
1535°F

2090 Btu/hr+ft2:.°F

18,200 Btu/hr-ft2.0F

55 Mw

160,000 in.2 (1110 £12)
107, 500|n 2(746 ft2)
'i3170|n. (1 83ff3) o

 

       
   
  

 

- velocn‘y profll

. with swurl chonflaer

 
  

":aCalcul'dfiorJ\s
,"b

At meqn op ‘ cmng

   
  

CTube sfre ses’

 

“The effect of one design is shown by the axial
,of"“ Flgs. 1.5, 1.6, and’ ] 7'”
‘ dota obtained for th
inlet only, that is, ho  inlet

_wh:ch

component.

 

 
 
 
 

Flgure ]5 pre

   

rotational component centered on the core axis  gradient is

 

’:Qrcdlenfs, which,”

   
    

lS cpproxlmately three tlmes the‘hax:al
" This gives rlse to strong radial
as a resul'r of fluid
’-frlctlon, induce an adverse axial pressure gradient
gw‘de vanes or turbulators. The flow has o  along the island wall. This induced axial pressure

‘additive to the already existing

25

T T I T

s ikl

 
 

 

 

 

ANP PROJECT PROGRESS REPORT

ORNL-LR-DWG 9348

 
 
   
 
  
 

 

STATION <
STAT\ON 8
| NO DATA TAKEN BELOW
STATICN 6
' sTllT\ON
! E.Q,UATOR STATION 5
e
¢ -
2]
a
o >
C
2
wl
>
o
STATION 3
STATION 1

 

Fig. 1.5. Axial Velocity Profiles from Test

Series 1.

gradient caused by the divergence of the core,

and thus large areas of reverse flow occur next
to the island.

The results of tests of the core with inlet guide
vanes for eliminating the rotational flow com-

ponent are presented in Fig. 1.6. The induced

axial pressure gradient was removed, and the

_amount of flow reversal along the inner wall was

decreased. As in the first series, the flow was

~essentially two-dimensional; that is, no appreci-
-able radial component existed.

Turbulators were added to the previous con-

i i‘igbratibh for another series of tests. The tur-
'_"_'__*‘V'bu]a'rors were expected to generate radial velocity
- 'ffcomponems ‘which would carry boundary-layer fluid
- fmto the mlds’rrecm and vice versa. It can be seen

 

ORNL-LR-DWG 9349

     
 

EQUATOR — STATION
5

INLET GUIDE
VANE DETAILS

NO DATA TAKEN
ABOVE STATION 7
OR BELOW STATION 4

 

o

q- ——
o
Q.
=
-

o

B
O
O
0
[}
>

o

Fig. 1.6. Axial Velocity Profiles from Test
Series 5. |

from Fig. 1.7 that the amount of flow reversal was
further reduced. '

In another series of tests (series 8) flow re-

versal was eliminated from station 6 downward
by use of a greater radial velocity component.
The configuration of series 8 was used for two
brief tests that simulated the one-fuel-pump-out

condition. Preliminary results’ md:cate that flow
conditions change very litile, ’rhe percentage of

flow separation remaining approxmately the same

as for the two-pump condition.

Data are taken from the metal core test rig by
means of wall static pressure taps and claw probe
traverses. The latter read total pressure and flow
direction but are limited to measuring flow which
is two-dimensional. As stronger turbulators are

designed and radial velocity components become
greater, accurate data will be mcreasmgly more

 

 

"

T T

,

 

 
3t PERIOD ENDING SEPTEMBER 10, 1955

 

i L kgt < electrolyte injected was a concentrated solution
. ORNL-LR-DWG 9320 . ..

of sodium chromate. Injections were made
manually by using a 30-ml glass syringe. Some
care was required to obtain adequate insulation
and to seal arcund the two wires of the conduc-
tivity probes. A satisfactory arrangement employs
Kovar tubing and wire with a glass insulator and
seal As now used, the probes consist of a
/16"'"‘ -dia Kovar tube with two 0.025-in.-dia Kovar
wires. The wires afe separated by approximately
}/8 in. and project about % in, from the seal. The
total length of the probe is 9 in. Some corrosion
of the probes occurs, but probe life is considered
to be satisfactory.

   

INLET GUIDE VANE AND
TURBULATCR DETAILS

   

 

The resistance bridge is a Wheatstone bridge
with fixed legs of 10,000 ohms each; the third leg
is adjustable to match the electrolyte resistance
sensed by the probe, which is the fourth leg. A
45-v battery supplies the bridge current. The

of the water, and therefore only a change in

NO DATA TAKEN .. . . .

ABOVE STATION 7 conductivity is passed as a signal to the ampli-
OR BELOW STATION 4 fiers.

Conductivity experiments have been made on
both the aluminum mode! and the transparent
plastic model. The water flow rate is approxi-
mately two-thirds the fuel flow rate for the
Reynolds number expected with the fuel flow.
Perhaps the most pertinent values derived from

‘ Lo i the tests are the estimates of transit times, For

Fig.. 1.7.  Axial Velocity Profiles from Test example, data were taken with the probes located

Series 6. _ , o very near the inlet and outlet of the aluminum

' R : P model core, and, if time is counted from the first

‘ difficult to obtam.' Two other hmltahons to the __w__appearance ‘of added salt at the inlet probes, the
metal core rig exist in that fhe probes “ {ime of appearance of the salt at the outlet probe

average values and do not respond to transients, " is ds shown in Fig. 1.8. The fastest transit time

. Qnd surveys‘ are I|m|fed>fo fwo po|nfs 9 deg Qparf ) ;_WQS O ?lhsec, whilafhe S‘IOWGSf medsured transit
' ; ' as ab9u1,2.8_ sec. There IS a ||mlt to the

L

 

 

VELOGITY (fps)

 
 

   
  
 
 
 
 

   

 

.7 doubtedly, some small arount of salt passes after
~ the time of no sngna[ on the Brush recorder. The
"":mean fransit time, compu’red as. the quohem‘ °f

 
 
 

 

 
 

plastic core

=

     

 

      

»fransn’"h_rmem acc rately. Howevér, fhe mmlmum

 

'duchwty o dran me

i ... The apparatus’ used includes “two conductlwtyd”-"Y"'frang;’r time was found to be 0.35 sec, and, by
L -~ probes, two resistance bridges, two Brush ampli- extrapolation, the maximum transit time was esti-
' fiers, and a two-channel Brush recorder. The mated to be at least 3.2 sec.

LT

27

 

 

 

bridge can be balanced for any initial conductivity

sensitivity of fhe measurlng apparatus “and, un-

systen{ is"s small the sa[’r was reCIrauiated rapldly, )

T

o

 
Saki e ks M i i

 

 

ANP PROJECT PROGRESS REPORT

 

- ORNL LR DWG 9321 7

 

 

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q oo
w
wy
oy
g
© ® RUN f _c*
z O RUN 2 o
> 80
<
I /
wi Q
e
wl
-J
&3 60
o
5 E /
O - |_DURATION OF .
53 SIGNAL ON /
23 40 |«NLET PROBE
a
a
= /
5 //
u 20
<I
=
= /
S
2 /
a 0 b |20 T ,
' 0 | 1.0 2.0 ' 30 40

TIME AFTER INITIAL SIGNAL AT INLET PROBE (sec)

Fig. 1.8. Transit Time for Flow Between Core Inlet and Outlet.

Studies were made of the flow of water through
the transparent plastic mode!l by injecting a dye
at various points, The header, at the time of
these tests, was designed to produce two,
meshing, helical vortexes through the flow
chamber, Several combinations of chemicals for
imparting color changes were considered. The
most satisfactory one, by far, was a combination
of starch, iodine, and sodium thiosulfate., For the
tests the model was filled with a starch solution,
which was transparent prior to the injection of
the dye, the dye being an iodine solution (a
safurated solution of iodine in methyl alcohol).
After the dye was injected, transparency could
again be obtained in the closed system by the
injection of a clearing agent. The clearing agent
was a saturated solution of sodium thiosulfate in
water, The study was made by alternately in-
jecting the dye and the clearing agent. The dark
blue cloud resulting from the dye injection was

' f_transported with the water. The general choracter-
‘istics of the flow could be observed, and, in
‘particular, the flow at various points could be
noted. In one way the action of the clearing agent
was ‘superior to that of the dye: those areas

where turbulence was least were the last ones

28

to be cleared of the dye. Photographs were taken
of the tests, but they have not yet been analyzed.
Visual observations indicated that the water
circulated about the vertical axis of the core, with
the greater portion of the mass mckmg from four
to seven revolutions in passage. There was
upward flow in the upper hemisphere near the
island, but very good mixing appeared to prevail
throughout most of the core. An extensive series
of pictures of flow in both cores is scheduled for
completion in September. |

Core Design Analyses

W. T. Furgerson
Aircraft Reactor Engineering Division
Interest in core designs with lower degrees of
divergence than that of the present 21-in. design

prompted a study of the effect of inlet annulus

dimensions on the adverse pressure gradients

encountered by the fuel in flowmg from the core
inlet to the equator. Any core desngn in which’

the area perpendicular to the flow mcreases along

the flow path will have adverse pressure gradlen’rs'

resulting from conversion of Velocdy head to
static head. If the flow entering the core has a
rotational component, an additional axial pressure

 

1‘)“:.: LT

-

W T

 

 

 

-y

 

.;‘.'-‘_ - i

 

 

-
 

 

 

 

 

 

 

' max:mum )

gradient will be induced through fluid friction
effects, |f the rotational component is centered
about the core axis, the induced gradient is
favorable along the outer wall and adverse along
the island, In each case, the gradient is algebrai-

cally additive to the gradients brought about by

area change. Since, all other things being equal,
flow separation is favored by higher relative
pressure gradients, the result is a tendency for
flow to separate from the island.

The effect of increasing the outside radius of
the present 21-in. core inlet on the two relative
pressure gradients along the island wall is illus-
trated in Fig. 1.9. The divergence gradient de-
creases because of the decreased divergence of
the core, but the induced gradient increases,
principally because of the increased ratio of inlet
radii. '

ORNL-LR-DWG 9322

 

o
N

 

 

INDUCED PRESSURE GRADENT

 

0.2 | ,
DIVERGENGE PRESSURE GRADIENT. /

-~

RELATIVE LOCAL PRESSURE GRADIENT

 

/ T——l
50 5.5 6.0 6.5 700 7.5
CORE INLET OUTSIDE RADIUS (in.)

o

 

 

 

 

 

 

(de/dx)/q,

 

S TR

- Fig. 1.9, Relatlve Dlvergence and Induced Pres- -
__sure Gradients for a 21- -in. Core wnh a 3405-|n.
Inside’ Radlus at the InIef. o o

A contour mop showmg the hnes of constant
: "_;'relohve pressure grodlent plotfed oga:nst both
;II‘ISIde _g d T

 

E chcmge ‘of inlei “inside and outside radii to 5.0

and 725 in., respechve!y,' could reduce the
‘/'"'}'_l_dhve gradlen’r from 0 19 to 0.15, but
an injudicious increase of inlet area could result
in higher relative gradients than those of the
present core design.

PERIOD ENDING SEPTEMBER 10, 1955

ORNL- LR-DWG 9323

 

8.0

7.5 » ////

 

 

O\
QNN O
S 83N,
2
A

(dp/dx)fg=0.4
\ £ LA £

/
7.0 Y 5,/ / ; e /’
L
/

6.0 / u

/ od

PRESENT INLET

|

30 3.5 4.0 4.5 5.0 5.5
CORE INLET INSIDE RADIUS (in.)

 

O
NN
]

 

\ ¢
N

 

 

CORE INLET QUTSIDE RADIUS (in.}

\\
s\\
o
3

\

 

 

 

 

 

 

 

 

5.0

Fig. 1.10. A Contour Map of Constant Relative
Pressure Gradient vs Core lnletInside and Outside
Radii.

Similar plots have been prepared for cores
having different equatorial dimensions, The value
of the minimum relative gradient decreases with
decreasing equatorial area, but the inlet radii af
which the minimum occurs remains as shown in
Fig. 1.10. It must be emphasized that the induced
gradient is present only where rotational flow

“components exist. Also, the equation by which

the induced gradient is calculated is highly

__,_theoretlcal ‘and has nof yet been quantitatively
| checked by experlmen’r

   
    
 

" Fuel Pump Performance

R. L. Brewster A. M. Smith
M E. Lackey G. D Whltmcm

Performcnce tests ‘of the fuel pump (model MF-2)
T with wqter as the c1rcu|cn‘ed flmd were concluded,

with a fofql of ]6 tests havmg ‘been conducted.
:‘These tests mvolved changes in lmpe“er desxgn,
in the axial and radial running clearances, in
volute design, and in suction geometry. The major.

changes in impeller design, suction geometry, and

29

 
i SR o e bk SRl

 

 

" ANP PROJECT PROGRESS REPORT

rcdml rtmnmg ‘clearance were discussed in the
prev:ous report, 1
Further study was made of the noise which was

~ present in the initial testing and which was
- suspected to be cavitation noise. It has been
- determined that the noise was not primarily due

fo4f|uid'cdv'iflfaf'i6n but, rather, to a local flow

'&condmon that ex:sted at the tongue of the pump
volute. This was established by the use of a

carbon microphone probe. A circuit was devised
‘50 fhat the microphone signal could be observed

~-ona cafhode-ray oscilloscope. With this arrange-
‘_ment the region of maximum noise was located
_near the volute tongue. The absence of cavitation

was further indicated by no observable decrease
in efficiency when the pump was operated at the
design point under conditions of varying pump

_suction pressure. The pump suction pressure was
'varled from approxlmafely 11 to O psig with a

a’rer temperature of approximately 70°F. The

-~ vapor pressure ‘of the water at 70°F corresponds

to that of fuel at a temperature of 1535°F, which
is well above the maximum fuel pump suction
temperature expected. '

Analyses of the data from the pump performance
tests indicated the need for additional flow area
in the pump volute, The particular evidence for
this was the imbalance of hydraulic forces around
the periphery of the impeller at the design point
and the occurrence of design-speed maximum ef-
ficiency at approximately 20% less than the design
flow.

A design change in the reactor north head for
structural reasons allowed the pump volute flow
area to be increased; however, this increase did
not compensate for the increase in the fuel flow

rate required for improved heat exchanger per-

formance. To obtain the additional required flow
area, the pump volute was redesigned without

- diffuser cones.

R_esulfs ‘of water performance tests with the
redesigned pump volute indicated excessive re-

- circulatory losses between the impeller discharge
" .and suction volumes. The impeller lower shroud

was therefore modified to allow the 12 radial

slingers to be lengthened to the full impeller
_diameter in order to increase their developed back
_ ’_:bi'ééjs,fire:"' The results of this modification are
fshown ln Flg. 'I 11. To further reduce the re-

 

 

]G D. Whitman, R. L. Brewster, and M. E. Lackey,

:’:'ANP Quar. Prog. Rep. June 10, 1955, ORNL- 1896, p 32.

_'_30_ .

 

circulatory loss, the impeller axial clearance was

reduced from 0.102 to 0.053 in. The results of
this modification are shown in Fig. 1.12, '

It may be seen from Fig. 1.12 that the desngn
point lies in the region of maximum efficiency.
The noise present with the earlier volute design
was eliminated, and the condition of hydraulic
force balance on the impeller occurs near design
speed and flow.

Controls and Instrumentation

J. M. Eastman

Bendix Products Division

E. R. Mann

Instrumentation and Controls Division

A flow sheet has been prepared that shows,
schematically, the locations of the instrumentation
and control components,  Control-panel layout
studies“are being made. Elementary-control elec-
trical wiring diagrams are being prepared to
provide control actions in accordance with infor-
mation block diagrams, which show control func-
tions independent of control techniques or hardware
types. The selection of control technigues and
the specifications for component hardware are
being worked out. Apparently no fundamentally
new development work, except for special sensors,
will be necessary to meet control requirements,
although heat load responses may require the use
of some controls not previously required for
reactors,

The linear equipment for an ORNL reactor
simulator has been assembled and tested, Com-
pletion of the nonlinear simulator equipment is
scheduled for September 1, 1955, with the delivery
of special components from Reeve Electronics,
Inc. One of the major objectives of the simulation
is the determination of the control system re-
sponse requirements under emergency conditions.

Two instrument test loops are being prepared
for checking special sensors, including thermo-
couples, flowmeters, pressure indicators, and the
tachometers. Arrangements are also being made
for testing the fuel-level indicator.

Conceptual designs have been worked out for
the rod-drive mechanism, the enricher actuating
and control mechanism, the moderator coolant
temperature control system, the flux-sensing
chambers, the radiator main louver actuating and
control mechanism, the fuel and sodium pump

-ty ‘-

 

-

 

 

 

 

 
 

 

 

 
   
   
  
   

 

- ispeed com‘rol mechamsm ‘and the fuel-level indi

PERIOD ENDING SEPTEMBER 10, 1955

 

ORNL~LR ~-DWG 9324

 

80

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

70 y
~I
AN — ———PUMP EFFICIENCY
® 2700 rpm
60
50 .
o 40 '
2 N
1
@ \\
: /
0\<ZOOO rpm ,/ .
/s
‘ \ o l.
30 ~ey 7 7 v\ .
L Tewk //’ s R
\. /'/ // /
7 \\ // .
. ® / -~
20 P N
1300 rpm / e - 68%
N l\)/ -’x
e °
\
\\. ‘/
10 -7‘:’\
0\4‘000 rpm /{ *
o‘\. — VOLUTE PRESSURE BALANCE LINE
600 r T—e<l
pm L g
0—L. .___./
0 i

0 © 100 200 300 400 500 800 700 800 900
CoLL T T ~ ' " CAPACITY (gpm)

  
   

 

Flg. 1 11." Fuel Pump Performance Characteristics with Modlfl.red lmpeller. Lower shroud of lmpeiler

i ;":modn‘ied to ollow ’rhe 12 rqdml shngers ’ro be Iengfhened to ful_l impeller diameter in order to increase
j"fdeveloped bock pressure Axml clecarcmce on shroud 0.102i |n

 

 

   
  

 

thermocouplés have been mc]uded ln ’rhe reactor

esign of 'rhe molr} Iouver and de5|gno -

 

Engmeermg Tesf Umtr
M. Bender

: Alrcraft Reactor Engmeermg Division

   

n fherll' actuating “A construction program for the Engmeermg Test
and coolmg eqb:bment “and the tod ‘actuating” Unit (ETU) has been established, which, based
mechanism will begin after pertinent reactor design on current equipment delivery information, makes
details are frozen, Fuel-temperature-sensing it feasible to start assembly operations on June 1,

31

 
S i ¢ o e M G, 5. B SR, i

 

 

 

ANP PROJECT PROGRESS REPORT

A E I

ORNL—LR—OWG 9325

 

S0

 

b
3000 rpm
80 \\(

\.

 

 

 

 

HEAD (ft)

 

 

 

 

 

 

N
~
~
\\\./\Q\VOLUTE PRESSURE BALANCE LINE

 

 

 

 

 

 

 

 

 

 

 

 

0 100 200 300 400

500 600 700 800 S00

CAPACITY {gpm)

Fig. 1.12. Fuel Pump Performance Characteristics with Modified Impeller and Axial Clearance on

Shroud Reduced from 0.102 to 0.053 in.

1956. On this basis the target date for completion
of assembly and initiation of operation of the ETU
has been set at September 1, 1956.

. The basic flow and heat transfer circuits have
. been established so that two of the main NaK
circuits will heat the fuel and the other two will
" cool the fuel. In each of the cooling circuits there

will be a one-fourth section of an ART radiator

arrangement. In each of the two heating circuits
there will be a 1-Mw gas furnace. The sodium
moderator circuit will be cooled by a 0.5-Mw
radiator. This arrangement will apply sufficient
heat to permit measurement of the response of the
system to power changes at low power levels.

 

..)'

-

s F

 

 

 

 
 

R

L |

 

"

 

 

The most important objectives of this experiment
are to develop assembly techniques for use on
the ART, to furnish some information on thermal
stresses in the reactor assembly, to obtain some
heat transfer information on the radiators and the
NaK-to-fuel and Na-to-NaK circuits, and to test
some of the instrumentation to be used on the

ART.

REACTOR PHYSICS
Control Rod Heating and Burnup

W. J. Fader
Pratt & Whitney Aircraft

A. M. Per‘ry

Electronuclear Research Division

Calculations of control rod heating and burnup
to be expected in the ART were completed. The
rod designs considered included cylindrical annuli
of various rare-earth oxides, ]3/8--in. OD, either
7/8- or ]]/B-in. ID, and ](4 or 1/8 in. in annular
thickness. The maximum density of the heat
generated by gamma-ray absorption in the 141—in.
annulus was found to be about 140 w/cm?, of
which 15 w/em® was due to gamma rays produced
in the core; the remainder was due to neutron
capture gamma rays produced in the rod. The
heat generated in a %-in. annulus was found to
be about two-thirds of that generated in the Z-in.
annulus. Temperatures in the rod material depend
on whether the rod is fabricated of fused oxide
or of a cermet. For a Z-in.-’rhick fused oxide rod,
the temperature rise is expected to be 620°F if
the rod is cooled at both the inner and the outer
surface and to be 2300°F if it is cooled only at
the outer surface. Correspondmg temperature
rises for the cermet are very much smaller: 20

. and 54°F, respectively, if the conductivity of the
cermet is assumed to be the same as that of iron.

The heat fluxes are large: 140,000 Btu/ft2.hr if
the rod is cooled at both surfaces, 230,000

Btu/ft2.hr if it is cooled at fhe= outer surfoce

only.
The thickness of the poison material required

to allow for burnup by neutron absorption depends |

on the'cbundcnce of high cross-section isotopes
in the rare earth and on the density of the rare-
ear’rh oxide, which would be greater for the fused
ceramic rod than for the cermet. Calculated pene-

 

2A. M. Perry; The Boron Layer of the ART, ORNL
CF-55-8-38 (Aug. 4, 1955).

PERIOD ENDING SEPTEMBER 10, 1955

trations for 500 hr at 60 Mw are tabulated below.
The cermet is assumed to be 33% oxide by volume.

Burnup Penetration (in.)

 

Ceramic Cermet
Sm203 0.051 0.150
Gd203 0.028 0,082
Eu203 0.011 0.032

For a ]/4-in. annular control rod of BAOC in
copper (30 vol % B,C), the burnup penetration
would be one-sixth as great as that in a europium
oxide cermet, but the heat produced would be
one-third greater. All these considerations are
based on a neutron current of 2.5 x 1014
neutrons/cm?.sec into the rod.

ART Boron Layer
A. M. Perry

Flectronuclear Research Division

The boron layer (assumed to be made of B;OC
tiles) between the reflector and the fuel-to-NaK
heat exchanger was examined with respect to heat
generation, flux attentuation, helium gas evo-
lution, and burnup.? Heating will be largely due
to the alpha particles from the B19(n,a)Li? re-
action and will amount to about 13 w/cm? of tile
(40,000 Btu/ft2.hr). The heat flux per unit area

will be one-half this amount, since both surfaces

of the tile will be adjacent to cooled surfaces;
thus the temperature rise for radiative heat -

transfer will be about 500°F. The absorption
mean free path of thermal neutrons (1200°F)

B;OC with a density of 2.5 g/em® is 0.0017 in.,
while that of 3-kev neutrons (near the sodium
resonance) is 0.145 in. If the BlOC layer were
made 1 in., thick to absorb sodium resonance
neutrons, gamma-ray heating of the layer would
increase the total heat produced by only 30%.
The flux of about 4 x 103 neu'rrons/cm2 ec

" absorbed over an area of 4 x 10¢ em? yields a

theoretical helium gas evolution rate of 5

liters/day (STP) Burnup of the boron will be

only 1.2 mg/cm of B0 in 500 hr at 60 Mw.,
ngh-Temperature Critical Experlmenf vs ART

A. M. Perry
Electronuclear Research Division

It is recognized that certain rather minor differ-
ences exist between the high-temperature critical

33

T T

 
 

 

 

 

 

ANP PROJECT PROGRESS REPORT

experiment and the ART with respect to the
factors that determine the critical concentration
of UF'4 in the fuel mixture. The differences that
may have a significant effect on the critical
concentration are: (1) control rod position: the
ART will be made critical with the control rod

inserted fcrther than it was in the high- -temperature

_ -'crlhcal experiment; (2) beryllium density: cooling
. holes in the ART reflector will reduce the moder-
--;'cn‘or densny, (3) reflector poisoning: sodium
 coolant in the reflector will increase neutron
dbs'orpfi'oris' in the reflector; (4) reflector size:

in order to minimize machining of beryllium for

" the experiment, the reflector was made in an

irregular rectangular shape circumscribing the
design reflector and had a somewhat larger volume
than that of the ART reflector. The effects of
these differences are estimated in the following
paragraphs. |

The ART will be made critical ot 1200°F with
the control rod withdrawn about 1.5% in A% from
its fully inserted position, The control rod cali-
bration of the high-temperature critical experiment,
carried to a point 4 in. above the mid-plane of
the reactor, appears to be consistent with an
over-all rod value of about 5% in Ak. Thus, the
Ak inserted at critical will be about 3.5%. The
increment in concenfration necessary to com-
pensate for the initial rod insertion is thus about
7 x 3.5, which is ~24%. The initial rod position,
of course, is, to some extent, at the discretion
of the operators, and the increment in critical
concentration could, if desired, be chosen to be
less than 24% — say 18 or 20% — to correspond
to an initial rod insertion of 2.5 to 3%.

Since the change in beryllium density is not
uniform, the cooling holes being more closely
spaced near the core, it is necessary to calculate
a weighted average density change. This was
done in two ways., From two-group perturbation
theory, the importance of a change in the fast-
group diffusion coefficient or in the group-transfer
cross section is approximately proportional to the
square of the fast flux. The quantity U,(r) was
obtained by adding multigroup fluxes, and 1,[!2 was
used as a weighting . function.  An c!ternate

‘weighting function was obtained by computing the
- slowing- down density by means of age theory.
““In this case the weighting function at radius r
Colis opphed to the average density change in the
o spherlcal she” befween the core and the spherical

surface of radius r, rather than to the local density
change at r. Average density changes computed
in these two ways are 4.2 and 3.5%, respectively.
The reactivity effects of a uniform density change
in beryllium have been computed by Curtiss-Wright
Corp. to be 0.42% in reactivity for 1% in density,>
and by Pratt & Whitney Aircraft to be 0.35% in
reactivity for 1% in density.? Based on these
figures, an average reactivity change of —-1.5%
is predicted. Data from the high- fempérafure
critical experiment indicate (AM/M)/(Ak/k) to be
about 7. Thus a change in critical concentration
of 10.5% is anticipated as a result of the reduction
of the beryllium density. |

The effect of thermal-neutron absorptions in the
sodium coolant was estimated by comparing the
calculations of the Curtiss-Wright Corp.> for
Inconel-lined and unlined cooling holes.  The
penalty in uranium concentration due to the
Inconel was reduced in the ratio of the macro-
scopic thermal neutron gbsorption cross sections
of the sodium and the Inconel. When allowance
was made for the greater number of cooling holes
presently planned for the ART, the increment of
fuel concentration required to compensate for the
sodium in the reflector was 5.6%.

The effect of the additional beryllium in the
reflector of the high-temperature critical experi-
ment has not yet been satisfactorily computed.
This is due, in part, to the irregular distribution
of the added beryllium and, in part, to some
uncertainty regarding the dependence of reactivity
on reflector thickness. Whether the entire re-
flector volume is considered or only the portion
between two planes 1 ft above and below the
equatorial plane of the reactor, the beryllium in
the experiment was about 3 in, thicker, on the
average, than the beryllium in the design reflector
of the ART. According to the parametric studies
of Curtiss-Wright Corp.,> such a decrease in
reflector thickness over the whole surface of a
spherical reactor would increase the critical con-
centration by a factor of 1.26. Results of the
cold critical experiments, however, indicate that
removing 3 in. of beryllium over the region ex-
tending 1 ft above and below the reactor mid-plane

 

3c. B. Mills and H. Reese, Jr., Design Study of an ANP
Circulating Fuel Reactor, WAD-1930 (Nov. 30, 1954),

D, G, Ott and A. Berman, private communication,
SH. Reese, Jr., S. Strauch, and J. T. Mihalczo,

Geometry Study for an ANP Circulating Fuel Reactor,
WAD-1901 (Sept. 1, 1954).

€

¥

%

i ol ki s

——

 

g

-

 

 

 

 
w

oy

 

 

should increase the conceniration by about a
factor of 1.12. The true effect is believed to lie
between these extremes.

~ The over-all factor by which the critical con-
centration of the ART is likely to exceed that
of the high-temperature critical experiment is

_dbtained by multiplying the four factors together.

The result is

F = (1,2)(1.])(],06)(1912) = 1.56 ,
if the low estimates for control rpd allowance and
reflector size effect 'qr'é; ‘empléyed, or

F o= (1.25)(L1)(1.06)(1.26) = 1.84 ,
if the high estimates are used. Since the clean
critical concentration of the experiment was 2.9
mole % UF , the critical conceniration of the ART
is expected to fall between the limits 4.6 and
5.4 mole %. Some multigroup calculations are
to be undertaken, in the near future, which should
help to establish the critical concentration of the
ART somewhat more reliably.

 

R. R. Bate, L. T.’Erinstei}nr, and W, E, Kifiney, The
Three-Group, Three-Region Reactor Code for Oracle,
ORNL CF-55-1-76 (Jan. 13, 1955).

 

PERIOD ENDING SEPTEMBER 10, 1955

Multigroup, Multiregion Reactor Calculation

W. E. Kinney

Aircraft Reactor Engineering Division

A new multigroup, multiregion reactor calcu-
lation is being programed for the Oracle, By
taking the consistent P, approximation to the
Boltzmann equation, group equations have been
developed and lethargy dependent coefficients
have been put into a form suitable for coding. In
the treatment of thermal neutrons, both neutron
and moderator temperatures will be considered.
The theory for the inclusion of shells and for the
spatial integration of the group equations is the
same as that described previously.b

The code will compute fluxes as a function of
one space variable in slab, cylindrical, or
spherical geometry. As presently planned, it will
allow up to 125 groups and 125 regions. The
calculation may be iterated on the concentration
of any specified element to obtain a multiplication

‘constant of unity, and self-shielding factors will

be available. Adjoint fluxes may be computed,

if desired.

35

 
 

 

 

ANP PROJECT PROGRESS REPORT

2. EXPERIMENTAL REACTOR ENGINEERING

H. W. Savage

E. S. Bettis

Aircraft Reactor Engineering Division

The difficulties encountered during the installa-
tion of 1he MTR in-pile loop and in startup of the
system are presented, and the modifications being
made in the loops now being fabricated are de-
scribed. The operating conditions for the 22 fused-
salt—Inconel forced-circulation loops operated
durmg the quarter are presented, as well as the
operating conditions for 9 forced-circulation loops

“operated with sodium in Inconel or in stainless
“steel tubing. A new test loop is described with
which more c:cciirm‘e information on the oxygen
content of fhe sodlum can be obtained during

R operohon.

Two stands for teshng ART pumps were de-
veloped ‘and tests of pump seals were made. The
tests performed with intermediate heat exchanger
test stand A are described, as well as the smali
heat exchanger tests,

The apparatus being built for tests of the thermal
stability of the ART outer core shell is described.
Also, a design for Inconel strain-cycling tests is
presented,

Several designs for cold traps for removing oxides
from sodium and NaK are presented, along with
designs of plug indicators for detecting and meas-
uring the oxide content,

IN-PILE LOOP DEVELOPMENT AND TESTS
D. B, Trauger

Aircraft Reactor Engineering Division

Loop Installation

C. W. Cunningham
Aircraft Reactor Engineering Division

The instrument panel for the MTR in-pile loop
was installed on the first level balcony on the
reactor north face. This location is advantageous
in that it is above the activity on the main floor
and relatively close to the HB-3 beam hole. Pro-
vision of a central balcony extension for the oper-
ator enabled him to easily observe the instrument
and control panel in the somewhat restricted space,
The additional instrument panels and cabinets
were Iocated on balcony extensions at each end of
the instrument and control panel.

" The auxiliary equipment and the loop, on its
cradle, were placed on the main floor, largely

36

under and adjacent to the east face stairway.

Interconnecting tubes and wires were protected
by running them in troughs hung beneath the bal-
cony and the stairway above head level. The
tubes and wires to the loop were lifted into these
troughs after the loop was inserted into the beam
hole.

A means for flux compensation had to be pro-
vided, since the flux profile at the HB-3 beam hole
was not known with certainty, A loop retractor
mechanism was built and was mounted in the HB-3
cubicle to permit withdrawal of the loop plug 6 in,
from the fully inserted position. This device
permits adjustment of the loop nose position in the
tapering flux region and, thus, variation of the
loop power level independently of reactor opera-
tion. The retractor also serves as a safety device.
By withdrawing the plug 6 in. and reducing the
process air flow, reactor operation could continue
in the event the in-pile loop experiment had to be
operated on the limited, emergency, cooling-air
supply. The ability to preset the position of the
plug greatly reduces the risk of freezing or over-
heating the loop during the reactor startup and
shutdown,

Orderly arrangement of lines and equipment in
the cubicle presented a difficult problem. The
lead shields, retractor mechanism, tubing cutoff
block, tubing and electrical wires, and hoses
almost completely filled the available cubicle
space. The problem was further complicated by
the movement required for retraction. A workable
arrangement was found by trial and error movement
of the lines and tubes into various coiled configu-
rations. The water hose had fo be carefully routed
to eliminate excessive strain on the quick-discon-
nect fittings provided to eliminate the leakage of
radioactive process water,

Loop No. 1 Cperation

L. P. Carpenter D. W. Magnuson
P. A. Gnadt

Aircraft Reactor Engineering Division
D. M. Haines
Pratt & Whitney Aircraft

The first MTR in-pile loop was completed on
June 20, and it was shipped, by air, to the MTR

 

 

il o

 

 

 
 

 

 

 

“This r revision wds

facility at NRTS.

It successfully passed the pre-

operational checks required by the MTR Reactor

Safeguards Committee, and it was then inserted
into the HB-3 beam hole from the special loading
cradle supplied as part of the experiment, Fuel
e[ements adjacent to the beam hole were removed
for the insertion, since no shielding was provided.

Although a rather high radiation beam was meas-

ured adjacent to the in-pile loop plug shield, it
was well collimated, and the insertion proceeded
rapidly and smoothly without the coffin,

. The startup of the loop proceeded systematically
through pump operation and preheating of the fuel
and the piping. At the final step, melting of the
freeze line and charging of the loop, the heater for
the freeze line was found to be inoperable. An
internal short, to ground, had developed after the
preoperational checkout. Efforts to clear the short
were unsuccessful, and it was impossible to melt
the line by increasing the power to adjacent
heaters, This terminated the loop operation, and
the loop was removed from the beam hole for return
to ORNL. Since this loop had been reworked re-
peatedly during the initial assembly, it was deemed
to be irrepairable and was therefore cut up for
evaluation,

The short was fo}uhd in a nipple, or extension,
from the pump bulkhead to the glass seal for the
power leads. Braided glass insulation had become
frayed, apparently during assembly, and the copper
lead wire was eprsed ‘Mechanical separation

befween the copper wire and the nipple probably
existed durmg the checkout but rnovement due fo _

 
 
  
  
 
  
 
 

of the thermocoup

tion, Experience with the flrsf ‘two Ioops shoulcl
provide adequate information for en‘abhshmg heat-

ing procedures,

;lw'ro tha'r of 'rhe in- -pile Ioop.
_fefm'"qfed bY a sudden pump stoppage caused by
:..;.,"h‘e c‘CZCUN’!U'fl'fIQr\ of fuel along the shaft in the

me'ral _surface,
-':by a more ilgh’rly etc ed surfuce, was SUff'C'e"”y
hlgh equxvclent to an increase of about 13 cm3

- ;"‘of flu
wfllnger

 
  

 

PERIOD ENDING SEPTEMBER 10, 1955

Horizontal-Shaft Sump Pump
J. A, Conlin

Aircraft Reactor Engineering Division

The difficulties previously encountered with the
shaft seals of the horizontal-shaft sump pump have
been corrected. The original seal used a thin,
brass bellows, which failed mechanically. It was
replaced with a stainless steel bellows, which
serves both as the flexible member of the seal
and the loading spring, It was necessary that the
beliows serve as the loading spring because the
combined spring loading of the steel bellows and
a spring would have been excessive. However,
the spring-loaded bellows caused chatter between
the graphite nose and the mating seal ring, and
the graphite crumbled. The installation of friction
dampers, in the form of spring-loaded steel plugs
that pressed radiqlly against the outside of the
seal nose, eliminated this difficulty.

The maximum gas diffusion leakage rate for this
seal has been found to be 100 ppm of argon dif-
fusing from the pump sump into the helium in the
bearing housing. The flow rate of both the helium
and the argon is 700 cfh, Satisfactory seal leakage
rates could not be obtained consistently with the
initial drip lubrication system; so the oil level in
the bearing housing was raised until the lower
edges of the running faces of the seal were im-
mersed, This provided a better oil film on the
seal mating faces, and satisfactory sealing condi-
tions were obtained.

A prototype model of the in-pile pump operated
satisfactorily, with the fuel mixture NaF- -ZrF ,-UF

(53.5-40.6.5 mole %) as the pumped fluid, for

173 hr at 1400°F in an isothermal loop identical

Operahon was then

Id region of the pump. A postrun ‘examination

 

 

'md»lrcated that there had been two levels of fuel

‘fl"u-e sump.‘ The first level was ‘at 'rhe normal
p opercn‘mg po:nt ds indicated by an etched
The se_cond level, as evidenced

 

T 'd, fo reach ’ro above fhe bo'r'rom of fhe shof’r: '

 

to collect ulong “the shaft cnd ultamately resulted
“in pump stoppage. “The change in sump level
probably occurred the day before the pump stopped.

-

37

 

 
 

 

 

ANP PROJECT PROGRESS REPORT

An m’rerruphon in the plant air supply thcn‘ was
bemg used to cool the fill-tank freeze line caused

“the fill-line temperature to rise to above the
melhng point of the fuel and the indicated fill-
__tcmk temperoture to increase about 25°F due to

 the absence of air movemenf in its v:cmlfy. The

| fill-tcnk tempercture rlse, plus any gas evolution

,frorri V'rhe fuel, could have increased the fill-tank

 

gas’ volume sufficnenfly to force fuel into the

~'pump”and flood the sump. A fill-tank temperature

“tise of 35°F would hove, alone, caused sufficnen’r
w’_:expcnsmn of the fuel to account for the 13-cm®

‘displacement. This type of failure could not occur
“with the MTR in-pile loop assembly, because the
_ freeze hne wnll be cooled by its proximity to the
- '4_‘water |ocket of the Ioop.

    
  

= Oll |rradmt|on .7

D. M Hames
Prat’r & Whltney Alrcrcft

Samples of Gulf Hcrmony ““A" oil were irradiated
at the gamma facility of the MTR to determine the
suitability of this oil for use in the lubrication and
hydraulic power systems of the in-pile loop. Cal-
culated doses of 107 r were obtained. The specific
gravity of the oil samples increased about 1%, and
the viscosity increased about 60 to 90%. In a few
of the samples a small amount of suspended par-
ticulate material was observed but not identified.
The radiation damage observed in these tests is
not considered to be seriously defrimentql

 
 

DEVELOPMENT AND OPERATION OF
FORCED- CIRCULATION CORROSION
AND MASS TRANSFER TESTS

Ww. B. McDonaId

Alrcrcft Reactor Engineering Division

Operahon of Fused-Sult—lnconel Loops

C. P Coughlen P. G. Smn‘h'
A:rcraff Reac_tqr Engineering Division
| R. A. Dreisbach
: ~ Pratt & Whltney Aerl’Gf‘l‘

Twenty two fused- sait—lnconel loops were oper-
ated durmg this quarter. A summary of the con-
dn‘nons of operation of these loops is given in
- _,Toble 2.1. The results of metallurgical examina-
'_'hons of these Ioops are presented in Sec. 5,

o “Corrosmn Research

S The mu|or cause of failure of these fused-salt

 

 

.Ioops has been the freezeup of fhe sclf in fhei

coolmg conl when flow was mferrupted for any

reason. " Most flow m’rerrupnons resuH from loss

of power, either at the test rig or in the building
as a whole. The freezing of the cooling coils has
occurred in times as short as 30 sec. Preliminary

tests have shown that a gas flame automatically

1gnn‘ed at the time of an interruption of the coolmg
air will prevent freezmg for perlods of 5 min or

more. Several methods of automatic |gmhon of

such a flome were tried, but the only one which
appears to give the surety of lgmhon requared is a
high-resistance heater coil in the gas stream; the
coil is energized from a 12-v, wet-cell circuit.
-Liquid Metals in Multimetal Loops
C. P. Coughlen
A:rcrcff Reocfor Engmeermg Dwns;on

R. A. Dreisbach
Pratt & Whitney Aircraft

Nme forced-circulation loops were operated with
sodium in Inconel and in stainless steel tubing.
A summary of the conditions of operation of these
loops is given in Table 2.2, The loops were
heated with electric heaters, and each loop had
an economizer section. The results of metal-
lurgical examination of these loops are presented
in Sec. 5, ‘‘Corrosion Research.”

A new test loop has been designed with which
it will be possible to obtain accurate information
on the oxygen content of sodium during operation.
The test system, as indicated in Fig. 2.1, consists
of a main loop, which includes the sodium-sampling
device and a bypass cold trap, and a plugging-
indicator loop, with its separate pump and flow-
meter, (The plug indicator is described in a sub-

sequent portion of this section,) The cooling rate
of the entire loop can be controlled to a fraction of

a degree per minute. The oxygen content of the
sodium circulating in the main loop is controlied

by use of the bypass cold trap. The plugging-

indicator loop and the sampling device are both

7 used to :ndependenfly defermme the oxygen content
" and thus to determine the effic:ency of ’rhe cold' B

trap.

In order to use the pluggmg-lndlcator loop, them

valve between that loop and the main loop is
opened, and sodium is allowed to flow from the
main loop until the plugging- mdlcofor loop is
filled to a predetermined level in the surge fank.

 

 

4 P

S -l

oW

 

 

 

 
 

 

TABLE 2.1. SUMMARY OF OPERATING CONDITIONS FOR 22 FUSED-SALT—INCONEL FORCED-CIRCULATION CORROSION AND MASS TRANSFER TESTS

 

 

 

   

 

 

Maximum Maximum .
Loop Method of Type of Reynclds Te?mperah..lre Recorded Fyel Recorded Tube Fused Salt Ope-ratmg o
No. Heating Heufed Number  Differential Temperature  Wall Tempera- Circulated Time Reason for Termination
Section {°F) (°F) ture (°F) (hr)
4950-2 Direct resistance  Straight 5,000 200 1500 1565 NaF-Z rF4-UF4“ 1000 Scheduled
4950-3 . Direct resistance Straight 10,000 200 1500 1690 NoF»Zer-UFAb 1000 Scheduled
4950-4  Direct resistance Straight 10,000 100 1500 1600 NaF-ZrF ,-UF ;% 1000 Scheduled
4950-5° Direct resistance Straight 10,000 200 1500 1575 NaF-ZrF ,-UF 1000 Scheduled
. 4950-6. - Direct Resistance  Straight  ~8,000 300 1500 1620 NaF-ZrF -UF,# 1000 Scheduled
7425-1 - Direct resistance ' Straight ~10,000 200 1500 1600 NaF-ZrF ,-UF ¢ 22 Lecked at 22 hr; motor
. failure
7425.1A - Direct resistance Straight  ~ 10,000 200 1500 1600 NaF-ZrF ,-UF ,“ 1000 Scheduled
74252 . Direct resistance Straight 10,000 200 1500 1575 NaF-KF-LiF® 550 Fuel evalyation test; pump
‘ shaft seized at 550 hr
74253 Direct resistance ; Straight ~23,500 125 1525 1600 NaF-ZrF ,-UF @ 73 Motor overloaded; loop
o g failed at 73 hr
7425-3A - Direct resistance ‘?,ti‘?VS?raight ~ 16,000 125 1525 1600 NaF-ZrF -UF @ 1000 Scheduled
742544 . Direct resistanceff“ Straight 10,000 200 1500 1575 NaF-KF.-LiF¢ 1000 Scheduled; fuel evaluation
4695-4D ;. Direct resistance : i Straight ¢ 10,000 200 1500 1595 NaF-ZrF -UF ,“ 20 Failed during power
o ’ ' B . failure at 20 hr
0 4_695-4D-2i-‘,'Direcf.;resis'rance :;’:‘:‘.“Struig‘ht 10,000 200 1500 1635 NaF-ZrF ,-UF 4 500 Scheduled
-‘4695-5(:-2};‘ Direct resistance;_fii Straight ~ 10,000 200 1500 1635 NaF-ZrF ,-UF @ 1000 Scheduled
" . 742541 [ Direct resistance;:’ Straight . 2,750 200 1650 1700 NaF-ZeF -UF ¢ 1000  Scheduled
49353 . Gas i Coiled = 1,000 100 1500 1540 NaF-ZrF -UF,* 1000  Scheduled
D 49354 . Gas Coiled 10,000 100 1500 1690 NaF-ZrF -UF % 486 Clutch failure caused
: T freezeup; failed on “
. re start m
4935.5 - Gas " Coiled 10,000 206 1500 1550 NaF-ZrF ,-UF ,* 682 Clutch failure caused E
.. .\. freezeup; failed on 8
. restart m
" 49357 ¢ - Gas . Coiled 6,000 200 1500 1700 NaF-ZrF -UF % 330 Power failure caused z
' freezeup; failed on E
o . restart =z
o " 4935-7B . Gas "Coiled 6,000 200 1500 1700 NaF-ZrF ,-UF % InTest Conditions noted were o
' ‘ ' new conditiens im- m
pressed at 600 hr; o
s scheduled for 1000 hr "“l
49358 Gas - Coiled 6,000 200 1500 1700 NaF-ZeF UF,« 435  Terminated at 435 hr by >
: thermocouple burnout o
4935.9 " Gas “ Coiled 4,000 200 1500 1800 NaF-ZrF ,-UF ,* 1000 Scheduled ;‘
4 Composition: 50-46+4 male %. o
bComposifion: 50-46-4 mote % with 2 wt % of the total uranium converted to U . —
w0 €Composition: 11.7-5¢.1-29.2 mole %. o
O th
| ;ANP PR’OJE‘CT P‘R'oc'REss R’E'P'O"R‘T'z

TABLE 2, 2 SUMMARY OF OPERATING CONDITIONS FOR LOOPS THAT CIRCULATED SODIUM

 

TemPemel’e T Maximem T Operatmg-:

' Material of ' Cold Réynolds Differential  Recorded Fluid Condition of Time

 

| LoopNo
T Cornstrdction Trap  Number CF) Temperature (°F) Sodium (hr)

 

   

49514 dnconel  No | 55000 30 1300 Commercialgrade 1000
49515 ncomel  Ne 59,000 300 1500 0.15% ox:de added”’“:" |
49516 dnconel  Yes 59,000 300 1500 Highpurty 500
'49‘51,"-7_’:"""Typ3316 © Ne 59,000 300 1500  Commercial grade ~ 475%

stainless steel 7 _ ’ 7 e

  

 

49518 lnconel  Yes 59,000 30 1s00 Commercial grade T 1000
4951.9 _‘lnconelf ~ No  ~59,000 300 1500 “1% bquum added S s00
- 7426-] ‘:ihtlncone[ A Yes "\’59,000 o 30'6 S ‘]500; g banum added o ]000 c

o B 7426;2“ ‘__;Z_Incone[ -_ N - No ~59,000 300 1500 0.05% oxide udded 1000

- -

     

 

 

 

*Operahon termmafed by a power faliure. ' e o S E

7 UNGLASSIFIED o R
ORNL-LR-DWG 2046

. a SURGE TANK
SURGE TANK

PLUGGING INDICATOR LOOP ,
vk
s FLOWMETER LOCATION b

 

E.M. PUMP

(ot
€

W\
Y

     
  

PLUGGING DISK

-y it

({
o

MAIN LOOP SODIUM-ANALYZING

AND -SAMPLING DEVICE

 

SUMP TANK

) COLD TRAP E

)

F|g.2'l Loop for C:rculatlng Sodium in Inconel or in Stcunless Steel Tubing to Study thethfect of
;:f?r fhe Oxule Confenf of fhe Sodium. )

£. M. PUMP

 

  

4o -

    
 

 

sy

 

‘The valve is then closed, and the sodium is cir-

culated by the electromagnetic pump in the un-
heated plugging-indicator loop. The plugging disk
collects the oxides precipitated as the tempera-
ture of the sodium decreases, and it eventually
plugs. The temperature at which pluggingoccurs

is compared with a calibration chart to determine

the oxide content of the sodium. The plugging-
indicator loop is then heated until flow is re-

" established. The valve is then reopened and the

sodium is forced, by pressure, back into the main
loop.

 

PERIOD ENDING SEPTEMBER 10, 1955

‘The data on oxygen content obtained with the
plugging-indicator loop are compared with those
obtained with the analyzing device attached to the
main loop and with results of chemical analyses
of samples removed from the main loop. The
analyzing device attached to the main loop is
described in Sec. 9, ‘“‘Analytical Chemistry of
Reactor Materials.”’

A loop in which NaK is to be circulated has
been designed with the same surface-to-volume
ratio as that of the primary NaK circuits in the

ART. A sketch of this loop is shown in Fig. 2.2.

UNCL ASSIFIED
ORNL-~LR—-DWG 9047

Fig. 2.2. Loop for Studying Mass Trunsfer in an lhconél’l.dbp Circulating NaK.

41

 

R T T T T T, T T T e R e e A TR R

TR R R R TR T

- T T

TR a

 

 

 

 
e L S el

 

 

ANP PROJECT PROGRESS REPORT

It is desigried to operate at a maximum NaK tem-

peraiure of 1600°F and a minimum NaK temperature

   
 
 
 
 

. ‘"':PUMP DEVELOPMENT
" E.R. Dytko .
Pratf & Whn‘ney Aircraft

 

Mechamcal Shakedown and Beurmg-and-Seal Tests .'

"A ‘G. Grindell

_ _*::Asrcraff Reuctor Engmeermg D|V|5|on

_ The ART-'rype MF-2 pump lncorporates two fuce-A
: '*type mechamca! seals, The seal specifications
. requu'e the upper seal to have a leakage rate of

- oil to the ctmosphere of not more than 20 cm3 per

24 hr at 70 psi. The lower unit is to have a leakage

o of onl ln'ro helium across a pressure differential of
-0 to 5 psi of not more than 2 cm3 per 24 hr. The
""i'-‘,fi'_f‘i""‘seoled fluid is a light, spindle oil having a vis-

cosity of 6'0'“SSUA (Saybolt seconds universal) at

]00°F Most of the manufacturers com‘acfed would':__
not bnd on the seals, and the three or four com-'
panies who did make bids would not gucram‘ee -
their seals to meet the s'rrmgen'r reqmrements of
‘Therefore a ‘seal evaluation |
) program! was initiated in which a small number of
seals from the Fulton Sylphon Dlv;slon, the Dura-

_ '_'f'of ]QOOOF The Reynolds number is to be com-
IR poroble to thct in prlmqry NcK circuits of the ART
Sl L " the specifications,

" Ine., will be tested.

metallic Corporaflon, and the Koppers Company,

that low leakage rates could probably not be
achieved with metal-to-metal seals, and, conse-
quently, the seals being tested are made of carbon
products and ceramics,

The evaluation of the Fulton Sylphon seals is
nearly complete., Nine of the 11 tests started
have been completed, and over 1600 hr of testing
time has been accumulated. The conditions of
the tests are given in Table 2.3, and the results

 

ID. R. Ward, W. C. Tunnell,and J. W. Kingsley, ANP
Quar, Prog. Rep. june 10, 1955, ORNL ]896 _p 33 _

 

TABLE 2.3. CONDITIONS OF TESTS OF MF-2 PUMP LOWER SEALS MANUFACTURED BY
THE FULTON SYLPHON DIVISION
Oil temperature: 200°F
Shaft speed: 3000 rpm

 

Test Material

Flatness? (bands)

Rcmo of

Preload Pressure Bearmgb Jownal Radius

 

No. Seal Nose Wear Ring

Seal Nose Wear Ring

(lpy  Differential \ 4 (p) to Radial

 

(psi) Clearance
1€ Sabeco 99  Ketos® 2 12 50 0.5 040300 800
2€  Graphitar 14 Ketos 6 3 30 0.5 0 to 300 900
3¢ Sabeco9  Ketos 8 3 30 0.5 010300 1250
4/ Sabeco 9 Case-hardened steel 4 4 50 0.5
5/ Sabeco 9 Ketos 2 50 0.5
6/  Sabeco 9 Case-hardened steel 3 1.5 20 1310 16
7/ Sabeco 9 Ketos 3 3 20 0.5 |
8¢  Graphitar 14 Case-hardened steel 4 3 20 0.5 0to200 810
12/ Graphitar 14 Ketos 8 6 20 0.5 o
6A Sabeco 9  Case-hardened steel 3 1.5 20  0.5t07.5
6 6 20 0.5

7Af \ __vVS'\ql\?Aeco‘ 9 7 Ketos

 

Meoeeiec-l‘ WI?.]"I Eehdm light; one band equivalent to 11.6 uin.

‘bBearmgs made of ASTM-B-144-49-36 bearing bronze.

' CTesfs made in beurlng and seal testing facility.

., ‘dSabecO 9i isa feaded bronze

| - eKettbs is an 18-4-] type of tool steel.

- ,,_fT_esf made m_cold shakedown facility; ne bearing loads applied.

 

Earlier tests had mdlcafed* '

 

o .

 

b

 

 

 
 

24

-

K

%
i

 

 

 

are presented as Fig. 2.3, It may be noted that
the low specified leakage rate has not been met,
and only four seals had rates lower than 1 em3/hr.
It was noted that the Fulton Sylphon seal was not
balanced, and the resulting pressure changes
caused variable performance of the seal, Repro-
ducibility of results from seal to seal has not been
possible. In similar tests the operation of the
upper seal was satisfactory, with leakage rates
of less than the specified 20 cm3 per 24 hr being
attained in six tests.

PERIOD ENDING SEPTEMBER 10, 1955

The lower journal bearing of the pump was de-
signed initially to carry loads of up to 600 b, but
hydraulic studies revealed that the expected
bearing load would be 150 ib or less.. With the
lower loads it appears to be feasible to employ
Inconel as the journal material and to thus obviate
the need for a hardened journal bushing, Tests 1,
2, 3, and 8 were conducted in the bearing and seal
testing facility, and loads up to 300 Ib were satis-
factorily carried by the Inconel shaft in a bronze
bearing.

UNCLASSIFIED
ORNL-LR~DWG 9048

 

700

 

TEST No.
TEST No.
TEST No.

 

. )
600 : : 7

TEST No.
TEST No.

 

 

500

TEST No.
TEST No.

TEST No

OO N oG ph

74
.42

A

 

TEST No. {

>
<P x o OO0 8> @ —]

 

400 — /

 

300

* ACCUMULATED SEAL LEAKAGE {cm®)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

- Frirlg. 2.3, Results of Leak&ge Tests of the MF-2 Fuel P"u'mp Lower Seal,

 

43

 

 

 

 
 

i G

 

 

 

 

ANP PROJECT PROGRESS REPORT

Short-Clrcmi Pump-Test Si'cmd ' |

5 M. DeCamp, Je.

A;rcrafi Reactor Engmeerlng DW|S|on '

~J.B. Kerchevcl
Prcn‘t & Whlfney Aircraft

Fobrrcahon cnd ossembly of the Ffirst short-
‘,__cnrcun' pump-fest ‘stand, described prewously,

have been essem‘:ally complefed ‘and water tests
have been started, These water tests are for

_ "_checkmg mechumcal ‘fits and interferences, check-
~ “ing pressure breakdown bypass flow rates, and
'"'correlahng head and flow data from this loop with
those obtained on the water test stand.’ '
Assembly of the pump and the volute indicated”
- _rhcn‘ the radial seals with metallic O-rings were
o uTvety sensmve d!menswna”y. A 0002-|n. intet-
'“"‘f'i’":fference on a 6-in.-dia O-ring was not enough to
o seal, while a 0008- to 0.010«in. interference
~ caused difficulty in assembly and disassembly.

Data were taken on bypass flow rates to deter-
mine the effect of the left-hand threads used in
the flow breakdown annulus. With a radial clear-
ance of 10 mils and thread depth of 37 mils, a by-
pass flow rate of 2.3 gpm was measured at a pump
speed of 2700 rpm; the main circuit flow rate was
approximately 630 gpm. Increasing the radial
clearance to 15 mils gave the desired flow rate of
4,77 gpm.

The data obtained in the water tests will be use-
ful in analyzing the data obtained at high tempera-
tures, since it will be difficult to measure dis-
charge pressures when the system is operating at
high temperature. Throttling orifices for the loop
were calculated to give a 50-ft head at 650 gpm.
Data obtained in the water tests indicate a flow
rate of 650 gpm at a 46-ft head. The test loop is
now being readied for operation at the design tem-
perature, 1400°F,

High-Temperature Pump-Performance-Test Stand

R. Cumry H. Young |
Pratt & Whitney Aircraft

The design layouts have been completed for two
loops for testing MF-2 ART-type pumps at tem-
peratures up to 1400°F, Callbrcn‘ton, shakedown,
cmd endurance tesfs on MF 2 pump rotary assem-

 

25, M. DeCcmp, ANP Quar. Prog. Rep. June 10, 1955,

ORNL-1896, p 35.

3G. D. Whitman, R. L. Brewster, and M. E. Lackey,

7 -"‘;ANP Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 32.

.:b-lies wnli Be: made wnth NaK or sodwmand wa'rh
“the fuel mixture NaF-ZrF4-UF4 (50-46-4 mole %)

as the circulated fluids. Data obtained at 1200
and 1400°F on shaft speed, head, flow rate, and
power input will be compared with water-perform.
ance data, and cavitation and vibration characteris-
tics will be noted. An endurance test will then be
run at design head and flow rate, The xenon-
removal system, inCIuding the helium-bleed and
o:l !eakage-removal sysfems, is to be incorporated
in the circuit for checking in conjunction with the
main fluid circuit. One of the two test stands will
be available for acceptance testing of the final

__MF-2 rotary assemblies with NaK.,

A concentric p[pe des:gn was chosen for the
test stand loop to eliminate thermal stresses from
nonuniform heating or cooling, to reduce the num-

bet of critical welds, und to provide a compact

assembly, The compactnessof the cssembly makes
possible a low fuel inventory.

A conventional venturi is located 15 pipe diam-
eters (60 in.) downstream from the pump volute
discharge, and a piston-type throttle valve is
located downstream of the venturi. The valve
travel was specified within the recommended travel
of a suitable bellows used as the valve-stem seal.
This valve cannot be fully closed; therefore, as
presently conceived, the minimum loop resistance,
with the valve wide open, will permit testing at
about one-half the rated pump head with rated flow,
and the maximum loop resistance, with the valve
closed, will permit testing at approximately one-
half the rated flow with the pump delivering the
rated head.

A screen-type, axial-baffle, flow=control device
is provided just below the inlet to the impeller to
prevent an uneven velocity profile and prerotation
in the stream entering the pump. To simulate the
fluid-expansion and xenon-removal region, a sepa-
rate chamber is provided above the top plate of the

pump volute. Fluid enters and leaves this chamber

only through the pump barrel wall, and this cham-
ber is not directly connected to the pump inlet

" chamber except through the impeller. The rise in

level of the fluid in this chamber for a 100°F in-
crease in the temperature of the fuel mixture will
be about the same as that for the ART.

During steady-state high-temperature operation,
the pumping power will be removed by blowing a
high-volume low-pressure air siream transversely

across the outer loop pipe. A power failure will

 

 

-

 

e T T T T Y I o

T

 

 
 

 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

cause automatic cutoff of the cooling air, and it is tent in the NaK was measured by means of a plug
anticipated that the fuel can be dumped before it  indicator and by chemical analyses of samples
freezes in any section of the loop. taken from the pump bowl, The plug indicator is a
- o _ concentric pipe stem connected to the NaK circuit,

HEAT EXCHANGER TESTS with a filter containing five 0.030-in.-dia holes

E. R. Dytko located at the bottom of the inner pipe, which plugs

Pratt & Whitney Aircraft with precipitated oxide at the saturationtempera-

R; E. MacPherson ture of the oxide in the NaK., Both the plug indi-

cator and the chemical analyses showed that the

Aircraft Reactor Engineering Division _
L . cold trap had reduced the initial oxide content of

| lntermefliaié Heat Ethangér Tests 1000 ppm to values of 150 and 650 ppm, respec-
R. D. Peak  H. M. Coopér: L. R. Enstice ?ivelhy, ofs found .by the two me:fhods after .'rhe fi‘rsf
Prat’r&Whltney Aircraft 56 hr o operaTmn. This discrepancy in oxide
content, as obtained by the two methods, cannot be

The new experlmentul assembly (sfond A) for explained at this time.
intermediate heat exchanger fests, described The second cold trap test was of 193-hr duration
previously, is shown in Fig. 2.4. This apparatus  (the period from 373 to 566 hr of operation) with
was operated for 690 hr in a series of furnace and NaK flow rates of 50 to 110 gpm and NaK tempera-

diffusion cold trap tests of the NaK system and @ tyres ranging from 1000 to 1400°F; the tempera-
2-hr cleaning cycle of the fluoride-fuel circuit. A tyres in the cold trap were below 750°F. In this

chronological description of the tests that were  test, only the plug indicator was used to measure
is given in Table 2.4, that the oxide content was reduced from an initial

As stated in Table 2.4, the 1-Mev gas-fired fur- 1000 pom 1o 150 ppm after the first 168 hr of NaK

nace (Struthers Wells Corp. ) for heating the NaK, as circulation, This level of oxide content is con-
installed, did not give satisfactory performance. sidered to be too high for a NaK-Inconel system.
The best over-all thermal efficiency that could be One tube bundle of the fuel-to-NaK intermediate
obtained was 28%, with 1.7 x 10® Btu/hr being  heqt exchanger (No. 2), described previously,
transferred to the NaK. In an effort to improve the  geveloped o leak after approximately 350 hr of
performance, a 12-ft-high stack extension was  gperation of the NaK circuit, The fuel had been
placed on top of the original 4-ft-high stack. As a  circylated for only 2 hr in the fuel circuit during

result, a thermal efficiency of 31% was obtained,  this period. At the time that the leak occurred,
with 2.83 x 108 Btu/hr being transferred to the  the fuel circuit was empty, except for small un-
NaK. Since the heaf transfer was still below the  drainable portions in the pressure transmitters, fuel
design requirement, a new, larger burner was then __pump, and the lower ends of the heat exchanger

installed, Wthh proved to be capable Of t"‘Z’V‘s't,_‘:_m‘ube bundles, With a pressure differential of 35
ferrlng 3.85 % 106 Btu/hr (] 13 MW) to the NaK ... psi across the NaK and fuel circuits, a consider-

'7 Wh'te bU"“'“Q 8400 ‘scfm Of:NU’fUW‘ 905; ‘the thermar able wquanhty of NaK flowed ln'ro the fuel circuit

efflmency wn‘h th“

 

 

er was 487 wnhi

 

i"e'__ the leak was detected, The leak was de-

 

  

_‘-',.';',_j:._furnace mief Aqnd exit jemperaf;lres of ]307 an&; ‘ fi_“"tectédflwhen the Iower spark plug in the fuel pump
1600°F and’ aN

 

 

rate of 140 gpm, o “shorfed . ‘Sbbsequent dumping of the NaK circuit
Two tests were ¢o dUC’fed fO evaluate the func- ‘_resulted in contamination of the NaK circuit and

-_'”°“ of the diff‘p_slon C°|d "C’P (0 cylmdrlcal C°"'.“_ the NaK dump tank with some of the fuel which
o tainer, W”h a»yo!ume 17 of that of the NaK circuit, could not be completely drained following the 2-hr
”i',connected to the main circuit by a pipe 2 in. in  operation of the fuel circuit.

 
  
  
 

“Q_fh) in removing sodlumm;w._" The fuel-to-NaK heat exchanger was removed
first, or prehmmary test,  from the assemb[y and exammed Pressure tests

and a NaK temperature of 1500°F.  The oxide con-

“ 4R, D. Peak, ANP Quar, Prog. Rep. June 10, 1955, 5P, Patriarca et al., ANP Quar. Prog. Rep. June 10,
ORNL-1896, p 37. 1955, ORNL-1896, p 131.

 

 

45

(Thé} Pe”Od from 50 to 219 h" showed that only one tube in one of the two 100- -
_°f °Pe"°'“°") W”h NaK flow rates of 55 10 85gpm  tube bundles leaked. The leak was found to be a

o

s

 
 

ANP PROJECT PROGRESS REPORT

TABLE 2.4, SUMMARY OF INTERMEDIATE HEAT EXCHANGER TESTS :

 

Hours of : ,
L Remarks : ¥
Operation _

i

 

0 B - Fllled system wnth NaK (56% Na-44% K) at room femperature. o

0to 16 7 . "'Nal( system elecfncally heated to 300°F NuK flow rai'e, 100 gpm, hlghesf NaK tem-
‘ ' perature oh‘mnable wnfh electric heufers found to be 300°F. |

16 to 50 NuK system operated mfermuflenfly at various flow rcn‘es cnd femperutures ‘to check
' opercblllfy of 1-Mw gas -fired furnace; furnace found to be unsqhsfoctory, ‘thermal
' efflclency, 28%, heat transferred fo NaK, 1.7 % 108 Bfu/l'lr. NqK pump had to be

replccecl at end ‘of 50 hr of operahon beccwse of fmlure of the onl-fo-helnum seal

50 to 219 ~ NaK system operoted |sotl1ermc|lly at lSOOoF wnfl’l NaK flow rafes of 55 to 85 gpm to o E
: - check operahon of diffusion cold trap, cold trap reduced oxygen content of NaK' from $

- 1000 to 150 ppm, accordmg to plug indicator, or 650 ppm occordlng to chemical ana Iy-

Csis. NaK replaced wnh new supply because of high oxygen confenf NaK pump agcnn '

'replaced at end of 219 hr of operahon becouse of beormg fcnlure

 

75t0 77 | Fuel circuit filled for cleamng wflh NaF ZrF UF (50 464 mole %) and Operated |
- lsofhermolly at lSOOOF for 2 hr at a fuel flcw rate of 60 gpm; fuel then dumped.

219 to 354 NaK system operafecl |m‘erm|flenfly at various flow rates and fempercn‘ures to check
gas-fired furnace; furnace found to be satisfactory after alterations; thermal efficiency,
48%; heat fronsfer to NaK, 3.85 X 106 B'l'u/hr {1.13 Mw).

354 Blower started with NaK c:rcuuf |sofhermal at 95°F fo check rud:utor air flow.

354 to 358 NaK system operated lso’rhermolly at 'l400°F at various NaK flow rates fo check

pressure drop in system., Test terminated because of a leak in the heat exchanger.

 

Heat exchanger removed for examination and fuel pump removed for cleaning. NaK

 

piping installed to feplace heat exchanger.
358 to 373 NaK circulated at 100 gpm with system isothermal at 1400°F for cleaning circuit; NaK

drained while hot; NaK replaced because it was found to be contaminated with fuel.

373 to 566 NaK system operated isothermally at 1000 to 1400°F ot various NaK flow rates to
check effectiveness of diffusion cold trap; cold trap reduced oxygen conten'r from En
1100 to 150 ppm, as determined by plug IndlCO‘l’Ol‘

566 to 621 Blower started to create temperature dtfferenhql of 200 to 2500F across NaK-to-orr

radiator; Nak sysl’em ‘operated at various NaK flow rates and NaK temperatures of

 

11000 to 1250°F Radiator gradually plugged with fuel contaminants. Plug mdlcator
_showecl oxygen confem to be too low to cause plugging. Test terminated because of

a leak in fhe radiator. Radlofor removed for exammahon. Cold trap, plug lndlcofor,

 

""‘_'.cnd Nal‘( sump removed for clennmg.

621 to 690 o _ " NaK system bemg cleaned

 

radial crack on the |n5|de of the tube bend. The was flcwing-'fhreugh fhe tubes at a .:cliffe‘re‘rllifeln-'
100-tube bundle and header that leaked is shown perature than that of the heat exchanger shell.

 

before and after operation with NaK for 358 hr in
Figs. 2.5 and 2.6. The tube that leaked is indi-
cated in Fig. 2.6.

The tube distortion shown in Fig. 2.6 was caused
by fhe thermal cycling that occurred when NaK

46

There were 73 instances in which this tempera-

ture difference ranged from 200 to 1000°F, Since
no fatigue cracks were found in the 36 tubes that
could be visually inspected, it is thought that the
one tube that failed must have had a flaw that

 
   

. {rihe, Mpu . gullEk. il G L e R
- ' -"n o B oom

-

 

; - . - . . LAk - .
- o ", L : Sam . : Lol e : «

  

       

      

D,
i

 

~‘,= ‘
L

pske

 

D 0
. s
L i ~
.4“‘4
s .

s

2

 

 

 

 

Ly
§S61 ‘0L ¥39WILdJ3S ONIAONI QOl¥3d

   
 

 

 

  
  

 

 

 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
     

  

' made it susceptible to stress failure,

After the heat exchanger had been removed and
replaced with piping, operahon of the NGK system
was resumed. The NaK was circulated for 15 hr
at 1500°F and then drained while hot, Since the
NaK contained fuel contamination, it was rep[aced
with fresh NaK, and the second cold trap test de-
scribed above was made. However, the operation

radiators, described prevnously,6 developed a leak.
During the last 55 hr of operation, a “steady in-
crease in the NaK pressure drop across the radiators
was noted, The pressure drop increase coincided
with startup of the radiator blowérs, which caused

the radiators., The increase in pressure drop,
along with a gradual decrease in NaK flow at a
constant pump speed, could only mean that some
of the r_adlc’ror tubes were pluggmg. These plugged

 

gl

tubes were cooled by the air flow from the blower,
with the result that there were low-temperature
regions in the otherwise high-temperature radiator

  

Fig. 2.5, The 100-Tube Bundle and Header of
..Intermed:ate Heat Exchanger No. 2 Which Leuked"'
After 358 hr of Operation with NaK Circulating in

the Tubes. | S1bid., p 134.

 

 

had to be stopped again when one of the NaK-to-air

a NaK temperature differential of 200°F _across

 

 

 

 

 
 

“

 

 

matrix. The high thermal stresses thus created
were probably of significance in the ultimate
failure of the one radiator unit, The radiator that
failed is shown in Fig. 2,7.

The fuel that leaked into the NaK circuit after
the heat exchanger failure was not flushed out
during the 15-hr period of circulation of the NaK
after the removal of the heat exchanger. The
analysis of the black material scraped from the
fins of the NaK-to-air radiator when it was being
examined after the leak occurred showed 34%
zirconium and 0.25% uranium; the remainder of the

. 'material was sodium, potassium, iron, nickel,

chromium, and copper.

The NaK circuit was cleaned by the Materials
Chemistry Division. A total of 4.2 g of zirconium,
0.46 g of uranium, and 3939 g (8.6 |b) of Inconel

‘was removed in the cleaning operation. Two 5%

nitric acid washes removed 87% of the uranium,
and one wash with 2% ammonium bifluoride and
2.5% nitric acid, a highly corrosive agent, removed

82% of the zirconium, 'qu ammonium bifluoride
and nitric acid washes removed 90% of the Inconel.

 

 

PERIOD ENDING SEPTEMBER 10, 1955

The Inconel removed by the cleaning process
reflects corrosive attack to a depth of 1.5 mils
on the inside walls of the Inconel piping throughout
the NaK system, if it is assumed that the attack
was uniform,

Test stand A is currently being rebuilt. Radiator
units made by the York Corp. and heat exchanger
bundles made by Black, Sivalls & Bryson, Inc. are
being used. Test stand B, which is almost identi-
cal to stand A, is now 70% complete. It will be
used to test two radiator units (500-kw size) built
by Pratt & Whitney Aircraft and two heat exchanger
bundles made by the Metallurgy Division of ORNL.,
Test stand C, another stand identical to stand A,
is now 10% complete.

Small Heat Exchanger Tests
J. C. Amos
Aircraft Reactor Engineering Division
L. H. Devlin J. S. Turner
| Pratt & Whitney Aircraft
The first of a series of tests of small fuel-to-
NaK heat exchangers was terminated on June 29,

UNCL ASSIFIED
Y- 16065

 

Fig. 2.7. NaK-to-Air Radiator Unit from Intermediate Heat Exchanger Test Stand A After NaK Leak

and Fire. The side cover has been cut away.

 

49

TE TR

 
S R St R B ekl i o

 

 

~ ANP PROJECT PROGRESS REPORT

.]955, dffer:-"Séo hr of opercfiofi. | The test as-

sembly, which was described previously, along

‘with a presentation of the preliminary results,” has
o been dismantled, and the components are curren'rly
o bemg sub|ecfed tometallurgical inspection. A sum-
) mary ‘of the cpprOXImofe operating conditions is
~given in Table 2,5, The heat transfer data obtained
_are presented in Fig. 2.8 and are compared w'ifh'the'
~résults from the theoretical relationship Nu/Pr0+4 =
. 0.023 Re® 8, with data obtained for a 100-tube
'jhecn‘ exchanger ‘operated with water,® and with a

' *curve showmg the data for the 20-tube heat ex-

' chcmger ad[usted for a IOO-tube heat exchonger

 

7'J C Amos, M. M. Yarosh, and R. I. Gray, ANP Quar.

. Prog. Rep. June 10, 1955, ORNL 1896, p 37.

8). L. Wantland, ANP Quar. Prog. Rep. June 10, 1955,

' '.*’,,_':ORNL 1396, P 149

 
  

The equivalent diameter used in calculating the
Reynolds number was based on the total wetted
perimeter of the tubes plus the side-wall area.
The fuel-side pressure-drop data obtained are com-
pared in "Fig. 2.9 with a theoretical pressure-drop

curve calculated from data obtained from water

’resfs carried out on similar tube bundles ’ro deter—

" mine the pressure drop created by the spacers.

The calculated pressure drops are higher than
those that would be predicted by fheory because

of the presence of spacers.

During the last 456 hr of operahon of fhls test
the NaK inlet and outlet temperatures were main-

" tained at approximately ART design conditions

to study mass-transfer effects. No appreciable
amount of mass transfer was detected by visuval

inspection when the Ioop was dismanfled. During

  

| TABLE 2.5, SUMMARY OF SMALL HEAT EXCHANGER OPERATING CONDITIONS”-J?' B

 

 

Hours Fuel

NaK NaK Mean Fuel

 

 

Fuel Mean Fuel
. of Reynelds  Temperature AT Reynolds Temperature AT
O peration Number (°F) (°F) Number (°F) °F Remarks
7% 1300-5500 1300 15 45,000 (min) 1300 30 Pressure drop data
184,000 (max) being obtained
2%%  1000-6500 1400 15 26,000 (min) 1400 30 Pressure drop data
187,000 (max) being obtained
444 1400-5500 1350 97 (min) 74,000 {min) 1225 (av) 142 (min) Héat transfer data
: ' 200 (max) 206,000 (max) 200 (max)  being obtained
272 500-6400 1350 90 (min) 40,000 (min) 1225 (av) 126 (min) Heat transfer data
228 {max) 290,000 (max) 240 (max) Being obtained
256 2850 1350 135 160,000 1275 130 Endurance run
472 3500 1450 115 145,000 1400 135 CEndurance run
168 3400 1360 20 135,000 1350 35 Thennalcychng
| | e L. test
16 600 1300 272 40,000 1225 290 ; Thennalcychng
6% 600 1325 280 40,000 1240 380 Thermal cychng
456 4000 1455 100 40,000 1250 490 .'Massfrunsferdata -

1557f'

bemg obfo:ned

 

 

INaK with fhe c0mp051t|on 22% Na-78% K used for the first three runs; NaK with the composfluon 56% Nq-44% K o

used for all other runs; fuel mixture for all runs was NaF-ZrF -U F (50+46-4 mole %).

Thermcd cycling discontinued after two complete cycles,

. c'-Buh,rl ‘alcohol circulated in NaK Loop for 19 hr after termination of test,

 

 

 
 

 

.Nu/ProA\‘:.’::;; i e \ ’ & :

-t

U[;‘_L;‘;P_RESS"URE‘{: DRQE’,‘;-ACROSSESHEAfi EXCHANGER - (psig)

 

 
  

 

 

 

 

S A e e “ ORNL-LR~DWG 9049
30 T T -
& DATA TAKEN DURING L -
[ INITIAL OPERATION” ' 1A
1 G DATA TAKEN AFTER S 7
[T 500 hr OF GPERATION ° i
' ® DATA TAKEN AFTER - A 4A
L 4000 hr OF OPERATION” v //
20 L -i 1 ,7 ’l - A.
R

o4 _ 0.8
Nu/Pr% = 0.023 Re‘. —a/
Ve

io. 1 / 7|
- ¥
o

v
-/ / :
5 ‘ ) X WATER DATA FOR
L .| 100-TUBE HEAT EXCHANGER

™~
| {™20-TUBE HEAT EXCHANGER
DATA ADJUSTED FOR 100 TUBES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10,000

 
   
 

- Flg. 8 Heat Transfer ‘Dutq for 20;T‘ube Fuel;
to-NuK Heat Exchunger. )

 

ORNL-LR-DWG 9050

 

 

 

 

 

 

 

R N — &
| A DATA TAKEN DURING 7
| INITIAL OPERATION ' o/
| O DATA TAKEN AFTER Y
500 hr OF CPERATION 9 /
50 |—® DATA TAKEN AFTER iy
1000 fr OF OPERATION o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig‘. 2.9. Fuel-Side Pressure Drop Data for
20 Tube Fuel fo-NaK Heat Exchanger.

 

PERIOD ENDING SEPTEMBER 10, 1955

this period of operation the efficiency of the NaK-
to-air radiator used as a heat dump dropped con-
tinuously for approximately 360 hr and then im-
proved suddenly. Samples taken from the NaK
sump tank at the completion of the test indicated
a very high oxygen content in the NaK (approxi-
mately 2500 ppm). A circulating cold trap has
been designed and will be installed on future
test assemblies in an effort to remove oxides
from the NaK.

Assembly of two, new, small heat exchanger
test stands is opproximately 25% complete. Each
of these stands has a 500-kw electric heat source
and will be used to test six 20.tube heat ex-
changers, Two heat exchangers are being fabri-
cated by ORNL and four have been ordered from
outside vendors.

A 25-tube heat exchanger is being designed,
with ART tube sizes and spacing, that will be
operated at ART design temperature and flow
conditions. [t is proposed to procure five of these
units from outside vendors. A block diagram of
the design operating conditions for this group
of heat exchanger tests is presented in Fig. 2.10.

STRUCTURAL TESTS

E. R. Dytko
Pratt & Whitney Aircraft

G. D. Whitman

Aircraft Reactor Engineering Division
Outer Core Shell Thermal Stability Test

D. W. Beli
th‘ & Whn‘ney Aircraft

A one-fourth scale model of the Iower half of
the 2]-m. recctor core shell was fabrlccn‘ed for
tesfs of fhermal “and’ s’rructural Sfdblllfyr when
sub;ected to cyclic themal stresses at reactor
operahng conditions. The model, shown in Fig.
2.11, was muchmed from Inconel bar stock and

"/__‘/_\gvd‘sfi__r_'jxsjl"_g_‘VS'"elleved before the_ flncl mochmlng

Wi ‘e helcl to 40.005

       

in, ’the dlcmeter cncl +0 003 in.on the thickness.

The - same ]ength-to-dmmefer ‘and” fhickness-to-
’7 those of the fU“ srze _

7 e"full-sne core she“ durmg reactor operation.
“An ]nconel housmg is being fabricated that will

form a /10"”' annulus on both sides of the shell.

51

 

it

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A
Co "ANP PROJECY PROGRESS REPORT
1 0 S sost”
1 - 400 kw _
A L
4 - 70°
4 - _ | NaK-To-AIR 1070°F
1 RADIATOR 25 pslg
§ APygk = T psi
g 25 psig y
&
§ . NaK
4 PUMP -
: ’ 34 gpm
A 144 psig
1 AT = 430°F
i BPyox = 82 psi
4 1070°F
; FUEL-TO-NaK 4 psig
; - HEAT EXCHANGER
: _ ¥F 1250°F
66 psig 25 psig
3 ATyg = 350°F
g AP e = M psi
4 25 psi
{ psig ¥
] FUEL
- PUMP ~
9.6 gpm
7 psig
APeyg, = 5 psi
4600°F RESISTANCE 4 1250°F
66 psig HEATER T4 psig

 

 

 

‘400 kw

Fig. 2.10. bes-tén'éonaitioee fet Test.s of 25.
Tube Fuel-to-_qu_ Heat Exchanger.

placed circumferentially every 60 deg. The large-

diameter end of the shell will be welded to the

housing, and the small-diameter end will be at-

tached to a bellows so that there will be no stresses

. on, ‘the sheli that are due to axml thermal expan-
- sion.

Sodlum er be used to 1mpose a rcdlal tempera-

  
  
   
  
  
 
   
 
 
  

‘:temperature ‘of the core ‘shell ot this point will
‘be between 1450 and 1500°F. ‘Sodium at 1550°F
‘will e enter the inner annulus at the smal! dtameter
end and [euve at T300°F Sodlum at. 985°F will

low counfercurrently in the outer onnulus and

‘eycled ‘between the above "conditions ‘and iso-
,-'therma]_operatton at 1200°F for 100 cycles, After

 

- of Zhfii‘ -ART Reactor Core'_‘_Shefl Fabncutecl for ¥

The outer annulus will contain six axial spacers

: ure dlfferenttu] of 300°F across “the core shell
‘at’ the small-dnameter end The msu!e surface' '

-“Iemfi/’e at 1065°F. The core shell will be thermally

s test | has ‘been. completed, |t has been planned 'Fig. 2.12. Fabrication has ‘been “started on twc»;'

ertain wBether the core shel! wnll buckle o

 
    
  

 

Fig. 2.11. One-Fourth-Scale Model of Lower Half L E

Tests of Thermal Stability.

under an extemal sodium pressure if the fuel
pumps should fail during reactor operation.

 

 

The test stand used for the first small heat
exchanger tests is being modified for these tests,
Heating and cooling of the sodium will be accomp-
lished with a 200-kw resistance heater and an' . ) i F

'ARE sodlum-to-cur rcdlctor. L

J. C. Amos LA M&nn '

Aircraft Reactor Engineering Division

C. H. Wells
Pratt & Whitney Aircrafi

A desagn has been c°mpieted of an “anvil bend-

| :mg test’’ apparatus for obtaining basic mformatlon

on the behavior of Inconel under strain cyclmcl
at elevated temperatures in both lnert cmd corrow
sive atmospheres. This opporotus is’ sflown

of these units. Affer shakedown operation of
o

    
   
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
   

 

 

 

 

 

    
 

 

 

 

 

 

 

 

 

  

 

 

 

 

   
 

 

 

 

 

 

i PERIOD ENDING SEPTEMBER 10, 1955
. +
1. UNGLASSIFIED
g CRNL-LR-DWG 9052
é -
' SPARK PLUG
PH_OB_E FILL LINE
THERMOCOUPLE THERMOCOUPLE
. FITTING FITTING
GAS CONNECTION
»
ch S1— RUBBER 0-RING, ,-in. 0D x ¥g-in. ID
x 0.070-in. DIA. {5 REQUIRED)
PEDESTAL-——__|
kR
- TOP FLANGE ASSEMBLY
] ,\TB\
r o ACTUATING ROD
|-t /
O-RING
&
L LIQUID LEVEL
 ALIGNMENT PL. /
UPPIfR ANVIL\ /TEST SPECIMEN
| |
. LOWER ANVIL - .' | TANK
*
- L ' .
: Fig. 2.12. Inconel Strain-Cycling Test Apparatus.

 

 

53

 

s e

e

 

 

 
 

 

 

 

 

ANP PROJECT PROGRESS REPORT

these two units the design will be reviewed and
revised, if necessary, and three additional units
will be fobricated and placed in operation.

 Eachtest assembly will accommodate two Inconel
sheet test specimens 2?8 X ]/4 X ]/8 in. Interchange-

able anvils with radii of 12, 15, 20, and 31 in.

will be used to prowde various strains,

lt is plcnned fo run the first test in an atmos-
o phere of heltum at 1200°F with 12-in.-radivs anvils
. that will impose approximately 1% strain. One
- specimen will be eycled on o 2-hr half cycle, and
' 'j\another will be cycled on a }’-hr half eycle, The
~purpose of this test will be to obtc:m an indication
Vof the reloflve relcxchon times.

B Dufc w:“ be ob’ramed on per cent of strain,
. length of cyc[e, number of cycles to fracture,
- effect of temperature, and effect of surrounding

'-'medxum {for example air, helium, sodium, NaK,

fuel). As the tests progress, the specific combi-
nations of conditions to be studied further will
be evaluated.

Thermal«Cycling Test of Sodium=Inconel-
Beryllium System

M. H. Cooper R. D. Pedk
Pratt & Whitney Aircraft

The third test to be operated in the sodium-
beryllium-Inconel compatibility testing apparatus,
described previously,” was started on July 14,
1955. The test section consists of a beryllium
cylinder drilled centrally with a };‘-in.-dia hole
through which sodium can flow. The loop was
operated isothermally at 1100°F for 100 hr, with
a sodium flow of 3 gpm, ond then shut down on
July 18, 1955, because of the failure of the 250-kw
transformer. The loop was restarted on August 27,
1955, with a projected operating period of 1000 hr
with 100 thermal cycles. One cycle represents
4 hr of operation at high power, during which the
sodium enters the test section at 1150°F and is
heated electrically to 1300°F, and 4 hr of operation
at low power, during which the sodium enters the
test section at slightly less than 1300°F and
leaves at 1300°F,

 

L 9P Patriarca et al., ANP Quar. Prog. Rep. Mar, 10, |
L 1955, ORNL 1864, p 134,

,__547

COLD TRAPS ANDPLUG |NVDICA-T0RVS'
F. A. Anderé_on
University of Mississippi
J. J. Milich
Pratt & Whitney Aircraft

A survey of the literature and of the available
experimental data was undertaken to obtain infor-

“mation on the use of cold traps as devices for
removing or controlling the presence of undesir

able impurities, especially oxygen, in sodium-

‘or NaK«filled Inconel systems and on the use of

plugging indicators as devices for determining
the oxygen content of sodium or MaK streams.
In addition to information obtained from the litera-
ture, supplementary data were obtained during

" visits to the Knolls Atomic Power Ldboratory,

Argonne National L aboratory, Mine Safety Appli-
ances Company, and North American Aviation,

Although both diffusion and circulating cold
traps have been shown to be capable of reducing
the oxide content of liquid sodium or NaK to values
corresponding to the saturation concentrations at
the cold trap temperatures, specific data are not
available which will permit sound engineering
designofcoldtraps. Diffusion or natural-convection
cold traps do their work much more slowly than cir-
culating or forced-circulation traps and, so far,
have been of primary interest only in relatively
small-volume systems (up to a few gallons capac-
ity). The existence of fairly strong eddy currents
in diffusion cold traps is suspected. If such
currents existed, their effects would overshadow
completely the effect of natural diffusion. It is
evident that experimental work would be required
to obtain a basis for the design of diffusion cold
traps.

Circulating cold traps have the important advan-
tage of lowering the oxide concentration of a sys-
tem rapidly. Present conventional designs that
provide for holdup times of 5 min and superficial
liquid velocities of the order of 3 fpm give system
cleanup times that correspond to three system
charge turnovers through the trap. Cold traps
having volumes ranging from about 3 to 10% of
the volume of the system have been utilized suc-
cessfully. Although microporous filters are con-
sidered to be unsatisfactory in cold traps, the
use of a packing material such as York demister
packing has been shown to raise cold trap effi-
ciencies from 68 to about 98%.

Packing is used

L

 

 

 

 

 

 
 

* to provide sufficient surface to hold the precipi-

. tated oxide, but it is not intended to function as
' a filter medium, _

On the basis of the collected data and arbitrary

i° specifications, three circulating cold traps have

been designed, one for a relatively large, 80-gal
system and two for a relatively small, 4000-ml
(~1-gal) system, Figs. 2.13, 2.14, and 2.15. These
designs are not necessarily optimum; that is, it
is felt that they will lower the oxygen content
: of the system, but they may be larger than neces-
sary and may have more cooling or heating capacity
than that required. The traps are expected to
operate at 400°F and to reduce the oxygen con-
tent of the main liquid metal system to the cor-
responding saturation value of 50 ppm. During
operation, precautions must be observed to prevent
the saturation temperature of the liquid metal
with respect to oxide (Na,0) content being reached
* ~ at any point in the system other than in the cold
trap. If the saturation temperature were reached
outside the cold trap, precipitation of oxide would
occur at that point and plugging of a pipe or some
other sysfem component would occur.

Because of the difficulty of making rehoble de-
terminations of the oxide content of molten sodium
or NaK by chemical procedures, considerable use
-~ has been made of the plugging indicator as a

- simple device for determining the oxide content
indirectly. Although there is some disagreement

'—fi

2

o

l—‘ My in,

3/g-in. PIPE, SCHEDULE 40
THERMOCOUPLE WELL, 2 in. LONG

       

% o] l* g

 

 

""'uss ‘/2 i DA HOLES T

  

4-in. PIPE,
SCHEDULE 40 .

 

" LEACH END TO BE PROVIDED [ &in
" WITH 1-kw WRAP-AROUND OR
CALROD HEATING ELEMENTS ‘_

UALL WELDS CRITICAL HELIARC

   
 
   
        
   

Yo-in. PERFORATED PLATES
(TACK WELD IN PLACE)

| PERIOD ENDING SEPTEMBER 10, 1955

as to the exact significance of the results, in-
vestigators at KAPL feel that the use of a plugging
indicator is a reliable and accurate method for
determining oxide contents to within 10.001%.
Unfortunately, as in the case of cold traps, the
design of plugging indicators is, at the present
time, more of an art than a science. Because the
design first used at KAPL proved to be satisfac-
tory, and because of lack of time, no effort has
been made to determine whether a better design
could be developed.

A plugging indicator consists of a perforated
plate (containing 15 to 21 holes 50 mils in diam-
eter) placed in a l-in. line through which molten
metal flows at the rate of 1 gpm (~10 fps through
the indicator plate holes). The flowing stream
is cooled until the oxide saturation point is reached,
at which time precipitation occurs. When the pre-
cipitated oxide partially plugs the perforated
plate, a change in the flow rate occurs. The
temperature at the plate corresponding to the
change in flow rate, measured with an electro-
magnetic flowmeter, is taken to be the plugging
temperature,  Although this temperature can be
used as a relative indication of the purity of the
liquid metal stream, it is common practice to
translate the plugging temperature into an oxide
concentration by means of a solubility curve.
A recommended design for a plugging indicator
and a bypass loop is shown in Fig. 2.16.

UNCLASSIFIED
ORNL-LR~DWG 2053

Y4-in. COPPER COOLING COILS

(BRAZE IN PLACE)
Ya- x 0.035-in. THERMOCOUPLE WELLS, 2% in. LONG
YORK DEMISTER PACKING {INCONEL)

 
 

Y4-x 0.035-in. THERMO-
COUPLE WELL, 2 in, LONG

¥g~in, PIPE, SCHEDULE 40

z

——

2in:

g

Y, in.»‘ }-«

DRAIN LINE, Y2~in. PIPE,
SCHEDULE 40 (END TO
BE CRIMPED AND WELDED)

Hn." r-—

8 in.—————

 

 

 

sgedion s i

. USE INCONEL EXCEPT AS NOTED [ S e

 

* Fig. 2.13. Circulating Cold Trap for Large-Yolume (80-gal) Systems.

55

 

 

 

 
 

N T

 

56 )

 

ANP PROJECT PROGRESS REPORT

" UNCLASSIFIED ~
ORNL—-LR —DWG 9054

 

NOTES :

{. EACH END TO BE PROVIDED WITH {-kw WRAP-AROUND OR
CALROD HEATING ELEMENTS.

2. ALL WELDS CRITICAL HELIARC,
3. USE INCONEL EXCEPT AS NOTED.

Y,-in. SCH 40 PIPE

'3 ~in. THERMOCOUPLE WELL

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Y4-in. COPPER COOLING COIL | .
J/ (BRAZE IN PLACE) ! din. s
/ l=-Yg-in. END PLATE
o
NOTE ,//2 in. SCH 40 PIPE - E
= > .
~—LIQUID IN
. = é —1in

= i , I »1' 7 ~8in
e T 4in——= YORK DEMISTER PACKING 1 | :
" PLATE DETAIL ! (INCONEL) ; :
e 3 R 3-in. SCH 40 PIPE ——DRAIN LINE (CRIMP AND WELD END) ,
. USE “g-in. DIA. HOLES _ Yg-in. PERFORATED PLATE E
L e e e (TACK WELD IN PLACE) E
3
. F
; r >

i 10 1t 2 3
18in. mmwmm onom s ans T e INCHES .

Fig. 2.14. Circulating Cold Trap for Small-Volume (~ 4000-ml) Systems.

UNGLASSIFIED -

ORNL-LR-DWG 9055

YORK DEMISTER PACKING
{INCONEL)

     
   
   
 
 

/4 -in. COPPER COOLING COIiL

V4-in. THERMOCOUPLE WELL, 1% ~in. LONG {BRAZE IN PLACE)

 

F— 134 in.

1-in. SCH 40 PIPE i

\

 

 

 

 

 

 

 

 

 

 

 

=Yg ~in. PLATE

 

 

 

 

 

3in

 

10in. 10in.

 

 

NOTES:
1. USE INCONEL EXCEPT AS NOTED.
2. ALL WELDS CRITICAL HELIARC.

 

f 0 1 2 3 .
INCHES
Fig. 2.15. Alternate Circulating Cold Trap for Smali-Volume (~ 4000-m!) Systems. i b

 

 

 
UNCLASSIFIED
ORNL~LR-DWG 2056

i‘ PERIOD ENDING SEPTEMBER 10, 1955
4
; . MAIN SYSTEM —~

VALVE ——= ~—— VAL VE

 

 

ELECTROMAGNETIC ELECTROMAGNETIC
FLOWMETER PUMP

 

 

 

 

 

 

 

 

 

PLUG INDICATOR
PLATE

- 4-in.TEE

 

 

f-in. SCH. 40 PIPE —"

 

it i R e it b

Y4-in. THERMOCOUPLE WELL

   

1-in. SCH. 40 PIPE

AACHINE END TO FIT INTO
TEE AND AGAINST PLUG
PLATE AS SHOWN)

   

 

 
    
     
 

L e

- 0049-in.
7 7 orR
* ’ . NO.16 BWG

PLUG INDICATOR PLATE v

  

 

| STANDARD {-in. TEE
' OR EQUIVALENT

|
|

|

2

|

|

L

. 1-in. SCH 40 PIPE
- (MACHINE END TO FIT
INTO TEE AS SHOWN)

  
 

sy
V/ 2
%

 

 

 

 

Y&-in. COVER PLATE _/
(WELDED IN PLACE) Y4-in. THERMOCOUPLE WELL
(TO EXTEND TO %-in. FROM PLUG PLATE)

 

USE WRAP-AROUND HEATERS ON ALL LINES TO CONTROL TEMPERATURES AND COOLING RATES

  

57

 

 

   
 

 

 

 

  
 

ANP PROJECT PROGRESS REPORT

3. CRlTICAL :E'X‘PVER)IMENTS_
A. D Cclllhan

V. G. Harness

J. J Lynn

- E, ‘R' Rohrer

Apphed Nuclear Physacs Division
R. M Spencer, Umfed States Air Force

D. Scott, Jr., Asrcraft Reactor Englneermg Dmsnon

J. S. Crudele

E. V. Sandln |

S. Snyder

Pratt & Whlfney Aircraft

S ROOM-TEMP ERATURE REFL ECTOR-
MODERATED-R EACTOR CRITICAL
¥ EXPERIMENTS "

 

A c:rmcal assembly of the reflector-moderated

- _circulating-fuel reactor loaded with sufficient
if-f"}siUz’?’fs to give about 3% excess reactivity was de-
'scrlbed prevuously.

After the completion of a
sferles “of experlmenfs in which the excess reac-

““tivity was utilized to measure certain reactivity

coefficients and the effects of various structure

changes, the assembly was reloaded in order to

determine the clean critical concentration more

exactly. With the same structure and dimensions

as those reported previously, the U235 concen-
tration was reduced from 0.416 to 0.345 g per cubic
centimeter of fuel region, equivalent to a loading
of 20,07 kg of U233, and the assembly (CA-21-2)
was critical with only 0.14% excess reactivity.
This assembly is compared in Table 3.1 with the
assembly (CA-21-1) which had 3% excess reac-
tivity.

Two structural changes in the assembly have
also been investigated. In one modification (as-
sembly CA-22) the average width of one of the
end ducts, the portion of the assembly that simu-
lates the fuel flow channels into the reactor core,
was increased from 1.29 to 2.80 in. to make its

volume almost 2.5 times that of the originally

constructed end duct. This change was made
because consideration is being given in the re-
actor design to the possibility of using end ducts
of larger cross section, The increased fuel capac-

ity of the assembly necessitated the addition of

U235 in the end ducts, which made a total of
28.35 kg of U233 gt an average density of 0.405 g

;_of_u U_2_35 per cubic centimeter of fuel material.

 

7 ]A.. D. Callihan et al.,, ANP Quan Prog. Rep. March

.10, 1955, ORNL-1864, p 41.

 

The array was crmcal wn‘h 3 2% excess recchvny, '

and the estimated ‘‘clean” critical mass was

24 £ 2kgof U235

In another experiment (CA-23), with end ducts of
the original size, the average radius of the central
psuedo-spherical beryllium region or ‘‘island’ was
increased from 5.18 to 7.19 in., with a correspond-
ing reduction in the fuel volume of from 58.2 to
44.3 liters. The use of a thinner fuel annulus is
being considered in the reactor design as a means
of reducing fuel self-shielding. The beryllium
content of the core was increased from 67 to 92 kg,
including that in the end ducts. This assembly
was critical with a loading of 18.62 kg of U235
at an average density of 0.420 g of U235 per cubic
centimeter of fuel region. The excess reactivity
was 0.19%, and the estimated critical mass, with-

out control rods, was 18.4 kg of U235, The di-

mensions of assemblies CA-22 and CA-23 are given
in Table 3.1, along with the data obtained with
these assemblies.

HIGH-TEMPERATURE
REFLECTOR-MODERATED-REACTOR
CRITICAL EXPERIMENTS

A low-nuclear-power, high-temperature, reflector-
moderated-reactor critical experiment is currently
under way, and some preliminary results have been
obtained. The liquid fuel in the assembly is at
1200°F. The reactor section of the equipment
closely resembles the current design of the ART.

It consists, essentially, of an annular fuel region

separated from the beryllium islané"an'cl' reflector
by ]/B-in.-'rhick Inconel core shells. The island,
partly surrounded by the inner Inconel shell, is
shown in Fig. 3.1. The outer shell, with some
of the beryllium reflector in place, is shown in
Fig. 3.2, A re-entrant tube is mounted along the

vertical axis of the reactor to serve as a guide

 

-

 

 

 

h

 

 
 

PERIOD ENDING SEPTEMBER 10, 1955

] . . ; ‘“N%
, TABLE 3"[ COMPOSITIONS AN PIM ENSIONS OF THREE-R EGION REFL ECTOR-MODERATED-REACTOR
CRlTICAL ASSEMBI.IES V"ITH /8-i n.-THICK INCON EL CORE SH ELLS AND DUCTS

 

 

 

 

 

 

 

   

{ S Assembly NUmber S CA-21-1‘ - CA-21-2 T cAa22  cas
Berylhum Islqnd | i
Volume, 3 - 1,27 1.27 ' 1.27 1.75
Averqge radlus e o o e - '
" Spherical section -~ 518 = 518 5.18 7.19
"Endduets 7386 3.86 3.86  3.86
qus, kg o e 670 7.0 67.0 92,2
o Fuel Reglon (excludmg shells and
: interface plates)
Volume, #° 2,06 2.06 2.47 1.56
liters 582 582 70.0 44,3
Average rqd:us, m o ' o T e S '
Spherlcql secflon o
Inside o 531 7 s31 531 7.32
_ Outside 944 T 944 944 9.4
o heside 399 T T 399 399 399
. : - Outsmiel A o 528 528 6,79 ' 5.28
) - Distance between fuel sheets, in. ' 0.142 ‘ . 0.173 0.142 0.142
. Mass of componem‘s, kg ‘ e I
. -  Teflon | 7 108.88  ° ~  108.18 126,77 81.54
' L Uranlum |oad|ng S 26,02 21.57 30.45 19.97
- U235 foading - 24.24 20.07 28.35 18.62
= . _ Uranium dens:ty,b g:;/cm3 | 0.446 : 0.370 0.435 0.451
ouBs density,® g/em’ 0.416 0.345 0.405 0.420
Uramum couhng materlal IR 0.25 0.21 0.30 0.19
Scofch 1qpe ' ' ' 0.15 0.15 0.15 0.15
" Core Shells and lnferfoce Plafes
¢ S o Mass‘of components, kg '
|
aOnfy one end duct was énlarged.
. Mass per unit volume of fuel region. :
e “Mass required for a critical system with the poison rods removed. §
; 59

 

     
 

 

 

 

 

~ :ANP PROJECT PROGRESS REPORT

Fig. 3.1. Partially Assembled 1sland Showing
Lower Half of Inconel Inner Core Shell and the
Upper Half of the Beryllium Reflector.

for the annular control-safety rod, which contains
a mixture of the oxides of the rare-earth elements.
The rod is magnetically supported and can be
positioned above the horizontal mid-plane for con-
trol. It is free to fall below the mid-plane for

 

reactor shutdown. The outside diameter of the
neutron-absorber section of the rod is 1.28 in.,
and the annulus is T/8 in. wide. The density of
the neutron-absorber compact is 6.5 g/cm?; its
principal constituents are Sm,0, (63.8 wt %) and
Gd,0, (26.3 wt %). The support and the driving
rod for the neutron source are coaxial with the
control rod. The beryllium blocks in the reflector
and the island are in an atmosphere of helium;
there is no sodium in the system.

The reactor is mounted above a reservoir for
the liquid fuel, which is a mixture of the fluorides
of sodium, zirconium, and enriched (93% U235)
uranium. The fuel is transferred to the reactor
by applying helium under pressure to the liquid
surface in the reservoir; the return to the reservoir
is by gravity. The temperature of the system has
been raised to 1350°F by electrical heaters located
external to the reflector and the fuel reservoir.
The completed assembly is shown in Fig. 3.3.

After an initial test of system operation and
leak tightness with an equimolar mixture of NaF
and ZrF,, successive increments of Na,UF, were
added until the reactor became critical, The fuel
concentration was then 6.30 wt % (2.87 mole %)
vranium, and the excess reactivity, determined
from the subsequent calibration of the control rod,
was about 0.13% Ak/k. A measurement of the
over-all temperature coefficient of reactivity be-
tween 1150 and 1350°F showed the value to be
negative and equal to 2 x 10=5 (AR/E)/°F. An
increase in the uranium concentration of the fuel
from 6.30 to 6.88 wt % resulted in an increase in
reactivity of 1.3% Ak/k. The control rod has a
value of 1.7% Ak/k when inserted to a point 4 in.
above the mid-plane. ' ‘

 

 
PERIOD ENDING SEPTEMBER 10, 1955

 

 

 

 

 

 

P,

 

: Fig. 3.2, Outer Core Shell and Partially Assembled Beryllium Reflector of High-Temperature Critical ‘;
Assembly.

 

   
 PHOTO 24440

 

 

 

 

3.

 

 

 

T

 

ircul&iing-F-uél R‘eddb"r_.:

Moderated Ci

ssem

al A

1C

 

Cr

h Temperature

3

blf of the Reflector-

 

 

 

 

 

 

           

 

 

i L el

 

 
T

    

   

B

 

3

   

 

*
4 -«
~
-

 

Part i

' MATERIALS RESEARCH |

 

 

 

 

 

 

 

  

   

   

 

 

 

 

  

\
.
&
) .
e
A
x :
=
.

 

   
 

o | .

 

 

 

 

L e o T T R T e Ty b s

 
 

.

[4]

RN M i . euSdSNEL G G
-
B

W

 

 

 

 

4, CHEMISTRY OF REACTOR MATERIALS

W. R. Grimes
Materials Chemistry Division

Phase equilibrium studies were made of the
LiF-ZrF,, UF,-ZrF,, NaF-LiF-ZrF,, NaF-LiF-
UF,, NaF-RbF-ZrF -UF,, KF-ZrF,, and NaF-KF-
ZrF, systems and of the BeF ,-containing systems
NGF-LIF BeF,, NaF-LiF-BeF,-UF,, NaF.BeF,-
UF,, K F- BeF and NaF-KF-BeF,. A revised
equalnbrlum d:agram for the LiF-ZrF, system is
presented., Additional data were obtained on the
solubility of UF, in BeF,-bearing compositions,

Additional work was done in investigating the
equilibrium reduction of FeF, by hydrogen in
NaZrF, the reduction of UF, by structural metals,
and the stability of chromium and iron fluorides in
molten fluorides. The experimental preparation of
NH,SnF; and of compounds of ZrF, with CrF,,
N|F2, and FeF, is described. The mveshga’r:ons
of the reaction of UF, with uranium in alkali fluo-
rides and the reaction of uranium metal with alkali
fluorides were continued. Studies of the reactions
between molten fluorides and metals included some
exploratory work on the equilibrium between so-
dium-potassium alloy and NaF-K F melts at 800°C,

Fuel purification and production research in-
cluded the investigation of methods for the re-
covery of confaminated fuel for re-use and the
conversion of ZrQ, to ZrF, in the present produc-
tion equipment, The studies of processing techs
niques for the purification of BeF,<bearing com-
positions were continued, Efforts to resolve the
problems in the procurement of adequate supplies

" of ZrF, are described, as well as improvements in
7 "bafchmg and dlspensmg operahons cmd locdmg

“and dralmng operaflons. e S e S
Pofenhal measwemem‘s were made v_y‘n‘ combm ,

      
 
  
  
 
 

metal IOI1 ln e molt

 

 

“":compounds Cin fluor{d‘_e sysf _3LiF.ZrF,

2LiF.ZeF, U 3LIF4ZeF,, 2KFBeF,, T KF- BeF
K F.NaF. ZrF , 3KF. 3NaF 2Z:F,, 3N0F 4ZrF ,, cmd
2NaF.LiF. QBeF

"Final results ob’rained for the electrical con-

 

  

“'products,_LrF ‘and lesz

o St GEEUF just above the decomposmon temperature,
Optical properhes ‘and x-ray

‘puflerns are presenfed for recently d:scovered'\_“”: EETRIES BLE T

: Jscosn‘y'measuremenfs*of a mixture

ductance and densityof all molten alkali chlorides,
bromides, and iodides have been obtained, and con-
clusions based on the measurements are summar-
ized, Measurements of freezing-point depression
are under way. Precise determinations of the self-
diffusion coefficients of sodium ion and of nitrate
ion in molten sodium nitrate have been completed.

PHASE EQUILIBRIUM STUDIES OF SYSTEMS
CONTAINING ZrF , AND/OR UF,

C. J. Barton F. F. Blankenship

Materials Chemistry Division

H. Insley, Consultant

The System LiF-Z:F

H. A. Friedman R. E. Thoma
R. E. Moore
Materials Chemistry Division
G. D, White, Metallurgy Division
D. L. Stockton, Merck & Co.

A revised phase diagram of the LiF-ZrF Sys-

tem! is presented in Fig. 4.1. The original,

tentative diagram? was based mainly on thermal
analysis data., The revised diagram is based on
thermal and differential thermal analysis datq,
high-temperature x-ray diffraction data, and petro-
graphic and x-ray diffraction examination of
quenched and slowly cooled samples. The Li ZrF,
crystals obtained by quenching and by slow cooling

‘were poorly formed, and it is possible that a rapid

inversion from a high-temperature crystal modifica-

tion occurred even during quenchmg. Highstemper-
“atu e:_'x-ray diffraction data on the compound indi-
s Egted that the high-temperature modification exists
-+ gbove 470°C. Below 470C, the pattern was identi-

- cal to the room-femperature pattern obtained from a

slowly cooled sample whlch contained the low-
temperaturé form and some pf the decomposmon
The inversion must

 

Quenches of the mixture with 31 mole % ZrF ,

 

. }R. E.V M'o.ore, ‘R. E. THgma, a'nd‘ D. L.. Stockton, ANP
Quar. Prog. Rep. June 10, 1955, ORNL-18%96, p 51.
2L. M. Bratcher, Y. S. Coleman, and C. J. Barton,

ANPéan:r. Prog. Rep. June 10, 1953, ORNL-1556, p 44,
Fig. 6.3.

65

T WY

 

 

 
 

 

i
Y
-;,4

 

 

ANP PROJECT PROGRESS REPORT

UNCLASSIFIED
ORNL-LR—DWG 9179 ~

 

f000

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

|
900 /,/
L~
. ~

_ ~

800 :

‘ AN 7

7~
7
. v
o 7
o .
=~ 700 - ~ ——ma-LigZrFy + LIQUID /
@
5 : i !
& LiF + LIaUID i / | Zrf, + LIQUID
& ¥
ul . . :
| CEL /\leszG + LIQUID i
[ LizZrgFg + LIQUID :
LiF +a-LigZrF, ¢ L'fZ”: |
C ‘ i T ,
‘ _""_____/:—"'—_"“'—' A Li,ZrFg + LizZr, Frg HiaZrafig T 2rhy
LiF + B-LiyZrF, B-LizZrfs; / | ] I I R A b
+
400 LiyZrF B
LiF 10 20 30 40 50 60 70 80 90 IeF,

ZrF, {mole %)

Fig. 4.1. Revised Equilibrium Diagram for the System LiF-ZrF .

established the primary phase as Li,ZrF , and the
liquidus temperature was found to be abouf 390°C.
Recently obtained thermal analysis data showed
that the Li, ZrF -Li,ZrF, eutectic melts at about
580°C. S:nce Li,ZrF ; is the primary phase of the
mleure con’rammg 3] mole % ZrF,, the Li,ZrF

i eufechc must contcun less than 31 mole % of ZrF4.

No L|3ZrF7 was found in quenched samples con-

rtammg 33.3 mole % ZrF, over the range 530 to
'619°C, and thus it was confirmed that Li o ZrF
_melts congruenfly. The melting point of Li ZrF
~was found, by thermal analysis, to be about 587°C

Thermal and differential thermal analysis data

_indicated that a eutectic is formed by the mixture
- with about 50 mole % ZrF,, The primary phase of

. the mixture with 48 mole % ZrF, was found, by

. 'quenchmg, to be leszé, the liquidus temperature
o was  522°C, and the solidus temperature was

f508°C The existence of L|22rF and L|3Zr4F19

%

in these samples was confirmed by x-ray diffrac-
tion examination, The Li,ZrF -Li,Zr F,, eutec-
tic appears to contain abouf 48 mole % ZrF, and
to melt at 508°C,

The experiments reported previously! seemed to
narrow down the composition of the unknown phase
in the mixture with more than 50 mole % ZrF
either Li,Zr,F,, (60 mole % ZrF,) or LisZr,Fy
(57.1 mole % ZrF4) Other experiments confirmed
this conclusion, but no decision between the two
formulas could be reached on the basis of the
quenching experiments. However, analogy to the
NoF-ZrF, system indicated LiyZr,F,, to be the
more probob[e, and it has been estcblished inother
experiments (described below) that Na,Zr F19 and
Li,Zr,F o belong to the same crystcl system and
form a complete solid-solution series,

Another phase (cubic) has sometimes appeared
in quenched samples in the region above 50 mole %

 

 

 

 

 
 

 

 

 

ZrF, ot temperatures at which liquid would be
expected to be present, A differential thermal
analysis break at 580°C was thought at first
to be associated with this phase. However,
quench growth and well-crysfalllzed ZrF, were
found in quenched samples containing 57] 60,
62.5, 66.7, and 75 mole % ZrF, above 520°C,
the incongruent melting point of Li Zr 4F 19+ There
appeared to be no difference in fhe somples taken
above and below 580°C, It has been decided that
the cubic material is either a metastable or o low-
temperature phase which sometimes precipitates
during quenching, The break at 580°C was prob-
ably due to nonequilibrium conditions.

The liquidus femperature of the mixture with
66 7 mole % ZrF, was found by quenching to be
rear 695°C. Thermai analysis data for the mix-
ture with 57.5 mole % ZrF , gave a liquidus tem-
perature of 618°C.

There is evidence that Li Zr,F,, decomposes
slowly into Li ZrF6 and ZrF4 upon cooling. A
relatively pure sample of LizZr 4F 19: prepared by
heating for 48 hr at 500°C, was reheated to 446°C
for several hours and quehched The sample then
gave a good x-ray pcflern for Li ZrF and ZrF ,,
although some Li Zr,F g remame-d The decom-
position temperature has not been definitely estab-

lished.

) The Sysfem UF -ZrF
R. P Mefcolf R E. Thomc

Motenals Chemlstry DIVISIOI"I

The system UF ZrF hcs been studied by using
the quenching technlque, and it has been analyzed

by using pefrographlc and x-rcy d:ffrachon _lmefh-"f'

g BRI

 

observed
(U, Zr)F ‘
cbout 765°C *

|n ’rhls system a

fh'

ed Vthy Et:et\\;io{us'ly.
fied mcongruenfly melting compound 3LiF. 4ZrF
whlch corresponds to the incongruently me!hng

<__:ompound 3NaF-4ZrF, in the NaF-ZrF, system, a

  

 

  
   
 
   
 
  
 
    
   
   
 
   

f pounds. “3NaF. UF , 2NaF. UF4, “and
= Since 2NaF. UF
- melting compound the LiF-2NaF-UF, join departs
from true uusn-bmary behawor ‘when the LiF con-

 

’rha’r\'are shown on" the diagram to have low mel'rmg
points contain too much UF  for them to be of use
as circulating reactor fuels.

PERIOD ENDING SEPTEMBER 10, 1955

series of cooling curves was carried out on mix-
tures of 3LiF4ZrF, and 3NaF.4ZrF,, A continu-
ous gradation of ophca[ properties from 3NaF.4ZrF

(n, = 1.420, ny = 1.432; biaxial positive; 2V = 30
deg) to 3LiF-4ZrF, (n, = 1.463, ny = 1.473; biaxial
positive; 2V = 20 to 30 deg) gave strong indica-
tion of a complete solid-solution series, The
primary phase along the join was ZrF,, and the
liquidus temperatures increased steadily from
550°C for 3NaF.4ZrF, to 618°C for 3LiF.4ZrF ,,
"The solidus temperatures, which corresponded 'ro
‘the incongruent melting temperatures of the solid-
solution series, exhibited a minimum at 475°C for
the mixture with about 25 mole % LiF. The
eutectic in the compatibility triangle formed by

6NaF.7ZrF ,, 2LiF.Z¢F ,, and the 3(Na,Li)F. 4ZrF4'

solid solution was found to be NaF-LiF- ZrF
(25-29-46 mole %) and to melt at 445°C,

An intensive study of the NaF-LiF-ZrF,.-UF,
quaternary system was interrupted when it was
learned that no improvement in viscosity resulted
when the mixtures contained relatively high LiF
and low ZrF, contents, At high zirconium con-
centrations, there was a marked reduction in melt-
ing point which was, however, accompanied by
unusually high ZrF , vapor pressures,

‘The System NaF-LiF-U F,

H. E. Friedman R. E. Thoma
Materials Chemlsfry Division
_ R. E. Cleary
Pratt & Whitney Aircraft
~T. N. McVay |
“ Consultant

 
 
   

aE U F 6UF4.

|s just barely an mcongruently

, Saddled

67

 

he compound LiF forms nearly compfe're quosn-: |

.. binary " systenis wn‘h each of the F-UF, com-

 

 

 

 

 

e

 
 

 

 

 

ANP PR OIJ_EICT: PROGRESS REPORT

The |om between 6NaF.7UF, and 6LiF. 7UF |

~ -appéars to form a side of a compatibility ’rnangle
- .and so does the join between 6NaF.7UF, and
- 3LiF. UF These two joins, along with the quasi
f'",fblnarles descrlbed in the foregoing paragraph,
. +-define the important compatibility triangles in the
S "':fportlon of the ternary mixture that contains 47 mole
7%, or less, UF In the region of UF, concentra-
"':*,rk",-’{tlons above 50 mole %, there is cn‘ least one
’{'temary compound the high UF 4-confent portion of
.- the diagram is -not yet undersfood but the first
';?{-rlnd:cahons are “that UF, is a compatible phase
-_""over d surprlsmg!y small area.
A.s a result of these studies it ap;flears that some
cdmpl:cc’nons prevsously unrecognized,3 appear in
the LiF-UF, binary system. In addition to the
L‘iF-‘UF’: and LiF-2UF, compounds previously
'E-observed there is another compound, 7LiF.6UF ,,
sling curves at composmons near the single
_,trc' (28 mole % UF 4)r the temperature may
" rise and fall 510 15°C fwo or three times after the
eutectic halt is under way. This abnormal be-
havior seems to be due to an exothermic decom-
position of fil'_iF"-'Ul:4 at about the eutectic tem-
perature, along with a tendency for the melt to
undercool and then rapidly decompose.

 

 
 
  
   
 
  
 
   

The System NaF-RbF-ZrF ,-UF,
H. A. Friedman R. E. Thoma

- Materials Chemistry Division

Preliminary studies of some systems containing
RbF and ZrF, were made in an attempt to obtain
materials with low viscosities for possible use as
reactor fuels, Previous studies? of the RbF-ZrF,
binary system had shown low melting points only
for mixtures with moderately high (greater than
40 mo!e %) ZrF concentrations; 3RbF. ZrF , melts
at 900°C. Consequenfly, low-melting mlxtures

: confcumng RbF and UF, with 20 to 40 mole %

: ZrF4 necessarily confam a fourth component.

The RbF available, at present, contains about
20 mole % of KF, and therefore all studies re-
_fr'r:béfi'eddhe're‘wéfe made with RbF containing that
- amount of KF. It was anticipated that the KF

- would subshfute isomorphically for RbF and that
" the phase dlagrams would not appear to involve
"rhe KF componem‘. Unforfunately, thls proved not

3:.R. E Moore et al., ANP Quar. Prog. Rep. Dec. 10,

1950 ORNL-919, p 245.

S L M. Bratcher, R. E. Traber, Jr., and C. J. Barton,
' ANP Quar. Prog. Rep June 10, 1952, ORNL-1294, p 92.

 

     
 
 
  

 

" to be the case, For example,BRbF.ZrF4formed

preferentially, and, if additional KF were present,
an excess of KF remained. In low-melting parts of
the NaF«(Rb,K)F-ZrF , system, unexpected phases
made their appearance, and therefore quasi binaries
could not be established. A join at 47 mole %
ZtF , appeared to pass through the phase fields of
two femary compounds and to give the lowest
melting points at the high RbF-content end, This
behavior, coupled with lack of information regard-
ing the similar NaF-KF-ZrF , system, caused the
postponement of a systematic study of the NaF-
RbF-ZrF, diagram, pending results from viscosity
measurements on an RbF-ZrF .UF, fuel mixture
and the procurement of better RbF,

In an effort to find a suitable fuel mixture for
viscosity measurements, a constant 4 mole % of

UF, was added to the RbF-ZrF, binary over a
range of composmons from 55 to 45 mole % RbF,

All these mixtures appeared to melt between 410

and 420°C and to give no detectable heat effects
above this range. This result led to the suspicion,
later confirmed, that small amounts of precipitates
with a very low temperature coefficient of solu-
bility were escaping detection, From the middle
of this range of apparently low-melting mixtures,
the composition RbF-ZrF ,.UF, (48-48-4 mole %)
was chosen as a preparation ?or viscosity meas-
urements, = The purified mixture was cooled to
450°C to precipitate any solidified crystals before
transfer; however, the transfer was carried out
through heated lines and a heated filter, so the
flowing melt was heated to 575°C. Crystals of
(Zr,U)F4 solid solution were found in the heel re-
maining in the purification vessel but not in the
filtered product. Hence, (Zr,U)F‘4 was presumed
to be the primary phase; it is possible that the
(Zr,U)F ; precipitated incompletely in the 15min
equilibration period at 450°C, since small parti-
cles were subsequently observed in the melted
product at temperatures above 500°C. However,
petrographic examination of the solidified melis
in which particles had been observed while they
were molten showed the presence of significant
amounts of oxyfluoride,

Viscosity measurements below 600°C were not
satisfactory because of these particles, but the
values obtained above 600°C were comparable with
those obtained, to date, for satisfactory fuel mix-
tures (see Sec. 7, *"Heat Transfer and Physicdl

Properties'). All the work described here was

 

 

-
. ! ¥ ]
T O TR e T

-l )

PR Y

Liiad oo ”h“?v

 

ap T

 

g

 

 

 
 

 

 

carried out with RbF containing about 20 mole %
KF.

The System KF-ZtF,
H: A, Friedman R, E. Thoma
Materials Chemistry Division

T. N. McVay

Consultant

The system KF-ZrF, is being studied in an at-
tempt to obtain background information regarding
the rather similar RbF-ZrF, system and to check
and refine techniques for handling these hygro-
scopic materials. Cooling curves obtained for ten
stirred KF-ZrF, melts with compositions in the
range 33.3 to 59 mole % ZrF, were in general
agreement with the phase diagram previously pub-
lished.> Thermal breaks in the range 45 to 50
mole % ZrF, indicate that there are two forms of
KZtFg, one of which is metastable.

| The System NaF-KF-ZrF,
H. A, Friedman R. E. Thoma

Ma’rerlcls Chemistry Division

F’reltmlncry studies indicate that the NaF-KF-
ZrF , system is much more complex than either the
NaF-ZrF -UF, system or the NaF-LiF-UF, sys-
tem., Neither of these latter systems shows ter-
nary compounds; in contrast, the NaF-KF-ZrF
system shows at least five such compounds. One
compound has the composition NaF:KF.ZrF,, and
another has the composition 3NaF.3KF.2ZrF .
Although the latter compound forms a compatibility
~ triangle with NaF and K, ZrF,, it cannot form a
quasi binary with K ZrF,, NasZrF,, or NaF, be-

cause it apparem‘ly melts mcongruenfiy at about

- 755°C to K ZrF and liquid. Cooling curves on
- the Na,Z:F -K ZrF join show a minimum in the
' !iquidus femperm‘ures' at about 800°C for the com-

. position with about 15 mole % KF; the liquidus

' temperatures then rise gradually to about 918°C
- for-the compound K ZrF and about 850°C for the
'VCompound Na ZrF7. Smce a slowly cooled melt of
-~ a compesition mldwcy between 3NaF.3KF.2ZrF,

* “and NaF- KF. ZrF4 consists of only these phqses,
it can be concluded that this join is a common
base of two, as yet undetermined, compatibility
triangles,

 

 

SL M. Bratcher, R. E. Traber, Jr., and C. J. Barion,
ANP Quar, Prog. Rep, June 10, 1952, ORNL-1294, p 91,
Fig. 43.

PERIOD ENDING SEPTEMBER 10, 1955

Eight melts were made along the 33.3 mole %
ZrF, join of compositions containing from 15 to
50 mole % KF and NaF as the remainder, Although
all these compositions contained the compound
NaF:KF.ZrF, as one of the phases in the com-
pletely crystallized preparation, liquidus-tempera«
ture relations indicate that the compound melts
incongruently and that therefore there is no quasi
binary along this join,

Melting relations and phase compositions of six
preparations along the 47 mole % ZrF, join indi-
cate that the mixture with 47 molé % ZrE, also is
not a quasi binary, because most of the composi-
tions consist of three solid phases, including a
new ternary compound of undetermined composition
that is probably near NaF.2KF.3ZrF,, The mini-
mum liquidus temperature along this join is about
420°C for a mixture with about 35 mole % KF, A
search in this general area has indicated that
there is a eutectic, with the approximate composi-
tion 38 mole % KF, 21 mole % NaF, 41 mole %
ZrF ,, which melts at about 400°C,

PHASE EQUILIBRIUM STUDIES OF SYSTEMS
CONTAINING BeF2
C. J. Barton and F. F, Blankenship
L. M. Bratcher R. J. Sheil
B. H. Clampitt R. E. Thoma

Materials Chemistry Division
R. E. Cleary, Pratt & Whitney Aircraft
T. N. McVay, Consultant
The System NaF-LiF-BeF,

Thermal analysis data were obtained with a num-
ber of compositions within a wedge-shaped areaq in
the temary diagram having the LiF-Na,BeF, eu-

~tectic (16 mole % LiF) at the apex and Na,BeF,

and NaF-BeF, (70-30 mole %) at the other corners.
The cva:lub[e data indicate that acceptably low
melting points can be obtained with low LiF con-
ceniration by moving into the ternary system along
the drainage path leading from the NaF-Na,BeF,
eufectic (approximately 31 mole % BeF,; melting
point, 570°C) toward the LiF-NGzBeF4 eutectic,
Further, viscosity data for BeF -bearing mixtures
indicate that, in order to obtain a melt with a
kinematic viscosity as low as that of the ARE-type
fuel (2.82 centistokes at 600°C and 1.34 centi-
stokes at 800°C),% the BeF, content of the melt

 

6s, 1. Cohen, ANP Quar. Prog. Rep. June 10, 1955,
ORNL-1896, p 157,

69

 
 

ANP PROJE’CT PROGRESS kEPORT

must not be greater than 30 mole %, to assure com-
plexing of BeF, as BeF,=~ ions in the melt, If
an NaF-LiF-BeF, mixture is desired, it also ap-
pears that the LiF content will probably have to
be less than 10 mole %. One composition that
‘meets these requirements is the mixture NaF-LiF-
BeF, (63.,5-7.5-29 mole %), which has a melting
point of about 525°C, It is expected that the
viscosity of this mixture will be determined in the
near future,

The low-melting ternary mixture NaF-LiF- BeF
(27-35-38 mole %), reported previously’ on thev
“basis of cooling<curve data to melt at 325 + 10°C,
was found to melt at 339 + 5°C by the use of dif-
ferential thermcl analyms techniques., The former
value was low because of undercooling. This mix-
ture has been reported8 to have kinematic viscosi-
ties that are too high — 3.85 centistokes at 600°C
and 1.47 centistokes at 800°C — to be considered
competitive with those of the ARE fuel but low
enough to be of interest for a reactor design re-
quiring @ substantially lower melting point fuel
than _can be obtamed with known ZrF -base com-
posmons.

quU|dus temperatures were determined for sev-
eral compositions by differential thermal analysis
of samples cooled rapidly from 675°C to room tem-

‘pefature in order to avoid the segregation of the

crystalline components that apparently occurs on
~slow cooling. The composition NaF-LiF-BeF,

(50-‘[63’-33‘/ mole %) showed a liquidus tempera-
ture of 478°C and a solidus temperature, deter-
‘mined from hecmng curves, of 329°C, The mixture

'NGF-LlF BeF (64-5-31 mole %) showed a melting

point of 545°C and a freezing point of 555°C. An
unpurified melt of this composition froze at 535°C,
Most mixtures in this system have exhibited much
lower freezmg points than melting points because
“of the prevalence of undercooling, but the 64-5-31
mole % mixture appears fo be exceptional in this
reSpect.

. The equimolar composition NaF-LiF-BeF,, pre-
"wously investigated by Jahn,? was studied by
“using quenching and differential thermal analysis
techniques. The latter technique gave a value of
'V'A"434°_C fbr the hqmdus temperature, as compared

 

7C J. Barton et al., ANP Quar. Prog. Rep. June 10
-1955 ORNL-1896, p 57.

o Prlvate communication from J. F. Eichelberger,
'~ Mound Laboratory.

_9Wf Jahn, Z, anorg. u. aligem, Chem, 276, 113 (1954).

70

'NG LiBe

with the approximate value of 460°C reported by

John, A crystalline phase having the x-ray pattern
described by Jahn for the NaLiBeF, compound pre-
pared from aqueous solution was observed in
samples of this composition quenched at tempera-
tures ranging from 406 to 562°C. At the lower

‘temperature, LiF and Na,LiBe,F, were also pres-

ent, while in samples quenched at temperatures
above the liquidus temperature, quench growth was
observed petrographically, Jahn's failure to ob-
tain NuLiBeF4 from melted samples by quenching
at temperatures below 550°C, which he ascribed to
formation of Be2F73" ions in the liquid, was more
likely due to poor quenching technique. Solid-
phase decomposmon of NaL.iBeF, to give LiF and
at 225°C, as repon‘ed by Jahn, was
conflrmeg by the quenchmg experiments,

The System NaF-LiF-BeF ,-UF,

Investigations of the NaF-LiF-BeF -UF system
were confined to liquidus remperature defermma-
tions with mixtures containing 2.5 mole % UF,,
which gives cpproxumately the same concem‘rohon
of UF, (in g/em3) at 800°C as that of NaF-ZrF,.
UF, (50-46-4 mole %). Various techniques, m-
cludmg thermal analysis, differential thermal anal-
ysis, filtration, and visual observation, were used.
The results are summarized in Table 4.1,

The first mixture listed in Table 4.1 was pre-
viously reported to have a liquidus temperature of
465 t 5°C, as determined by visual observation,

.The filtration experiment demonstrated that the
small, high-temperature thermal effects observed

on cooling curves were spurious,

Analyses of the filtrates obtained with the 64.5-
31 mole % mixture showed an increasing concentra-
tion of uranium from 540 to 520°C, which indicated
that a nonuranium phase was precipitating in this
temperature range. The filtration of the 27-35-38
mole % mixture showed that 9.65 wt % uranium,
equivalent to about 1.8 mole % UF 4+ Was in solu-

tion at 335°C,

The System NaF-BeF ;~UF,

The reported low v:scosnfy of the NaF-BeF,
(70-30 mole %) composition!® prompted interest |n
determining liquidus temperatures of mixtures

formed by adding 2.5 mole % UF, to NaF-BeF,

 

IOPrivafe communication from J. F. Eichelberger,
Mound L aboratory.

 

 
 

ne

 

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 4.1. LIQUIDUS TEMPERATURE DETERMINATIONS WITH NaF-LiF-BeF »~(2.5 MOLE % UF ) MIXTURES

 

Solvent Composition

 

Liquidus Temperature (°C)

 

 

(mo-]e %) T .
- ‘ echnique
NaF LiF "BeF. NoF-LiF-BeF,-UF , Mixture Solvent
"5 16 28 Filtration <491 480
64 5 31 " Filtration <548, >540 540
27 35 38  Visual 355 | 339
S Filtration >355 339
63.5 .5 31,_5 .7 fDane-renhal thermal analysns - 537 545 to 555

 

   

 

mixtures °,f. dbpm*iffi?}e!y'V'fHé’_?O-’BO mole,-%:.éom-
‘posifion. Differential thermal analysis data ob-

tained with rapidly cooled samples of the mixture

- NaF-BeF ,-UF, (68.25-29.25-2.5 mole %) indicated

a probable liquidus temperature of 579 + 5°C,
Filtration data obtained wn‘h the mixture NoF-

_Ber-UF4 (66 8-30.7-2.5 ‘mole %) showed the

liquidus temperature fo be below 562°C.  The
highest thermal effect observed on cooling curves
with the lafter composition was ot about 555°C.

There appears to be little hope of obtalnmg an

acceptable hqmdus fempera’rure with a ternary
mixture of this fype confammg 25 mo[e % UF,
cnd 30 mole %, or Iess, BeF

The System KF-BeF

Thermcl 0na|ys:s ‘data recem‘ly ob'ramed wn‘h_'_
punfled mixtures in the KF-BeF2 system showed
sllghfly h|gher melhng pomi's than those obtmned{ "

      
 

,,,,,

P

. mcongruem‘ly at 390°C and g;vesfl‘K

BeF £

 

Thermal anulys:s data on. abouf:"40 unpurified

" melis in the NaF-KF-BeF, system were obtained
- prewously.H The mcomplete inve .tlgcmon showed

 

rather high liquidus temperatures, as compared

with those of the NaF-LiF-BeF, system, and no
promising fuel carrier compositions were formu-
lated as a result of these studies., Recent interest
in obtaining a BeF -buse fuel cartrier with low
kinematic viscosity prompted a reinvestigation of
the l\]c::[:--KF--BeF:2 system. Thermal analysis data
obtained with temary compositions prepared from
purified binary mixtures have served to confirm the
earlier indications of high liquidus temperatures in
this system, particularly for the mixtures con-
taining 33 mole %, or less, BeF,. The mixtures
containing 33 mole %, or [ess, BeF are considered
to be the most promising maxtures from the vis-
cosity standpoint, Preliminary results of petro-
graphic and x-ray diffraction examinations of a

number of slowly cooled melts show that composi-

‘tions on the NaF-K. ,BeF, join contain only these
two componem‘s. Thermal analysis data for this
system, shown in Fig. 4.2, indicate that it is a
_simple eutectic " systém, with the eutectic con-

- L R e

g a ouf{»_’?g‘_AS mole 7 NoF cmd havmg a melt-

    
 
 

 
  
    

 

  

 

J P. Blakely, L. M. Bratcher, and C. J. Barton,
ANP Qua:r. Prog. Rep. Dec, 10, 1951, ORNL-1170, p 87.

71

 

compounds are”’NaKBeF ancf Na K(BeF4)2. An e
I f th_e rpha ehevedr_ 'rqv be N

K BeF “and the fernarym

 
 

 

 

 

 

ANP PROJECT PROGRESS REPORT

“c UUNCLASSIFED ¢
ORNL-LR-DWG 9480

 

1000 o=

 

900 : ™~

 

700

800

 

 

O
e
- L
o
3
|._
<
e
Sw
S
=
Lo
'_

 

 

 

 

 

 

 

 

 

© 500

 

 

 

 

 

 

 

CNeF T 40 20 30 40

50 60 70 80 790 KyBeR,

K,BeF, (mole %)

Fig. 4.2. Thermal Data for the System NaF-K BeF,.

Solubility of UF; in BeF,-Bearing Compositions

Data on the solublllty of UF; in a number of
BeF .bearing compositions were reported earlier.12
The composition(in mole %) designated 69 NaF-31
BeF, in the previous work was actually 69 LiF-31
BeF,. All the earlier data were obtained with
mckel filters.1® In view of the greater stability of
UF, dissolved in LiF-NaF in copper containers
(discussed below in this section under the heading
““Reaction of UF, with Uranium in Alkali Fluo-

rides’’), it seemed desirable to see whether more

consistent data could be obtained in copper ap-

paratus equipped with a bronze filter medium, The
data obtcuned are shown in Table 4.2, together
‘with some addmonol dm‘u obtained with LiF-BeF,
mixtures in ‘nickel apparatus. The uranium addt-
~ tions were made erh UF, and uranium metal or by
'addmg large excesses of uranium metal to UF,-
beormg melts,
It is difficult to draw definite conclusions from
such scah‘ered data, but it is apparent that alloying
o of uramum wnh the nlckel filters and containers

 

o ]2L. M Bratcher et al, ANP Quar. Prog. Rep. June
10 1955, ORNL-1896, Table 4. 2, p 38,

R L M. Bratcher et al, ANP Quar. Prog Rep. March
10 1955. ORNL 1864 p 51

 

was not the sole cause of the poor reproducibility
of the UF .solubility data obtained in nickel ap-
paratus. The large amount of tetravalent uranium
found in all the filtrates, regardiess of valence
form of the uranium added to the soivent, could be
due to oxidizing impurities in the melt, dispropor-
tionation of UF,, or a reaction between UF; and
the alkali fluoricj’e or beryllium fluoride in the melt,
Further experimentation will be required to deter-
mine the controlling factor in the production of

UF, in BeF,-bearing compositions.
CHEMICAL REACTIONS IN MOLTEN SALTS
L. G. Overholser G. M. Watson
Materials Chemistry Division .
Equilibrium Reduction of FeF, by H, in NaZ¢F,
C. M. Blood

Materials Chemistry Division

An apparent equilibrium ‘;ldln"stfir_i'f (molefrochon )
of dissolved species used as activity) of 5.7 at
800°C was previously re.-pm'\‘ec!14 for ’rhe reachon

FeF, + H, ;—.:_‘ Fe + 2HF

 

Y. M. Blood, ANP Quar, Prog. Rep. June 10, 1955,
ORNL-189, p 6.

 

il

-y

 

T T

 

 

 

 
PERIOD ENDING SEPTEMBER 10, 1955

TABLE 4.2, SOLUBILITY OF UF, IN BeF,-BEARING COMPOSITIONS

 

Temperature

Container

Analysis of Filtrate
Uranium

 

 

Solvent Composition o) Haterial Added (wt %)
us? Total U
LiF-BeF, (69-31 mole %) 700 Nickel UF5 + U 2,79 4.80
700 Nickel UF, +U 3.38 5.66
a | 800 Nickel UF, +U 8.35 14.1
i 800 Nickel UF, +U 8.28 12.7
600 Copper UF, +uU 0.46 3.12
. : 700 Copper UF, + U 3.09 5.34
800 Copper UF, +U 5.02 12.7
NaF-BeF , (70-30 mole %) 600 Copper UF, +U 0.40 2.50
600 Copper UF, +U 0.19 7.05
700 Copper UF3 + U 2,75 5,24
‘ 700 Copper UF, + U 4.82 10,1
800 Copper UF, +U 10,1 14.7
’ 800 Copper UF, +U 12.2 19,1
° when the materials were contained in a mild steel  of iron. From all apparent indications, the ap-

system. This value was obtained by passing hy-
- - drogen at different gas flow rates through the salt
mixture containing FeF, and measuring the cor-
responding partial pressure of HF in the effluent,
An approach to the equilibrium partial pressure of
HF was made by extrapolation to the slowest flow
rate, As has been stated before, !5 the extrapola-
. tion does not c:ppear to be saflsfaci’ory.

 

H2 through The melt con

     
  

 

 

 

| '5c. M. Blood ANP Quar.'
ORNL-1864, p 57.

* ' 16C, M. Blood and G. M. Watson, ANP Quar, Prog
Rep. Sept. 10, 1954, ORNL-1771, p 66.

 

i order to avoud fhls exfrapolatlon, measure-
ments made bY Us'ng an ethbruhon fecthue:m
: - "prewously descrlbed 16 have been continved, The
, **“method cons:sfs in bubbhng a mixture of HF and
‘ | nmg FeF2 in contactv'::"‘j"”"'“:"'
ntil the HF <.:on't:em‘rc|1hons»\oW}c the '"'-.,,.,

p o S corresponchng ’ro 6490 ppml

proach to equilibrium in this instance was better
than that for any previous measurement obtained by
this technique. Unfortunately, the existence of
equilibrium cannot be demonstrated with certainty,
since the present equipment cannot provide a gase-
ous influent of constant composition for long
periods of time. An alternate method of formation
of H,-HF mixtures of constant composition is

 presently being studied.

Reduction of UF, by Structural Metals

J. D. Redman C. F. Weaver
Materlals Chemlsfry DlVISIon

The techmques used for measurmg ‘the reduchon

| “of UF by Ct° or Fe® in molten fluorides were de-

scrlbed in earlier reporfs.17 Equilibrium data for

<these reactions when the Nc:tFaZrF4 compound, 18

 

]7L G. Overholser, J D Redmon, and C. F Weaver.
ANP Quar. Prog. Rep. March 10, 1954, ORNL- ]692

~-.p 56; J. D. Redman and C. F. Weaver, ANP Quar. Prog.
Rep. June 10, 1954, ORNL-1729, p 50.

18), D. Redman and C .F. Weaver, ANP Quar. Prog.
Rep. June 10, 1954, ORNL- 1729, p 50; Sept. 10, 1954,
ORNL-1771, p 60.

73

 

 

 
b i i i e, i s o il

bl sy

IQJ'

 

 

ANP PROJECT PROGRESS REPORT

the NaF-LiFKF eutectic,!? or The NQF-ZrF
(53-47 mole %) mixture?® was used as a rectchon
medium have also been presented. These studies
have been continued with experiments in which

‘NaF- -LiF<ZrF, (22-55-23 mole %) was used as the
“reaction medlum. A study of the reduction of UF,

by vanadium metal with the NaF-LiF-KF eutectic

as the solvent has been started, but no analytical
‘data are yet available. Similar studies, planned

for the near future, with metallic niobium, tantalum,
and tungsten will be conducted in an aflempt to

. "demonstrate whether or not these metals are com-
“ patible  with UF dissolved in the NaF-LiF-KF
“eutectic at 800°C
The results of some studies of the reduction of

UF, by Cr° at 600 and 800°C with NaF-LiF- ZrF

(22-55-23 mole %)} as the reaction medium are g:ven
in chle 4 3 In a” ‘these experiments 2 g of hy-

 

D. Redman and C. F. Weaver, ANP Quar. Prog.

" Rep. March 10, 1955, ORNL-1864, p 58.

205, D. Redman and C. F. Weaver, ANP Quar. Prog.
Rep. June 10, 1955, ORNL-1896, p 60.

TA_BLE 4,3, EQUILIBRIUM DATA FOR THE
REACTION OF UF, WITH Cr® IN MOLTEN
NaF.LiF.ZrF , (22.55.23 MOLE %)

AT 600 AND 800°C

 

Conditions of Anolysi‘s of Filtrate

 

Equilibration

 

Total Total  Total
Temperature Time U Cr* Ni
(°c) (hr) (wt %) (ppm) (ppm)
600 3 8.6 580 40
3 8.5 540 30
3 8.8 430 15
3 8.9 680 30
5 8.9 650 15
5 8.3 460 35
5 8.6 450 115
5 8.6 560 40
800 3 9.4 720 15
3 9.1 690 20
5 8.5 810 15
5

8.7 800 15

 

 

. *ijc‘!'fi‘k' of 250 ppm of éhrc:lmiur:; at SOOO C.

44

~ drogen-fired chromium was reacted with 114 wt %

UF, (2.5 mole %) dissolved in approximately 30 g

_of fhe NaF-LrF-ZrF4 mixture contained in nickel.

The averages for the chromium values given in
Table 4.3 — 570 and 750 ppm at 600 and 800°C,
respectively — are the lowest found for any of the
reaction media studied. For comparison, chromium

values at 600 and 800°C in NaF ZrF were 2250

“and 2550 ppm; in NaF-ZrF (53-47 mole %), 1750

and 2150 ppm; in NaF.LiF- KF (11.5-46.5-42 mole
%), 1100 and 2700 ppm. Although the UF, con-
centration was somewhat lower (2, 5 compcred with
4.0 mole %) in the recent experiments than in the
previous experiments, this variable can account for
only a small part of the difference found in chro-
mium content., |t seems more probable that the
most important factor, for the ZrF -containing
mixtures, is that with a decrease in ZrF, con-
centration there is also a decrease in the activity
of the UF, because of increased complexing by
the alkali fluorides. The results cannot be com-
pared with those obtained in the NaF-LiF-KF
eutectic, because virtually all the chromium is
present as the trivalent form in the eutectic, As
reported below, a large fraction of the chromium is
divalent in the NaF-LiF-Z¢F , mixture,

Data for the reaction of UF ; with Fe® in nickel
containers at 600 and 800°C with NaF-LiF-Z:F ,
(22-55.23 mole %) as the reaction medium are gwen
in Table 4.4, In all these runs, 2 g of hydrogen-
fired iron wire and approximately 30 g of the
solvent containing 11.4 wt % (2.5 mole %) UF
were employed,

The equilibrium iron values reported in Table 4.4
fall in the same general range as those found in
the other reaction media. The one dlfference to be
noted is that the iron values given in Table 4.4 are

slightly higher at 800 than at 600°C, whereas in

the NaF-ZrF ; mixtures and the NaF.LiF-KF eutec-
tic slightly hlgher values were observed at 600
than at 800°C

Stability of Chromium and Iron Fluondes in
Various Molten Fluorldes B o
J. D. Redman C. F. Wedver
Materials Chemistry Division
Previous experiments indicated that CrF, was

not stable in the eutectic mixture NaF-LiF-KF
(11.5-46.5-42 mole %) but that when the soluflon

~ was equilibrated at 600 or 800°C 'n nickel ap- L
paratus the product was mainly Cr3*, Additional

 

 

-y

T I

al

o

-

-

 

gt

 

 

 

 
 

! PERIOD ENDING SEPTEMBER 10, 1955

 

 

 

 

 

 

 

 

 

  

     

 

 

is ' TABLE 4.4, EQUILIBRIUMDATA FOR THE =~ data that indicate the stability of CrF, in this
3" - REACTION OF UF4 WITH Fe® IN MOLTEN solvent in nickel and Inconel equipment are pre-
i . ' NaF.LiF-ZrF , (22.55-23 MOLE %) "~ sented in Table 4.5. The data suggest that Cr3*
- AT 60°_AND 800°C under these conditions is completely stable to
. e e — ' reduction by Ni°® but that Cr® (and possibly Fe9)
Conditions of Analy5|s of Fllfra're' from the Inconel reacts with the CrF,, Whether
Equilibration equilibrium has been achieved in the latter case is
: Total Tgtcl Tofoi
Temperature  Time U Fe* Ni not known. Quantitatively, however, the solubility
(°C) (hr) (wt%)  {ppm)  (ppm) of CrF3 is considerably less than was indicated in
| ) e S . previous studies for which shorter equilibration
g ce 600 3 7.9 460 30  times were used. If does not seem likely that the
= : 9. 8.4 420 35 temperat'ure meczsuremenfs are sufficiently in error
S . A to explain the discrepancy,
. 5 . 86 375 35 a R
S5 g5 440 30 Data obtained by adding CrF, or CrF; to an
o e NaF-ZrF, (53-47 mole %) mixture, equilibrating in
800 3 89 700 30 nickel apparatus for 5 hr at 600 or 800°C, filtering,
g 54 650 70 ‘and analyzing the product chemically are presented
L e ', in Table 4.6. Similar values for CrF2 or CrF,
‘ S 86660 25 additions to an NaF-LiF-ZrF, (22-55-23 mole %)
: 5 8.6 540 25 mixture are presented in Table 4.7,
»9' 3 5 ,- 83 7 ‘:-6‘30,‘ :45:'_.' It is evident from these tables that GrF, is not
[ 5 .85 . 580 55 entirely stable in either of these solvents, al-
: — e e ‘ though, when CtF, |s the added constituent, the
- *Blank of 100 ppm of Fe at 800°C.. 7 concentration of Cr is generally higher than the
] TABLE 4.5. ‘S'TABIL'ITY OF »CrFsrlN'MOL‘TEN NaoF-LiF-KF (11,5-46,5-42 MOLE %) AT 600 AND 800°C
T e e Analysis of Filtrate
Conditions of Equilibration - CrFyq —
—_— : - : Total Total
. N Added 2+
Temperature Time (wt % Co) Cr Cr Ni
Or ) ) t r
_(.C) (hr) v - (ppm) (wt %) (ppm)
600 12 48 5 0.24 110
| 12 4.8 5 0.24 65
: 12 4.8 10 0.23 25
12 4.8 30 0.35 115
80 2 48 5 29 65
Sl 48 s 29 85
48 0 3o
600*' . 35 o s
'“'jwoo* 40 30
e i!r:i;:o.r;.e.il contqmerused,aliofherruns /i‘n .n.f.ckél confomers. )

75

 

 

T

 
 

 

 

'ANP PROJECT PROGRESS REPORT

TABLE 4.6, STABILITY OF CHROMIUM FLUORIDES IN NaF-Z:F , (5347 MOLE %) AT 600 AND 800°C

 

Analysis of Filtrate

 

 

7 Equullbrahon' A CHromiQm » :
- Tempercn‘ure" ' Fluoride Added o2t Total Total
. P : o Cr Ni
( C) (wt % Cr) |
UL : (_"‘” %) (wt %) (ppm)
600 o 1.0 cr2¥ 0.46 80
SR 1.0 a2t 0.40 0.60 50
. 800 . 0.88 cr2* 0.58 0.75 40
o ossa?t 0.30 0.86 65
120 o2t 9.1 1.7 305
12.0 cr2* 9.2 11.6 | 270
600 0.1 c3* 0.08 0.08 70
. S st 0.05 0.08 . 70
S so0 4.0 cr3* 1.5 3.6 280
R a0 c3t 1.7 3.6 285

 

TABLE 4.7, STABILITY OF CHROMIUM FLUORIDE S IN NaF-LiF-ZrF , (22.55.23 MOLE %) AT 600 AND 800°C

 

Analysis of Filtrate

 

 

E qui libra?.io'n | Chromium

Temperature Fluoride Added o3 Total Total

(°C) {(wt % Cr) Cr - Ni
(wt %) (wt %) (ppm)

600 1.0 2t 0.61 0.78 1

1.0 Cr2* 0.59 0.97 1

2.0 Cr2* 1.01 1.91 155

2.0 cr2? 0.99 2.15 30

800 1.0 cr2* 0.87 15

1.0 cr2¥ 0.65 1 0:95 | 1
5.0 Cr2t 3.3 45 40
5.0 Cr2* 3.1 4.6 25
1.0 ce3? 0.37 1.05 200
4.0 ¢t |

1.2 4.2 160

oy

 

concentration of Cr3t, When Cr®* is added, ap-
preciable reduction to the divalent state is ob-
served, even though no Ni?* appears in the filtrate
and no other reducing agent is supposed to be

"'presen’r. Further experimentation will be required
_before conclusions can be reached regarding the
behavuor of elfher CrF, or CrF, in these solvents.

76

Data obtained by adding FeF, or Fe‘F.3 ’ro an

NaF-ZrF, (53-47 mole %) mleure, equ;hbrahng in
nickel opparctus for 5 hr at 600 or 800°C, filtering,
and analyzing the product chemically are presented
in Table 4,8, Similar values for FeF, and FeF

additions to an NaF-LiF-Z:F, (22-55-23 mole %)

mixture are presented in Table 4.9.

i

 

-y ‘

 

 

 

=

 

 
 

-y

PERIOD ENDING SEPTEMBER 10, 1955

 

 

 

 

 

 

 

    
   

 

%- TABLE 4.8, STABILITY OF FeF, AND FeF, IN NoF.ZrF , (53-47 MOLE %) AT 600 AND 800°C
3' . . : Lo Analysis of Filtrate
{ Equilibration Iron
’L Temperature " Fluoride Added Fe2+ - Total Tof-al
(°C) (wt % Fe) (wt % Fe Ni
o wta) (wt %) (ppm)
600 06 FeY 015 0.25 | 65
‘ | | | 0.6 Fe?* | | 0.18 55
1. | . LoFe?t 0.06 0.36 25
| 1.0 Fe2* 0.24 60
: 800 1.0 Fe2t 0.63 129 15
’ 1.0 Fe2* 0.63 1.17 80
6.0 F2* 4.9 5.9 340
6.0 Fe2t 5.2 6.2 150
600 0.4 Fe3t 0.18 0.25 500
0.4 Fe3¥ 0.7 0.21 600
|, g0 C40FS3t 3.0 3.5 14,800
i 4.0 FS3Y 3.4 39 22,300
TABLE 4.9, STABILITY OF FeF, AND FeFy IN NaF-LiF-ZrF , (22-55-23 MOLE %) AT 600 AND 800°C
: . ; Analysis of Filtrate
Equilibration lron
Temperature " Fluoride Added Fo2+ Total T;ml
. : Fe : i
. {°0) (wt % Fe)
| | - (wt %) (wt %) (ppim)
i 600 10Fe2* 0.83 104 65
| el o - 0.80 1.14 90
" 0,90 e ‘85
77

 

 
i, o o, . i

oo

 

 

 

.

 

 

. ; Sl s L b sl e skiady i

ANP PROJECT PROGRESS REPORT

From the results glven in Tables 4.8 and 4.9, n‘n' )
is seen that roughly 80% of the FeF, added re-
mains as Fe wnh the remainder apparently being

oxidized to Fe , although no logical mechanism

for the oxndqhon can be postulated, From the re-

sults for the runs in whuch FeF, was added, it is
ewdent from the increase in the nickel content of
the system, that the nickel container reduces Fe3*

to Fe2®, No saflsfac’rory balance between Fe2?

“and N|2+ is shown by these data; it is therefore
possflale that the reaction is not the simple one
" proposed. Some rather extensive experimental

‘work would be required fo obtain an insight into

the poss:ble equ:l:bna mvolved

Expeflmem‘a! Preparahon of Pure Fluorldes
LT B J. Stum |
'Matenols Chemlsfry Division

Prepomhon of NH SnF In order to produce
abouf 500 g of NH SnF 3 essenhally free of chlo-

  

" ride ion, stannous hydrox;de was precipitated by

the addition of aqueous ammonia fo a hot aqueous
solution of SnCl, in the presence of Sn°% The
precipitate was washed by decantation, and the
raffinate was removed by centrifugation. After
dissolution of the precipitate in aqueous HF, the
precipitation, washing, ond centrifugation were
repeated. The precipitate was then treated with
the stoichiometric quantity of NH,HF, and was
dissolved in the minimum quantity of aqueous HF,
Crystals of NH,SnF, were recovered by partial
evaporation of this solution.

Compounds of ZrF, with CrF,, NiF,, and FeF,.
Evidence from various experiments has indicated
that the simple fluorides of chromium, iron, or
nickel are not the primary phase when solutions of
these materials in NaF-ZrF, mixtures are cooled;
also, petrographic and x-ray examinations of such
cooled melts show the solid phases recovered to
be quite similar, 1t was believed that primary
phases were complex compounds of ZrF, with the
“structural metal fluorides. Accordingly, ZrF, was
heated at 950°C in sealed nickel capsuies with
each of the compounds CrF, CrF,, FeF, and
N|F2 in 2:1, 1:1, and 1:2 stoichiometric propor-
~tions. The CI'F3 dld not react with ZrF ,, and no
“evidence of compound formation was observed in
-~ the three proportlons tested. All the NiF -ZrF,
fus:ons showed compound formation, When 2 moles
" of NIF2 per mole of ZrF, was used, the compound
cnd ‘excess. NIF were observed; when the molar

  

”

“tatio was reversed, ZrF, and the compound ap-

peared. The compound seemed to be NiF,.ZrF,,
and the constancy of cell size with composition
of the mixture indicated very hmfied or no solid
solution of either NiF, or ZrF, in the compound,

The ZrF FeF fusions showed a compound that
was isomorphous withl\lin'ZrF4 but had a slightly
larger cell size. The compound FeF,.ZrF,, how-
ever, unlike its NiF,.ZrF, counterpart, forms
solid solutions with ZrF . the cell size of the
compound appeared to be lowered by the ZiF,.
When the FeF, was in excess (2 moles of FeF,
to 1 of ZrF ), an additional phase, as yet unidenti-
f:ed was observed; the unchanged cell size of the

FeF,-ZrF, present indicated virtually no solid

solution of these compounds.
Petrographic examination indicated that the com-

pound CrF,ZrF, existed and that it was ap-

parently lsomorphous with its FeF, and NiF,
counterparts, No xeray diffraction data and no
observations of solid solution formation are yet
available. Because of the value of these studies
to the interpretation of some electrochemical
studies (also reported in this section), these
studies will be extended to include systems con-
taining NaF, as well as ZrF ,

Reaction of UF4 with Uranium in Alkali Fluorides
C. J. Barton B. H. Clampitt

Materials Chemistry Division

It was previously reported?! that the reduction
of Uf:4 in the NaF-LiF (40-60 mole %) eutectic
with metallic uranium was more nearly complete in
copper than in nickel apparatus., In those experi-
ments the copper apparatus was equipped with a
filter medium of bronze (82% Cu, 16% Sn, 2% Zn).
The experiments were recently repeated, with a
filter medium of pure copper being used; the data
from these experiments are presented in Table
4.10. The higher UF,/UF, ratios obtained in
these experiments suggest that some dispropor-
tionation of UF, occurred on contact of the melt
with the bronze filters,

However, when the reduction of UF with ex-

cess U° was carried out with the NaF-KF-LiF
eutectic (11,5-42-46.5 mole %) as the solvent, the
reaction yielded U3*/total U ratios of 0,55 to 0.7
regardless of whether copper or nickel containers

 

21c, J. Barton and B, H. Clamplfl ANP Quar. Prog.
Rep. June 10, 1955, ORNL-1896, p68

 

 

-

 

 

 
Ml el ol o il

 

 

 

-

we

    
   

_ TABLE 4.10, REDUCTION OF NuFuL:F UF, wmr
- 100% EXCESS URANIUM IN ALL-COPPER
' APPARATUS AT 750° '

 

Apalysie of

 

_ s Theoreficol_;“v. | Retio of

Filtrate (wt %) u3t U3*te

ud*  Total U (wt %) Total U
249 266 271 0.93
24.5 25.0 25’.2 T 0,98

 

*Coalculated by assummg complete reducfron of UF

7 WI'I'hU .

cnd frlrers were used Exumlnctlon of rhe lnrerlor

of the copper containers used in these experiments
showed the ex:s’rence of a mirrorlike metallic

“coating on the copper ‘that had been in conract”

with the melt, The thickness of the courmg ap-
peared to increase with time and femperature; x-ray

diffraction exommcmon of this material showed it

to be o mlxrure of copper and UCUS. ‘
- The copper in the uraniferous alloy oxidizes

quu're rapldly in contact. wrrh morsf air, Tamished
specrmens “showed the presence oF copper oxides,

in addition to UCU but no U0, lines could be
found in the x-ray drffrccnon parrern.

- Alloying of this type was also observed when
NaF-KF-LiF mixtures were heated with either UF,
or with metalhc vranium. In both cases, cna!ysrs
of the filtrate obtained cr fhe end of the equilibra-
tion period showed the presence of both U3+ nd
U4t in the melt; dlsproporhonahon of_ UF

S TR

ro‘have béén_ mvolved in the formation of rhekUCu

 

The reacflon of metallic uranium wrrh the NaF-
KF-LiF eutectic at high temperatures, reported

prev:ously, 22 has been shown fo occur at 650°C

 

seemsfi

1

PERIOD ENDING SEPTEMBER 10, 1955

a similar reaction occurs at 750°C when the
NaF-LiF eutectic is used.
ments the purified salt mixture and metallic ure-
nium were placed in a nickel opparatus, and, after
formation and decomposition of UH, to subdivide
the uranium metal, the temperature was raised
to the desired value. The contents of the appa-

In the recent experi-

~ ratus were agitated by sparging with helium. An

air-cooled condenser maintained at slightly above
room temperature was included in the exit gas
line to condense any alkati metal or alkali fluoride
The trapped
in water and analyzed

swept out by the flowing helium.
material was dissolved
chemically for potassium, sodlum, lithium, fluorine,
and total alkalinity.

The results obtained with the NaF-KF-LiF eu-
tectic and with the NaF-LiF eutectic are presented
in Table 4.11. It is obvious from the data that
most of the trapped material was alkali metol;
less than 5% of rhe‘,condensqre was alkali fluoride.
Also, both the quantity of alkali metal evolved
and the uranium content of the melt increased
with time in a quite regular manner. = As was ex-
pected, the reaction preduced more K° than Na®
{when both were constituents of the melt) and
more Na® than Li% However, the amount of alkali
metal collected did not, by a large factor, cor-
respond to the amount of uranium oxidized. It
appetrs certain that some condensation of the
alkali metals occurred in cooler portions of the
apparatus before the condenser was reached. The
apparatus is being remodeled to eliminate this
possibility and to demonstrate the stoichiometry
of the process.

The presence “of UF4 in these melts suggests
of the rype ‘ '

d meta[s?fmve lncluded;tsome explororory ‘work
on the ethbrrum between sodlum-pofcssmm alloy

 

22, E, .C[.eary, ANI;'J Quar. Prog. Rep. June 10, 1955,

ORNL-1896, p 71.

79

 

 

 

 

 
 

 

i .. e i

 

ANP PROJECT PROGRESS REPORT

. TABLE 4,11, 'REACTION OF URANIUM METAL AND NaF-KF-LiF AND NoF-LiF EUTECTICS =~

 

’ Ahalysis- of Filtrate

Analysis of Trapped Maier}ra'l“.‘ o

 

 

  

Reoction - Anelvee o0 —
o Total U o N L P Akalinity
S . o mg) (mg) (mg) {mg)  (meq)
o ' 'NuF-KF'-L_iI“:‘E.:;f‘gaéii'c': at 650°C -
3 325 408 126 48 002 105 3 50_""
4 229 5.20 42 69 0.03 255 4.0
5 13,34 5.07 o 8.6 0.07 1.65 5.79
7 ” _'"”_‘5.91'" 08 245 ns 018 377 712
14 © 5.25 | 11.5 340 ‘1o 5. 03 2.75 T 9.24 o
L NeFeLiF Eufechc** at 750° -
20w am0 o 2 oo o0 0se
e s 028 75 oes om0 Taer
e 2.12 3.83 0.49 134 0.15 0.50 63
15 4.22 540 1,63

 

257 0.15 1.30 OMe

 

*Since considerable amounts of alkali fluorides were found in the trapped moterial, the results of the anISrses are
not accurate enough for reporting; however, it is felt that the contaminants did not affecf the total alkalinity deter-

mination appreciably.

**This eutectic was found, by analysis, to contain 0,18% K.

and NaF-KF melts at 800°C., Samples of purified

- potassium metal and NaF were loaded into a

hydrogen-fired nickel capsule in a helivm atmos-
phere. After a welded capsule had been held at
800°C for several hours, it was guenched and
then opened under an atmosphere of helium. Sam-
ples of the metal phase were easy to obtain in
that they could be poured without difficulty into
a test tube having a rubber stopper and a stopcock.
A good sample of the salt phase was much more
difficult to obtain, beccuse droplets of metal
clung tenacmusly to :'r

A more’ cerfcm ‘separation of the phases was

' ”’";'5'-’7'obtolned by evacuating sealed test tubes that
“had been filled in a helium atmosphere with sam-
ples of ‘each phase and allowing the evacuated

tubes to partially fill with anhydrous isopropyl
alcohol. The c|coh07| reacted with the metal but
not with the salt. Since the salts were insoluble

;}"_t;fln the lsopropyl alcohol, the alcohol fraction was
R f_;l_fgred_ and analyzed to give an approximation
- of the Na¥K® ratio, and the residue was dissolved

80

in water to give a water fraction that was analyzed

for the NaF/KF ratio. Results for the one experi-
ment which was carried to completion are pre-
sented in Table 4,12, |

On the assumption that quenching effectively
froze the 800°C equilibrium and that no significant
exchange occurred during the short exposure of the
salt to isopropyl alcohol, the results in Table
4,12 lead to an equilibrium constant, based on
mole fractions of salt in the salt phcse and metal
in the metal phase, of

[Nd] [KF]
* = K] [NaF]

 

 

= 0.2 .,

This compares with the valve of 0.29 obtained by

Rinck?® and the valve of 0.5 + 0.3 based on

standard free energy estimates and idédi aéf‘i\'/‘ivfiés‘. o
It was noted that the salt assoc:ated w:fh 'rhe

metal phase, presumably that salt which was dis-

solved in the metal ot 800°C, contained much

 

23E. Rinck, Ann. chim. (Paris) 18, p 429-434 (1932),

 

LY IR D

 

 

 
 

 

  
 
 

W

TABLE 4,12, ANALYSES FOR THE' EQUILIBRIUM
K + NuF‘-—bKF + Na

 

Metal Phase Salt Phase

 

{megq) {meq)
Na K Na K

 

Alcohol fraction  1.47  3.42  0.26  0.52

. Water fraction .78 0.05  3.22  1.36

 

lf the CrF2 can ‘be re

more NaF than KF; this p'rrobdbly reflects a weak-

ness in the sampling, the separation, or the anal-
ysis, since KF has been shown to be ten times??
more soluble in K° than NaF is in Na% The ratio
of salts in the méfél phase did not affect the
value of Kx; since the salt ratio was based on
analysis of the water fraction of the salt phase.

PRODUCTION OF PURIFIED FLUORIDES

G. J. Nessle F. F. Biankensth
G. M. Watson

Mo'rerlcls Chemls’rry Division
Recovery of Contummcn‘ed Fuel for Re-use

J. E. Eorgcm J. P. Blokely

Materials Chemistry Division

Fuel used for testing “reactor components at
ORNL and at Pratt & Whitney Aircraoft is expected

to become available for reprocessing and re-use

at the rate of 2000 to 3000 Ib per month. If this

material is dlscharged from the testing equ:pmenf

in a manner that wuli mlmm:ze o‘udahon and hy-

.droly51s, ’rhe ma]or con'rcrrmcm’r will be CrFZ..'I
ed before the rmatencl_\ )
'mg cycle, it is

for re-use WOUH-\- . and ré/di_js_s'o_lrve's‘ “the Cr° as CrF,; accordingly,

 

  

  

 

'24M. A. Bredig, H. R. Bronstein, and W, T. Smith, Jr.,
J. A. Chem. Soc. 77, 1454 (1955).

  
 
   
  
 

= 1 mbeing assemb[ed'm
' _‘for frea’rmg 750 to 1000 b of recldlmed fuel w:’rhw:’
~ zirconium_ )
:'mel'r dlrecfly into fhe 250-1b productlon apparafus.

PERIOD ENDING SEPTEMBER 10, 1955

Removal of CrF2 from I*ic:F-ZrF“-l.H"‘4 Mixtures
F. L. Daley

Materials Chemistry Division

The standard hydrofluorination-hydrogenation
method for the purification of fluoride mixtures
will remove NiF, and FeF, by reduction of these
materials to the metals, but it is largely ineffec-
tive in removing CrF,. Accordingly, if the mo-
terials charged to the processing plant contain
appreciable concentrations of CrF,, an additional
processing step must be included. Since the raw
materials NaF, ZrF,, and UF, contain only traces
of CrF,, this contamination is not a problem in
the production of virgin fuel. However, the fuel
materials that will be available for reprocessing
after having been used in engineering test unifs
at ORNL and ot Pratt & Whitney wili contain
CrF, as the major contaminant.

If the used material is stored in contact with
air, considerable amounts of ZrQ, and/or ZrOF,
will be formed by hydrolysis when it is reheated,
and these compounds could also cause trouble
in the reprocessing steps, It is possible, however,
to handle and to store the material so that such
hydrolysis is negligibly small. In that event the
only contaminant present that would not be re-
moved by the normal processing procedure would
be the CrF,, which would probably be present
to the extent of 500 to 1500 ppm chromium.

To achieve nearly quantitative reduction of
CiF, to Cr% a considerable fraction of the UF
must be reduced to UF;. Since the solubility
of UF, in NaF-ZrF, melts is relatively low, even
at 800°C, the amount of reductant must be closely

controlled. If this reduction is performed in the

. standard production egquipment, the subsequent

hydrofluorination merely reoxides the UF; to UF,

the reduction must be accomplished in auxiliary

- equipment, and the reduced melt must be filtered

info 'rhe producti on equ;pmenf
~Two experlmeni's were performed on 3- kg bcn‘ches

'of purified NcF-ZrF -UF, mixture to test” the
<+ feasibility of these operohons on oxide-free ma-
. terial. The batches consisted of NaF- ZrF -UF,
+ (50,8445, 2-4 mole %) mixed and contammafed wrfh
¢ 2 wi % CrF (” 500 p‘pm Cr*tY and zirconium

metal chips of the desired weight. For the re-
duction step the charges were placed in unlined

nickel reactors, heated, held for 3 hr ot 800°C

81

T

 

 
o

 

 

‘mitted for cmoly5|s.
" and filtered material; the other sample was repre-
. sentative of the séfne salt after hydrofluorination
"cnd hydrogencn‘lon ,
~are summcr:zed in Tcxb[e 4 13.

. conium mefuf
"_';‘_-because of the msolublln‘y of UF
As was shown previously,

of Cr

ANP PROJECT PROGRESS REPORT

with continuous helium sparging, and then trans-

ferred through filters into standard hydrofluoring.

tion equipment., The hydrofluorination step and
subsequent hydrogen reduction were carried out
in graphlte-hned mckel receivers.

Forr each experlmem‘ two samples were sub-
One represented the reduced

"The results of the analyses

These resu[fs mdlcate thcn‘, in the absence of

oxyfluondes and oxides, high concentrations of

. Cr™t may be subsfcnhal[y reduced from molten

NGF ZrF ,-UF, mixtures by the addition of zir-
Furthermore, the loss of uranium
is negligible.
the concentration
remaining in the filtrate is a function of
the ratio of U3*/U** obtcined by the particular
excess of zirconium used. The concentrations
of Cr'" obtained in the two experiments are lower
than those predicted from equilibrium calculations,
for which a K, of 4 x 10~4 was used, if the Cr®
is assumed to be at unit activity. This assump-
tion may not be justified in view of the possibility
of the formation of chromium-zirconium compounds.

25

Sufficient zirconium to produce 1.0 mole % UF
reduced the Cr** concentration to 100 ppm, with

 

25, @, Overholser, J. D. Redman, and C. F. Weaver,
ANP Quar. Prog. Rep. March 10, 1954, ORNL-1692, p 56.

_version of ZrQO, to ZrF,

neglligibie vranium loss. This amount of zirconium
appears to be adequate for similar l\lc:F-Zrl'—"‘-UF4
mixtures.

Effect of ZrO, in Fuel Preparahon |

F. L. Daley
Materials Chemistry Division

Since ZrO is expected to be a common con-
taminant in commercml ZrF,, studies have been
made in order to ascertain the efficiency of con-
in the present equipment.
Appropriate 5- |b mixtures of NaF-ZrF ,-ZrO,-UF
which would yield, after complete hydrofluorlna-
tion, 50 mole % NaF, 46 mole % ZrF,, and 4 mole %
UF, were prepared. A maximum of 10 wt % ZrO,
was used. The samples were hydrofluorinated in
the usual manner at 800°C, and the exit gases
were periodically sampled by absorption of H,O
and HF in pyridine. The resulting liquid was
analyzed for fluoride and water. These analyses,
though rather erratic during early stages of treat-
ment, indicated that the hydrofluorination effi-
ciency was not greater than 20% with the agite-
tion possible in the equipment used,

[f only 20% efficiency were obtained in the
production plant, a 50-hr hydrofluorination period
would be required for complete conversion of a
100-kg batch containing 10% ZrO,. Such a pro-
longed period is quite impractical. 1t is obvious
that ZrO, contamination ‘in the ZrF, must be kept
to a low figure if high throughput of fuel that
meets specifications is to be maintained.

TABLE 4.13. CHROMIUM AND URANIUM CONCENTRATIONS IN FILTRATE AFTER REDUCTION WITH
ZIRCONIUM METAL OF NaF-ZrF -UF , (50.8-45.2-4 MOLE %) CONTAINING 2 wt % CrF, AT 800°C

initial Mixture:

Total U, 8.63 wt % (4 mole %); Cr**

(ppm), 11,500

 

Test No, 2**

 

 

 

+

I

. -

-

 

 

 

 

 

Test No. 1%
After After Entire After After Entire
Reduction Treatment - Reduction Treufmen_t
Total U (wt %) 8.38 ' 8.53 8.48 8.49
Udtiwt %) 2.52 0.1 0.97 0.1
Cr?¥ (ppm) 90 130 1400 1355

 

*Exces-e zirconium metal added equivalent to 1.0 mole % UFS'

| *’_‘fIExcese'_zirconium metal added equivalent to 0.5 mole % UF..

8

 

 
 

 

 

-

Pilot-Scale Purification Operations

J. P. Blakely J. Truitt
C. R. Croft W. T. Ward
Materials Chemistry Division

Fifty-three batches totaling approximately 500
Ib of purified fluorides were processed in the
pilot-scale equipment during the quarter. This
material, including 22 batches of BeF ,-bearing
compositions, was made up entirely of specidl
batches for use in physical property determina-
tions, phase equilibrium studies, and other pro-
grams requiring small batches of fluoride mixtures
of possible interest as fuels. The study of proc-
essing techniques for the purification of BeF,-
bearing compositions has continued to indicate
that further refinements will be necessary before
consistently high- qualify material can be pro-
duced. The BeF2 used as charge material has
been found to contain carbonaceous occlusions,
which are believed to contribute to the high sulfur
content of the material. The relation between
carbon and sulfur content in some typical, finished
batches is shown in the following:

Carbon (wt %) Sulfur (ppm)

0.10 208
- 0,37 460
0.27 523
0.25 _ 435
0.05 20
0.01 7
0.01 ‘ 74
0. 01 - 5 |

Since sulfur in amounts of over ]00 ppm is known

to be very detrimental to the life expectancy of

the processing equipment, it may be seen that
rigid spectflcahons must be met before any large-

“seale processing of BeF becrmg fuels cah be
| undertaken.

    

 

J.E.Eorgan  JP. Biakely
Mcterlois Chemlstry Dw;snon

The large-scal e (250-|b bcn‘ch) fluon de-firbcessmg:

facility continued in operation for approximately
one-half of this quarter. Because of an acute
shortage of ZrF ,, operations ceased during July
to allow stockpiling of a sufficient supply to per-

PERIOD ENDING SEPTEMBER 10, 1955

mit full-time production, Operation was resumed
in mid-August, and a total of 6960 Ib of purified
fluoride compositions was produced during the
quarter. [t has been shown that the facility con
produce 1000 Ib of processed fluorides per week.
No attempts have been made to exceed this rate
because of the inadequate supply of ZrF .

Five of the 28 batches processed during this
period were the thF-ZrF4 (50.50 mole %) that
was required as the fuel carrier for the ART high-
temperature critical experiment, The impurity
analyses given in Table 4.14 show that the moa-
terial was acceptable for use.

TABLE 4.14, IMPURITY CONTENT OF ART
CRITICAL EXPERIMENT FUEL SOLVENT

 

tmpurity Content {ppm)

 

 

Batch
No. ]
Ni Cr Fe Hf B
1 10 75 55 125 1
2 10 55 35 100 1
3 20 60 165 100 1
4 50 80 170 100 1
5 10 75 55 80 1

 

 

Preparation of ZrF ,
A. J. Gully

Materials Chemistry Division

Zirconium tetrafluoride has been obtained in the

posf from the Y-12 Production Division, which has
_\op'é'rdfe'c! a facility for converting ZrCl, to ZrF,
" at a maximum rate of 250 Ib per week. While mos’r

of the demands for processed fluorides have been
met thus far, it appears that a source of supply
capable of producing 400 to 600 b of ZrF, per
week will be necessary.

Pilot-scale apparatus for converting ZrCl4 to

- ZrF by direct hydrofluorination at a relatively
Produch: »n Sca|e Qperoilons ’ = low temperq’rure (200 to 300°F) has been designed,
~ built, and tested.
._,____‘Jfrom this equipment have made the process appear
to be very afl'rachve, a unit capable of yleldmg

1000 1b of Zrl':4 per week has been designed, and

Since analyses of the product

fabrication of the equipment has begun. This
facility is expected to be in operation by mid-
September 1955, Experimental operations are being

83

 
 

 

 

 

ANP PROJECT PROGRESS REPORT

continued in the small-scale equapmenf to deter-
mine ophmum operating conditions.

Various private vendors have also been con-
~ tacted as possible suppliers of suitable material.
" To date, only one vendor has submitted samples
which appear to be reasonably promising, and he
has been requested to submit a larger sample, for
~actual charging to small-scale processing equip-
- ment to verify that the material will yield a suit-
‘able final producf Tests of this material will
be made in the near future, At present, it is not

' 'EpOSSIb[e to state whether purchosed ZrF, will be

" more économical than mofencl prepared at ORNL
-from hofnlum-free ZrCI

Butchmg and Dlspensmg Operahons
J P Blakely "~ F. A Doss

Matenals Chemlstry Division

Approxma’rely 100 batches totaling 4129 ib of
_ various fluoride compositions were dispensed in

 batch sizes ranging from 1 to 250 |b. Operational

difficulties plus a downward revision in estimated
usage of processed fluorides in the ANP program
by both ORNL and off-area consumers have, it
appears, postponed the time of maximum fluoride
consumption to October—November of this year.
With presently known testing schedules, an aver-
age demand of between 4000 and 4500 Ib of proc-
essed fluorides per month should prevail during
the fiscal year 1956.

A substantial gain in stock inventory was made
during the quarter, despite the fact that the in-
adequate ZrF, supply curtailed production by
several thousand pounds. A material balance
for this quarter follows:

 

Pounds

Material on hand af beginning of quarter 3,818
Production during quc;rter | 6,961
Total 10,779
Material dispensed during quarter | 4, 129
Material on hand ot end of quarter 6,650

Pratt & Whitney Aircraft reduced their c->riginc1|I
request for 4750 b to 1150 Ib of processed fluo-
rides for this period; however, to prevent a pos-
 sible shortage of material for ORNL-ANP usage

~while the ZrF, supply problem is being resolved,

- “only 750 1b of materlcl was shipped to them. Other

';.—']off-qreq shipments included 44 Ib to Wright Air

M

Developmem‘ Center and 73 Ib to Baflelle Memonal
Institute.

Successful attempts to fill four 50-|b cans simul-
taneously from a 250-1b batch were made, and
the practice was extended to smaller batches.
The time required to perform these opercmons
was, accordingly, reduced '

Loading and Draining Operations
N, V. Smith
Mcn‘enols Chemlstry Dwns;on _

The operohons necessary for fl!llng, drcumng,
and sampling of charge material in all test equip-

- ment other than the thermal-convection loops

have continued at a rate comparoble to that of
the previous quarter. Over 50% of the operations
have involved the handling of alkali metals.

With the rapid increase in festing of alkali
metals, con5|derqt|on of safe, rapid methods
of disposal of used metals became necessary
Commercial vendors will not accept the metals
for reprocessing. The small-scale methods of
disposal formerly used have become inadequate
for the quantities now involved. Information ob-
tained from the Mine Safety Appliances Co. led
to the installation of an underwater jet through
which NaK is forced at 30 psi under 10 ft of water
in the disposal quarry. It is now possible to
dispose of NaK at a rate of about 30 |b/min. With
proper heating of the lines leading to the jet, this
installation could be used for sodium disposal.

The ART high-temperature critical assembly was
loaded with approximately 750 Ib of NaF-ZrF,
(50-50 mole %) and sufficient NaF- UF, (66.7-33.3
mole %) to reach criticality. Most of ?he NaF-UF
was added by helium pressure transfer of fhe
molten material in approximately 5-1b increments.
The final titration to criticality was made by using
the enriching system designed by the ARE Division
for the experiment; the enrlchmg system had pre-

viously been filled with fuel concentrate. After

each significant addition of N02UF6, the ‘sump
was sampled; the uranium content was determined
by the ANP Analytical Chemistry Group.

Enriched Fuel Preparations

J. P. Blakely F.A. Doss
'J. E. Eorgan
Materials Chemistry Division

The relocation and installation of processing

equipment sunfdble for prepara'rlon of enrlched'

 

 

 

 
i alicu SREL
- e

[STCHPL T X

 

 

  

fuel batches was completed. Two batches of
NaF-UF, (66.7-33.3 mole %) containing enriched
uranium were prepared for use in loading of the
ART high-temperature critical assembly; four
batches of material remaining from the ARE were

 

PERIOD ENDING SEPTEMBER 10, 1955

tained in recrystallized alumina crucibles; elec-
trical contact between the half cells was effected
by a porous bridge of ZrO, impregnated with the
NaF-ZrF , mixture,

Attempts have been made to examine cells of
the type

ZT02

M| MF NGF(a]), ZrF

2{(satd)’

@)\l NaF-ZrF,

, MF M,

NGF(a by ZrF 2satd)

 

also included in the stock available for this ex-
periment, . -

- One preparation of NaF-ZrF,-UF, (53.5-40-6.5
mole %) was subdivided into three small batches;
one of these was transferred ihte the first in-pile
loop to be sent to the MTR for testing. A batch
of NaF-ZrF,-UF, (63.0-25.0-12.0 mole %) to be
used in other rad:ohon damage studies was also
proces sed.

FUNDAMENTAL CH EM|STRY OF FUSED SALTS
So]ubllrty of Xenon in Molten Sults
" R. F. Newton '

Resecrch Dlrecfor s Dzvlsron

Recent experlments have shown ’rhcn‘ the gas
stripped from fused salts and measured as xenon
was contaminated with SiF, and with some organic
material that was volcmle at room temperature
but trapped in liquid nitrogen. Accordingly, the
previously reported?® value of about 10~7 mole
of xenon per cubic centimeter of solvent for the
solubility of xenon in the NaF-LiF-KF eutectic
is too hlgh

2 The orgamc contammanf is presumed to ar:se,__l_,'f
,'from sfopcock grease. "The" apparatus is belng”"“m
modn‘led to substn‘u A
. ;Athe sfopcocks.'

 
 

 

 

 

phere was maintained over the half ceils con-

 

26R, F. Newton and D. G. Hill, ANP Quar. Prog. Rep.
Sept. 10, 1954, ORNL-1771, p 70.

y “cutoff traps for -

A barlum oxide ‘ltrqp. Is bemg m.-»-«-_ Fe® and the Fe®Ni® cells agrees reasonably well

   
 
 
  
 
 
 
 
  

where M and M' are CrS Fe% and Ni°% These
cells appear to be reversible, and they are quite
reproducible. However, interpretation of the data
from these cells is complicated because the solid
phase in equilibrium with the melt is not the simple
metal fluoride; evidence reported in a previous
portion of this section suggests strongly that
complex compounds of which NiF,:ZrF, is typical
‘are formed. Accordingly, when M and M! are
both Cr® for example, and when the two half cells
contain CrF, in different amounts, both sufficient
to saturate the solution at temperature, small but
reproducible and significant potentials are ob-
served. For example, when one half cell contains
7.8% and the other contains 12.6% CrF, (solubility
of CrF, at 700°C is 5.8%), the emf of the cell
varies from 15 mv at 550°C to 8 mv at 700°C.
This potential arises because the CrF, added
in excess of the saturation concentration ‘‘pre-
cipitates’’ ZrF 4 and thus changes the composition
of the solvent and affects the activities of the
ZrF, and of the NaF to a different extent in the
two half cells,

Potentials measured at various temperatures for
- cells of this type are shown in Table 4.15. It
‘may be noted that the sum of values for the Cro-

at all temperatures with the values for the Cr°-

Brawer's27 estimates of free energy of formatlon

Tare assumed to be correct, E0 ‘values of 0.35,
0.25, “and 0.60 v are obtained for the CroFe®
. Fe®Ni% and Cro-Nj° cells; these values should
"_,be nearly mdependent of temperature over this
interval.

 

27) . Brewer e: al., p 107 in Chemistry and Metallurgy
of Mz.scellaneous Materzals, Thermodynamics (ed. by

L. Quill), McGraw-Hill, New York, 1950.

85

 

N| cell If the solld phase in ethbr:um' with
__'rhe'\' “melt were the simple fluoride MF and if

 

 
Sl e g e kel ok

B

g B s

b ok G R i

 

 

 

"ANP PROJECT PROGRESS REPORT

Reproducible emf values have been obtained for cells of the type o : .
. Zr0, . o
Cr Cer(Cl)' NaF(a]), ZrF4(a2) NGF(“%" ZrFMa;), CrF2(C2) Cr° , 3

where ¢, is less than the saturation concenfra-
tion of CrF, and c, is more, Data for three such
cells for which ¢, was 0.90 wt % Crf, and c,
had the values shown are presented in Table 4.16.

Additional studies of similar, but less complex,

cells of this general type are to be made, While
a partial interpretation of the available data would

NdF-ZrF4

 

 

~are _qflfg_ampied, | b

% . '
T T T T TR TR YO WP RIS T PR

additional experimental effort appears to be de-
sirable before such evaluation and interpretation

-~

A number of cells were run to determine the
feasibility of employing a platinum wire as an
electrode for half cells containing mixtures of 4

be possiblé by making use of several assumptions,

FeF, and FeF, in NaF-ZrF,. These cells, as

 TABLE 4.15. POTENTIALS OF CELLS OF THETYPE

Bl il bt i i o o i

 

 

 

 

Zr0
: 2
. i ]
M IMF o(saray NoF(q y ZrF 4q,) NaF-ZiF NaFady ZF o ly M Fa(sara) | M - b
a ' B c Summation of ’
Measured EMF (v)* ‘ >
Temperature - Cro%Fe® and Fe2Ni° ¥
(OC) Cfo‘Feo** Feo’Nio** Cf 'Nio** Values (V) -3
550 0.342 0.408 0.754 - 0.750
600 0.345 0.415 0763 0.760
650 0.361 0.422 0.773 0.783
700 0.424

0.374 0.788 0,798

 

*Mean of values from two similar cells,
*_*Cer concentration, 7.6 wt %.
Fe F2 concentration, 7.3 wt %,
" Ni F2 concentration, 5.0 wt %.

TABLE 4.16. POTENTIALS OF CELLS OF THE TYPE

Zr02

' O Wiel = ¢
- e,y NQF(alj' ZFF4(02) NaF-ZeF , NaF Cr,Wnrhc] 0.90 wt %

o
C!" CrF (a})’ ZrF4(a;), ch"""z)

 

Measured EMF {v)

 

 

: rTein'ilpercrfl.ure -'
°C) cy = 6.6 Wt % cQ TTTwt% T ey = 90wt %
550 ' 0.008 0.013 0.014
600 0.032 0.038 0.040
650 0.050 0.058 0062
0.061 0063

700 0.052

 

86

 
-

-

+

e it . e ey QLA WG .y G g ofbod

 

Lk o

-

a®

 

“a lu'rfi‘i'n'd ,

 

Vfoneousiy as part of a ce

“that cell.
electrode becomes effechvely a chromium elec-

well as others, such as

|, UF,
Pt and Pt
CrF, UF,

gave irrepreducible potentials.
- Two uranium rods dipped into a solution of
UF, in molten NcF-ZrF4 showed approximately
zero emf from 550 to 700°C, It is possible that
uranium acts as a reversible electrode at these
temperatures.,
Activity of Chromium in Inconel
M. B. Panish
Materials Chemlsfry Division
As «d foundaflon for future studies of thermo-
galvanic effects in fused-salt melts contained in
Inconel systems, an attempt has been made to
determine the activity of chromium in Inconel.
The elec‘rromohve forces of several cells of two
types have been determined over a temperature
range of 550 to 800°C., The cells being studied

are:

(D Cr |NaF, ZrF,, CrF,

Inconel

and
(2 Cr|NaF, ZeF,, CrF,

The fused-salt melt used in ffiis work was NaF-
. ZrF4 (53‘47 mole %), Wl'llch wcv:. sa'run_'ated w:th_v_'_
Cri':2 to ensure an |‘dent|ca|“ chromlum concentrc_

R4

~ tion in the anode and ccfh e ccm‘lpdr fme"f5° The

 

'>unde_r a__dry he[ium qtmqsphere.

e T e e fi’«f—,f_ SARAEREN

 

   

Thls seems t

AlL0,

(satd with melt)

PERIOD ENDING SEPTEMBER 10, 1955

trode as a result of an internal discharge in cells
of type 1.
A reaction that might possibly occur is

3CrF2-T-__"2CrF3 + Cr°®

If such an equilibrium did exist, there would be
a different equilibrium concentration of CrF, and
CrF, around each electrode because of the differ-
ence in concentration of metallic chromium in the
electrodes. If, as in cells of type 1, it is possible
for convection currents to carry solution from one
electrode to the other, an internagl discharge of
the cell may occur,

Cells of type 2 are obtained experimentally by
using two recrystallized alumina containers, one
within the other. The electrodes are placed so
that one is in the inner container and one is be-
tween the two container walls. The salt mixture
is placed in and around the inner container.

The results obtained with cells of type 2 are
somewhat erratic, but there is a general trend
that is shown by the data plotted in Fig, 4.4.
Further investigation will be necessary to clarify
these results. Particular attention will be paid
to the purification of starting materials and to

 

Inconel .

NaF, ZrF ,, CrF,

 

 

fhe chromous-fo-chromlc jon raflosl m fhe ce” com

 

' under‘ihe c'ondmons of 'rhl's work._ |
T Very little 'has been done fo determine the polar-
lzablhty_ of the lnconel elecfrodes. After a small

indicate fha’r the lncone] “amount of curren'r “is passed ina ceH of type 2,

  
  

 

 

the momtencnce of a water- qnd oxygen-free atmos-
phere. It will also be necessary to investigate

    
      
   

a fyp:cal ‘melt used for cells in this work
_ |_s shown in Flg. 4.5, Alfhough the current across

i, S

y electrocles wQs reversed after q co_nSlderable

 

  

e‘no’r

the cell appears to recover slowly,

87

 

E
E
i
.
.§.

 
e R R ki ks e dae de 2 ¢ ol

 

bl Rk | i

 

 

 

 

 

'ANP PROJECT PROGRESS REPORT

.. * 'UNCLASSIFIED
“ "ORNL—LR-DWG 9181

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.2
1.0 DATA TAKEN AT
7 s T00°%
HE o T00°C
A 750°¢C
. ..08 A 700°C
W
0.6
04
\"-..._._‘_
'--..,_._‘
\‘
"-.._- \
e
\"\A \-..___._.
e ".__-_‘_
. . — Se—A
: "—'-—--——.a...._.:;—. e
. T A e e e A —_——=
0
0 5 10 15 20 ) 25
TIME (hr) . .

Fig. 4.3. Discharge of Type 1Cells.

VYiscosities of Molten Nitrates

F. A. Knox F. Kertesz
Materials Chemistry Division

The capillary viscometer, previously described,?®
has been calibrated by using pure LiNO, and
KNO, and has been used to measure the viscosity
of a mixture of these materials. Attack by LiNO,
on glass at high temperatures has necessitated sub-
stitution of a nickel capillary for these studies.

Measurements obtained?® for KNO, are in close
agreement with those reported by Dantuma?? and
by Goodwin and Mailey.3? Values for LiNO,
agree reasonably well with those presented by
Goodwin and Mailey but are nearly 20% below
those of Donfuma.

A mixture of the two salts (62.2 wt % KNO )

shows a |1necr reiatronshlp befween |og vnscosn‘y

 

 

 

e 2%,
28 (1 908)

 

F. A Knox,F Kerfesz, und N. V. Smlth ANP Quar,

"jf,Prog Rep. Dec. 10, 1954, ORNL-1816, p 75.
29

L 33-4(1928)

R. §. Dantuma, Z. anorg. u. allgem. Chem. 175,

30 M. Goodwin and R. D. Mailey, Pbys. Rev. 26,

and reciprocal temperature over the temperature

interval 230 to 492°C (Fig. 4.6). Below 230°C,
however, the viscosity increases more rapidly
than would be predicted by this relationship, The
values obtained for the mixture are lower by about
15% than the values for the pure components.

Optical Properties and X-Ray Patterns for
Recently Discovered Compounds in
Fluoride Systems

R. E. Thoma

Materials Chemistry Division

G. D. White
Metailurgy Division

- T. N. McVay ‘H, Insley

Consultants

The identifying characteristics of some new

~ compounds encountered m phose studies are hsfed;' I

below. The symbol d(A) means the distance be-
tween reflecting planes measured in angstroms;

I/I] refers to the relative intensity as compared

with an arbitrary value of 100 for the strongest

line; under optical properties, N, and N refer fo

 

 

 
 

 

kit Mt G s o, SN, it NIRRT L A SR
T « ay
: -

0.4

 

T ' Fig. 4.5.
Electrodes inan NaF-ZrF
Saturated with CrF, ot 706 °C.

0.8

046

N

PERIOD ENDING SEPTEMBER 10, 1955

UNCLASSIFIED
ORNL—LR-DWG 9482

 

 

 

 

\\\\\\\

N\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

//

 

 

/ -

2

 

 

 

 

 

 

 

 

 

 

 

 

fo 2o %0 40 50 6
R fma)

70 80 90

Elecfro!fsm Across Two Chromium

(53-47 mole %) Melt

 

1.2 1.4 1.6 1.8

4
TPK)

2.0 2.2 2.4 2.6

X 403

0.40 \\\\\\\\\\
\\ T NS DR
=
5 E et !!\\
___.——-":—_— \
A "~ = \\\\\\\\\
™~ THEORETICAL
0.06
0.04
' 0.02
> 0
550 600 650 700 750 800
TEMPERATURE (°C)
Fig. 4.4. Theoretical and Experimental EMF Values for Cells of Type 2 at Various Temperatures.
- ‘ . UNCELASSIFIED . UNCLASSIFIED
ORNL-LR-DWG 9183 ORNL—LR—DWG 9434
360
320
3 280 @ POINTS TAKEN IN INITIAL RUN e
o - O POINTS TAKEN AFTER REVERSAL OF
L * CURRENT '
. 240 -
. 2 o ~
¢ R : a
. A / L
. E 200 /;7 . i
W i &
- A 3
ca S
>

Fig. 4.6. Viscosities of the LiNOs-KNos (62.2-
37.8 wt %) Mixture at Various Temperatures.

89

 
 

 

 

 

 

 

 

the lowest and highest indices of refraction, re-
spectively; 2V refers to the acute angle between
the optic axes of biaxial erystals; and O and E
" refer to the ordinary and extraordinary indices of
‘refraction of uniaxial crystals.

3L|F' ZrF4 (low=temperature form)

- 4.88
3.67
3.43
2.9
2.79
2.67
2.40
2.07
1.94
1.82
1.80
1.78
1.65
1.59
1.57

 

N
© Ny = 1.465

. Biaxial negative;

ANP PROJECT PROGRESS REPORT

O'pi;icol data:
1.445

a

2V = ~10 deg

,_ Co VV-X-ray data:
549
U 5.40

I/I]
55
35
50
25

6
14
8
15
6

100
19
25
12
12

4
4
4

2LiF-ZrF4
Optical data:
0 1.468
E = 1.478

Uniaxial positive

It

X-ray data:

10
10
25
100
100

14
11
11

60

 

2.15 23

2.05 19
1.95 27
1.70 - 31
1.63 | 45
1.58 | 10
1.54 10
3LiF-4ZrF4 :
Optical data:
N, = 1.463
Ny = 1.473

Biaxial positive;
2y = ™25 deg

Xeray data:

0
d(A) /1,
611 | 26
5.24 34
4.90 14
4.21 28
4.00 ' 10
3.90 94
3.77 ' ' 18
3.69 6
3.33 60
3,29 20
3.26 ' 22
3.16 100
2.615 16
2.303 10
2,248 10
2.227 ' 6
2.194 86
2.159 16
2.043 12
2.130 34
1.947 36
1.912 22
1.883 10
1.721 10
2KF « BeF,

Optical data:

Average refractive index = '[357

Biaxial positive
2v = ~20 deg

X-ray data:

o
d(A) 74 I

 

.).‘..

-y

 

 
PERIOD ENDING SEPTEMBER 10, 1955

 

 

3.24 12 2.292 15
3.14 40 2.220 7
. 3.07 - 8 2.190 4
.J : 2,99 12 | 2.010 35
1. 2.95 37 1.979 23
2.84 100 1.745 7
2.73 18 1.641 10
2.61 20 |
2.58 12 KF+NaF « ZrF
2.465 85 Optical data:
. 2,359 | 85 N, = 1.375
2.332 55 Ny = 1.382
2.259 7 Biaxial negative;
¢ 2242 15 2V = ~60 dog
2.184 35
2.159 : 25 ' . Xeray data:
2.106 5 d(A) /1,
2.060 7 5.99 10
2.038 12 5.37 50
. 1.955 12 4.98 16
1.920 12 4.85 60
1.905 10 4.50 50
| ¥ 1.894 8 4.41 6
1.854 12 4.25 30
. 1.815 15 4.09 64
e 1.736 15 3.69 28
1.708 | 35 3.60 12
. 1.657 7 3.52 6
1.641 7 3.34 100
3.26 44
KF:BeF, 3.18 40
Optical data: 3.12 12
Average refractive index = 1,315 3.07 12
z Biaxial positive; 3.01 38
2y =~60deg S 2.747 | 4
B -‘.-X;m.y dm . : e 2,690 24
3 N : 2.556 20
- - dA o v 2.489 24
Tess o s | 2405 | 20
599 60 o 2.258 20
L ogee P 2154 1
358 T T e 225 20
333 90 ' 2.050 o 34
3.23 35 1.947 20
3.01 100 1.928 12
2.92 6 1.886 12
S 21 1792 14
| 263 200 0 L743 8
| 2.542 5 ' 1.723 14
2.442 5 1.694 6
2.410 7 1.651 24

91

 

T R I T T T

 
   

 

 

 

 

 

'ANP PROJECT PROGRESS REFORT '
_— ~ 3KF«3NaF «2ZF 3.78 . S ' E
i S T S " Optical data: | | 3.74 | VE
; N D L | 2;192 :3 :
N) = 1.422 : 1}
, " Biaxial positive; 3.36 62 *
T2V = ™30 deg 2.630 13
e R - | 2.074 .75
... . Xeray data: _ 1.935 12
d(z) v 1.766 52
VX I 17 1.506 | 25 :
490 | 100 | o :
4.82 25 2NaF+LiF «2BeF, =
B 455 . 15 Optical data: _ .
' 4.'27 , 33 Average refractive index = 1.312
7.4'09 75 Uniaxial negative;
3,'78 : , 12 Low birefringence
3,40 21 | .
- 3.28 | 21 . X-tay datat
2,99 21 d(A) | Vi,
Con297 88 5.37 24 T
| e 12 4.07 15 |
| e 12 378 n ,
S e 585 6 3.57 18 "y
a_ 2,455 - 25 3.39 21 |
1 2.417 21 3.08 7 ) ‘.
2.368 12 2.99 100
2,270 25 2.614 12 | '
" 2,226 52 2 417 15 ' =
< 2,135 42 2.343 57 b
- 2.051 49 2.303 07
1,931 15 0.240 20
1.822 42 2.707 ‘ ‘
1.748 21 2.149 26 -
1.721 25 9111 o .
‘ ' 2.039 ' 7 '
3NaF - 4ZrF 1.999 6 b
Optical data: 1.968 ' 100 o
N, = 1.420 1.928 9 '
Ny = 1.432 1.894 10
. Biaxial poéifive,‘ 1.840 30
oy =30 deg 1.790 19
_ x-roy data: 1.767 4
e 1.748 7
- dlA) /1 1.720 19
7.5 - 1.704 s
7.42 ” 45 1.676 ' e
547 14 1.612 | 12
4,15 ' 100 1.589 e

 

 

 
High-Temperature X-Ray Spectrometer Studies
G. D. White, Metallurgy Division
* T. N. McVay, ansultaht

by ’ Samples of 3LiF.ZrF, were x-rayed at elevated

' temperatures by using the furnace attachment to
the x-ray spectrometer. The investigation was
conducted to confirm the decomposition tempera-
ture of Li,ZrF,, which had already been deter-
mined by quench methods and thermal data, and
also to determine whether there was a high-
temperature polymorph of Li,ZrF, that was not
retained when the material was quenched.

At room ’remperai’ure the samples contained three
phases: Li ,ZrF g, LigZrF,, and LiF. In this
study the procedure qu to heat the mounted
sample in the evacuated furnace attachment to
550°C and then maintain it at that temperature

~ until the x-ray diffraction pattern contained none
. - ofthe Li2sz6 or LiF peaks. The reacted sample
o was then cooled to various temperatures until the
x-ray pattern mdlcated the presence of Li ZrF
andLiF. In this manner ’rhe decomposition temperq-
ture was deTermmed as being 470°C, which is
to be compared with the value of approximately

 

S

sl  475°C obtained by quench methods, At no tem-
_ perature above 470°C was the x-ray pattern for
. the low-temperature form of Li,ZrF, obtained.

The high-temperature x-ray pattern, presented be-
low, is evidently the pattern for a polymorph of
Li,ZrF,, which inverts to the lower temperature
form at a temperature just slightly above the de-
compos ition temperature: | h

aw

  
  
 
  
 
 
  
  

   

g

59

  

 

Physical Chemlstry of Fused Sults
e g - E - . .

     

Final results have been obtamed for the elec-

trical conductance and density of all molten alkali
chlorides, bromides, and iodides. The equivalent

 

 

     
  
 
    

. R. ,VGn:A‘rfsda]eh” -
Chemlsfry Dlvnsuon

PERIOD ENDING SEPTEMBER 10, 1955

conductance at corresponding temperatures has
been correlated with such properties as ion size
and mass. Conductance has been treated as a
rate process, and values of the heat of activation
and entropy of activation have been computed
from experimental data. In general, the heat of
activation is somewhat temperature dependent;
this dependence is approximately linear for lithium,
sodium, and potassium salts, while it deviates
progressively more from linearity with rubidium
and cesium salts. The entropy of activation for
these salts is in the range of —6 to —8 eu. This
small negative entropy of activation is reason-
able and indicates a similar conduction mechanism
for all the salts, but no detailed quantitative
significance can be attached to it at this time.

An extensive series of freezing-point-depression
measurements has been made in which molten
sodium nitrate was the solvent, The heavy-metal
halides CdCl,, ZnCl,, CuCl,, PbCIz, and CdBr,
all show less than complete ionic dlssocmhon
in molten sodium nitrate; dissociation decreases
as conceniration rises. However, the slopes
of the curves indicate complete dissociation at
infinite dilution, The significant discovery has
been made that the scecalled ‘““common ion effect’’
is generally applicable to these salts. Thus the
addition of any of a large number of completely
dissociated chlorides greatly represses ionic
dissociation of partially dissociated chlerides such
as PbCl, and CdCl,. It has been demonstrated
in the case of CdCl, that the complex ion CdCl, ™~
is formed in the presence of relatively low con-

" centrations of excess common chloride ion. Re-
___actions have been proposed to account for the
““results, and equilibrium constants have been cal-
‘ culated.  Apparently those chlorocadmium com-

plexes containing even numbers of chlorines are
the more stable.
—~Several preliminary measurements have been

.- made of freezing-point depression by K2Zr|:6 and
Kz'i'iF‘5 in NaNO3. The salts were prepared by

~ wet chemical methods. The results indicate that

the complex ions ZrF ;== and TiF ;=7 are reason-

’ably ‘'stable ‘at a concentration of about 0.1 molal
.. _in_molten NaNQ, at its melting point. However,
_,___H_‘Gthere appears to be some slight dissociation,
:_whlch presumably ylelds F~ and ZrF or TiF,~

‘Dissociation is greater with htanlum than wH'h

 

31Details of this work will be published in separate
reports and articles by the ORNL Chemistry Division,

93

 
 

 

" ANP PROJECT PROGRESS REPORT

~zirconium, A similar result was obtained pre-

_fvnously with K ZrF‘S prepared by a dry fusion

F’reC|se determlnctlons of the self-diffusion co-
7,_'jeffl<:|en'rs of sodium ion and of nitrate ion have
“been completed in molten sodium nitrate, A radlo-
chemical frccer technique employing Na?? was
used for Na*, and a mass spectrographic tracer
technlque employing O'® was used for NO, ™.
The self-diffusion coefficients are expressed with-
in 1 to 2% by 'rhe fo”owmg equahons

1.288 x 10=3 (A9TO/RT

‘Il"

D+

8974 x 104 o=5083/RT

.~ThIS is the flrsf 'rlme ’that self-dn‘fusmn of both

""flons of a smgle fused salt has been measured,

g ft is highly significant that the heats of activa-
- tion for 's}elf-vdkiffdsion of both the cation and the
anion ore ‘the same within experlmental error
- ("-'80 cul) Thls indicates that there is a single
- frictional coefflment for diffusion within the melt.
The resu[fs show that the simple Nernst-Einstein
 equation is inapplicable to this molten salt; it
s probably not applicable to any molten salt.
" The ratio of self-diffusion coefficients for Na¥
~and NO,~ is somewhat less than the inverse ratio
- of the square roots of the masses of the two ions,

It is expected that both mass and size are im-

‘portant factors in determining diffusion, Experi-
ments are in progress with other fused salts, and
~attempts will be made to obtain generalizations
“concerning diffusional properties.

 

Diffraction Studies ofnL.icjb'i'dér
P. C. Sharrah P, Agron
H. A, Levy M. Danford
M. A, Bredig

Chemistry Division

The previously described32:33 liquid diffrac-
tometer has been thoroughly aligned and tested
and is being applied to studies of molten salts,
X-ray diffraction patterns obtained from liquid
mercury at room temperature were used to test the
instrument; the pattcrns were satisfactory. These
diffraction patterns and the analysis giving in-
formation concerning the distribution of atoms
within the liquid have been presented, 33 It has
been possible to obtain data which appear to be
reliable to a somewhat larger value of the variable

= (47 sin 6)/A than that reported in the litera-
ture.  Work is under way with molten lithium
chloride. |

Neutron diffraction work on molten salts is also
being carried on so that information from the two
technlques can be coordinated. The equnpmem‘34
consists of the” Chemistry Division neutron spec-
trometer and a furnace for handling the molten
materials., Diffraction patterns of KCI and Li7l
have been obtained.

 

32p, C. Sharrah et al., ANP Quar. Prog. Rep, June 10,
1955. ORNL.-1896, p 81.

Cbem. Semiann, Prog. Rep. June 20, 1955, ORNL-
1940 (in press).

34P. C. Sharrah and G. P. Smith, J. Chem Phys. 21,
228 (1953).

 

 

 

 

 
 

 

 

- “with sodlum-pota

PERIOD ENDING SEPTEMBER 10, 1955

5. CORROSION RESEARCH

W. D. Manly

G. M. Adamson

Metallurgy Division

W. R. Grimes

F. Kertesz

Materials Chemistry Division

Several Inconel forced-circulation loops that
were operated with fluoride mixtures and with
sodium as the circulated fluids were examined.
The effect of operating time on corrosion and
mass transfer under the dynamic conditions was
studied for loops that circulated NaF-ZrF -UF ,
as well as the effects of the method of heating the
loop and of the length of the heated section, The
loops in which sodium was circulated were used
to study the effects on mass transfer of varying
the temperature differential in the system and of
varying the oxide content of the sodium and were
used to compare mass fransfer in Inconel and in
type 316 stainless steel,

Additional Inconel thermal-convection loops
were examined to determine the effects on cor-
rosion and mass transfer, in loops that circulated
fluoride mixtures, of varying the loop cleaning
method, of using direct resistance heating, and of
applying electromotive forces. In one loop, the
wall temperatures in the heated zone were
measured, The effects of oxide additions were
studied in loops that circulated sodium.

Several hot-pressed metal-bonded tungsten
carbide cermets were screen tested in NaF-ZrF -UF
and in sodium, and additional solid-phase bonding
tests of cermets were made, Inconel plc'red with

’ruthenlum was sub|ec’red to cree-p-rupfure fes’rs,_

- and oddmonal tesfs of brazed T-joints in fluoride
f'mleures and in sodlum Were made. A Hastelloy
| "",B—Inconel system was “checked for dlSSlmllar-
L metal mass fransfer in a fluoride. mleure. /
¢ A study of mass transfer in an Inconel sys’rem‘ '
_'__"‘;‘cwculahng “sodium  was initiated,
- corrosion tests wer

and seesaw

   
 

-----

tic hthlum._

 

 

spectra in fused hydroxides at high temperatures,
mass transfer and corrosion in fused hydroxides,
and thermal dissociation of sodium hydroxide,

¢ made on Inconel tubes loaded
-lithiom mixtures. In other
icorrosnon tests molybdenum, vanadium, and niobium
' '""ff'were fes e_d A
" The >furri'domen16| corrosion research repon‘ed
included additional s’rudles of fiim formchon on
metals, techmques for measuring absorphon'w

Chemical studies were made of corrosion of Inconel
by NaF-LiF-ZrF -UF,, the stability of UF, in
NaF-KF-LiF, ang the effect of chromium on 'rhe
mass transfer of nickel in NaOH.

FORCED-CIRCULATION STUDIES

G. M. Adamson R. S. Crouse
A, Taboada
Metallurgy Division

Fluorides in Inconel

G. M. Adamson R. S, Crouse
Metallurgy Division

Examinations were completed of several Inconel
forced-circulation loops in which NaF-ZrF ,-UF
(50-46-4 mole %) was circulated. The loops anj
the operating conditions are described in Sec. 2,
‘‘Experimental Reactor Engineering'’ and in the
previous report.! The corrosion data reported in
Table 5.1 and the analyses of the fluoride mixtures
given in Table 5.2 were for Inconel loops operated
with a temperature differential of 200°F, a
Reynolds number of the fluoride mixture of 10,000,
a maximum fluoride-mixture temperature of 1500°F,
and a maximum wall temperature of between 1600
and 1625°F. These loops were heated by direct
resistance of the Inconel tubing. Loops 4695-4A,
-4B, -4C, -4D-1, and -4D-2 were a single loop in
which the two heated legs? were replaced at the
end of each experiment. A new batch of fluoride
mixture was used for cleaning and operating each
test, The loop was cleaned by circulating the
fluoride mixture for 2 hr at 1300°F. The loop was
then filled for the test with a fresh batch of
fluoride mixture, which was circulated iscthermally
for 25 hr before the temperature differential was
applied.
~The period of isothermal operation was for the
purpose of establishing chemical equilibrium,

 

16. M. Adamson and R. S. Crouse, ANP Quar, Prog.
Rep. June 10, 1955, ORNL-1896, p 83.

2For experimental arrangement see Fig. 5.2, p 86,
of the previous report (ORNL~1896).

95

PR T O v

T T

TR T TR

 

T

 
 

st RS S s

 

 

 

granular voids to a depth of
5 mils

ANP PROJECT PROGRESS REPORT
TABLE 5.1, EFFECT OF OPERATlNG T|ME ON DEPTH OF ATTACK IN |NCONEL FORCED-ClRCULATION
LOOPS IN WHICH NoF-ZrF -UF (50-46 -4 mole %) WAS CIRCULATED
Operating
‘Loop .
" Ne Time Attack of First Heated Leg Attack of Second Heated Leg
' (hr)
4695-4A 0 | Light, general, subsurfaice voids to Light, genercxl,r uii"ét‘granular subsurface
: S voids to a depth of 0.5 mil . vonds to a depfh of 0.5 mil
3 -5A - 10 Heavy, genet"a.l voids to .0. .débfh of Heavy, eneral voids to a depfh of
- 3.5 mils _3m|{$_
3 -4D-1 ~20.’5' ' Herovy,'gene;al;r i:r;fefgfdnfiiaf voids ‘Hercl.v'y, general, ‘|'n"re|.;g'rr;:'lAn;Jlla'r voids toa
i L to a depth of 3 mils depth of 3 mils
4 -4B 50 Heavy, generc;l voids to a depth of Heavy, general v0|ds to a depth of
3 3 mils _ 3 mils
- =4C 100 Heavy, general voids to a depth of Heavy, general vo'ids‘fo'a:clepth of
3 mils 3.5 mils | |
3 -SB 241 Heavy, general, intergranular voids Heavy; general, lnf:ergfdnblaf voids to a
o ' o to a depth of 4 mils depfh of 5 mxls
'.»,-4D'-2 ' 500 Moderate to lighf, general, inter- Moderofe, eneral, mtergranulur volds o
e : granular voids to a depth of to a depth of 5 mils
3n5 mils
;-5C-_2 1000 Moderate to heavy, general, inter- Moderate to heavy, general, intergranuiar

voids to a depth of 7 mils

 

*Time after temperature differential imposed.

TABLE 5.2, ANALYSES OF FLUORIDE MIXTURES BEFORE AND AFTER CIRCULATICN IN LOOPS

 

 

 

 

'Aft‘ér termination

Uranium ‘s
_ Impurities (ppm)
Loop No. When Sampled Content -
(wt %) Ni Cr Fe
4695-4A During filling 8.36 6 70 70
| After termination 8.26 9 245 30
-5A During filling 8.71 15 70 45
After termination 8.77 25 635 30
-4D During filling 8.46 15 65 40
_ After termination 8.57 40 520 50

-4B During filling 8.80 50 65 | 30
After termination 8.85 10 800 25
-4C During filling 8.33 15 35 20
o After termination 8.84 8 725 50
-5B During filling 8.94 7 60 30

After first termination 9.15 20 ' 765 , * 50" ‘
o o After Seéond termination 8.95 30 725 75
. -4D-2 During filling 8.83 30 90 10
Cee After termination 9.14 55 365(?) 70
. -5C-2 ~ During filling 9.12 5 105 60

e 9.15

56 505 45

 

s it Skl e L Gk - Sekil

 

 

 

 

 
 

 

 

 
 

the depfh of a'rtacrk and the ch

were still qun‘e low, Durmg fhe ubsequent first

few hours of operdtion with a temperature dlffer-
ential and a high waH temperm‘ure, the attack was
quite rapid, with more a’rtuck bemg found after the

first 10 hr of opercmon thcn had prev:ously'

occutred in 25 hr of isothermal operation. After
the first 50 hr fhe chromrum contenf of the fluoride
mixture remcmed constant, bu’r the depfh of attack
increased, in conflrmohon of the thermai-con-

vechon-loop data on mass transfer. Considerable

scatter is present in the data, but a depth of attack
of between 3 and 4 mils per 1000 hr of circulation
seems to be a reasonable value for the second
stage of attack. Typical hot-leq sections from
these loops are shown in Flg. 3.1,

Two “other loops (4950 2 “and_ 4935- 2) wefe
operated in as nearly ‘an’ Jdenhcal manner as‘v‘
possible, except that one was heated in a gas

furnace and the other was heated by the direct
electrical res:stance of the pipe wall, Both loops
were operated for 1000 hr, with NaF- ZrF4 UF
(50-46-4 mole %) as the cwculcfed fluid, a temper-
ature differential of 200°F, a maximum fluoride
temperature of 1500°F, and a fluorlde-mleure
Reynolds number of 5000. In both |<:ops a moderate

concentration of subsurface ‘voids to a max1mum

depth of 5 mils was found. A typical area from

the electrically heated loop |s shown in Fig. 5.2.
These loops had 17-ft-long heated sections to
keep the wall temperatures down c:nd to show that

 
 

 

 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

 
 

dn‘ferenthl condn‘no s, it _
increase fhe powe p _run:t of heq ,e'f 'engfh whlch
resulted in a 100°F mcrease in wall temperature,
This loop showed heavy ‘subsurface-void attack
to a depth of 18 mils. These data are additional
evidence that the wall temperature is a more
critical variable than is the fluoride-mixture
térfiperaturé. A series of loops with varying,
but contro“ed wall temperatures are now being
operated.

   

Sodium in Inconel and in Stainless Steel

G. M. Adamson A, Tabeoada
Metallurgy Division

Two Inconel forced-circulation loops (49512
and 4951-3) in which sodium was circulated were
examined after operating for 500 hr with a hot-leg
temperature of 1500°F. Loop 4951-2 had a 300°F

temperature differential, and loop 4951-3 had a
- 150°F temperature differential.

The two loops
were of a test series which included loop 4951-1,
operated previously, which had a temperature
differential of 200°F Table 5.3 presents the
metallographic and chemical dota obtained in
this series of tests.

No correlation can be observed between the
amount of mass transfer and the temperature
‘differential. However, the three loops were not
identical, and high oxide impurities of different
amounts were found in these loops. The results
of these differences could have obscured the
effects of the different temperature differentials.
d gLrpulahon Ioops were

   

 

 

 

n 0.5 450
s us 240

 

 

97

 

 

 
 

ANP PROJECT PROGRESS REPORT

 

 

   

 

 

 

-

i

 

 

 

e P

 

 

o

 

- .

 

 

 

 

 

 

 

 

100 hr _ 241 hr

 

 

 

 

 

 

 

 

 

 

Ll 500 hr 1000 hr

" Fig. 5.1. Chdnges in Attack with Increasing Operating Time in Forced-Circulation Inconel Loops.
i Fluorlde mleure circulated, NaF- ZrfF, UF (50-46-4 mole %); maximum fluoride-mixture temperature,
]500°F temperature differential, 200°F f[uorlde-mleure Reynolds number, 5000.

  

 

S

 

 

 

         

   

 

 
 

 

hot-leg temperature of 1500°F, o temperature
difference of 300°F, and a Reynolds number of
15,000, This loop showed the maximum mass
transfer found to date, There was a 30-mil-thick
deposit in the economizer, and there was attack
to a depth of 2 mils in the hot leg. Three sections
from the economizer and one from the cold leg
are shown in Fig. 5.3.

The second loop, 4951-6, which was operated for
1000 hr, included a bypass cold trap for removing
oxides. The hot-leg temperature was 1500°F,
and the temperature differential was 300°F
Metallographic examination showed mass-trans-
ferred deposits to a maximum thickness of 11 mils
(Figs. 5.4 and 5.5), which is comparable with
the thickness of the deposits found in the loop
with no cold trap. There was attack in the hot leg
to a depth of 1.5 mils (Fig. 5.6) that was of the
intergranular type found previously in sodium-
Inconel systems., Analyses of the sodium after
operation of the loop showed from 150 to 290 ppm
O,, and thus very’ little of the oxide had been
removed,

In an effort to obtain a more qualitative picture
of the amount of mass transfer in the sodium
loops, all the sodium was melted out and the

 

 

 

mum fluorlde-mleure temperature, 1500°F; tempera-
ture differential, 200°F; fluoride-mixture Reynolds
number, 5000.

       

: __jFié. 5 2, Typicul Attack in Direct-Resistance--

" Heated Leg of Inconel Forced-Clrculuhon Loop

- (4950-2) Operoted for 1000 hr. “Fluoride mlxture'i'\
- circulated, NaF-ZrF UF4 (50-46-4 mole %); maxi-

PERIOD ENDING SEPTEMBER I‘O, 1955

deposited metallic crystals were brushed out and
weighed, With this procedure, any well-bonded
crystals or layers were left in the loop. The data
thus obtained from the loops operated to date are
presented in Table 5.4,

Operation of the type 316 stainless steel loop
(4951-7) listed in Table 5.4 was terminated by a
power failure after sodium had been circulated
for 476 hr. There was no oxide filter used in this
loop. It was the third loop operated in a series of
tests; the other two loops (4689-5 and -6) were
Inconel with type 316 stainless steel cold legs.,
Loops 4689-5 and «6 operated for 1000 hr under
similar temperature conditions, The maximum
thickness of the mass-transferred layer in the

UNCLASSIFIED
T-8097

HONI

 

 

ECONOMIZER SECTIONS

 

 

 

 

5.3. Three Sec‘l'loi»'ls-‘-f'q-fbm_'fhe Economizer

Fig.

“and One Section fiéifi'n‘tl"ié"”'C-o'iawLeé “of Inconel
Forced Clrculuhon Loop 4951.5 Which Circulated
for 1000 hr Sodium to Which 0.15% 0, Had Been

Added as Na,0,. Hot-leg femperature, 1500°F;
temperature differential, 300°F; Reynolds number,
15,000.

99

 
 

 

 

e o b e ety

ANP PROJECT PROGRESS REPORT

UNCLASSIFIED
-7?46

 

COLD-LEG SECTION

  

 

ECONOMIZER SECTIONS

Fig. 5.4. Three Sections from the Economizer
and One Section from the Cold Leg of Inconel
Forced-Circulation Loop 4951.6 Which Included
a Bypass Cold Trap and Which Circulated High-
Purity Sodium for 1000 hr. Hot-leg temperature,
1500°F; temperature differential, 300°F; Reynolds
number, 15,000.

all-stamless-steel loop (495] 7) was 0.8 mil

,_'-(Flg. 57), which is much less than the 9 mils
“found in Inconel-and-sfclnless-steel loops (4689-5
‘and -6). Two different layers were present in the
“all-stainless-steel loop.

The majority of the
deposited material was in the economizer, and it
was found by chemical analysis to be 14.9% Ni,
57.5% Cr, and 20. 0% Fe. The second layer was

f_-'j',}llmifed to the electromagnetic flowmeter area in
*;;the cold Ieg, and it was a smooth, adherent
“deposit that was found by chemical analysis to be

- 146% Nl, 19._]%_ Cr, and 55.2% Fe.

i00

 

 

 

Fig. 5.5.

1000 hr. 250X. Reduced 34%.

 

Fig. 5.6. Typical Hot-Leg Attack in Loop 4951-6.
250X. Reduced 33%.

THERMAL.- CONVECTION STUDIES

G. M. Adamson E. A. Kovacewch
Metallurgy Division

T. C. Price
Pratt & Whitney Aircraft

Effect of Various Loop Cleaning Methods

The standard procedure for cieahingnlh‘c'dhelw“

thermal-convection loops has been the prehmmary
circulation of a fluoride mixture for 2 hr with the

system isothermal at 1350°F, The cleaning was

undertaken to assure correlation of data between

loops; it was not expected to decrease corrosive

Deposited Layer in Economizer of

‘Inconel Loop 4951-6 Which Circulated Sodium for

 

~y

 

 

 

 

 

 

 

 
o B, B N

 

kbl . BARRIR.

 

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 5.4, WEIGHTS OF DEPOSITED LAYERS IN VARIOUS LOOPS WHICH CIRCULATED SODIUM

 

Oxide Content (wt %)

 

 

l.oop I .W;igi-wt"of 'Deposit Maximum Layer Thickness
Difference from Control Loop ] Before After
No. (9) (mils)
Operation Operation
49511 200°F temperature 7.9 1 0.034 0.046
differential ' '
-3 150°F temperature 9.0 8 0.031 0.024
' - differential -
S e s | 0.021
-5 Oxide added, operated 25.8 30 0.036 0.027
1000 hr
=6 Bypass cold trop in 7.6 11 e 0.041 0.017
7 Type 316 stainless 08 1 0.087 0.04

Lo s'ree_lrtt_:ubing 7

 

 

 

*Inconel loop operated for 500 hr; temperature differential

15,000. '

 

attack during experiments. Over the past few
mdnfhs,'“hb’wléver",' th‘e"'ddt& obtained have not been
S0 reproduc:ble as those obtained prevnously.
To determme whether the clecmmg operation was
respon51ble for this lack of reproducrblllty, a
series of Inconel loops that had been clecmed by
‘various me'rhods were operatecl for 250 br with
NaF- ZrF4 UF4 (50-46-4 mole %) as fhe “circulated

f_l;u_d The data from these loops ure ‘presented in

 

u,\ o

, ifiréwously fOUl‘Id‘ The:resulfs

   
   
      
    

 

      
  

DV RS

 the ¢ K __M. caflef of the
results being obtal‘hed with stancfard, |oops. ‘

 
   

 

 

;:These Ioopswall operated ‘with c“stcndurd hot-legjt
temperature  of 1500°F, and they circulated

| i\']dF-Zer'-UF:'4 (50-46-4 mole %). Loop 618

  

" that the wall ‘temperatures under the heaters may

, 300°F; hot-leg tempéroturé; 1500°F; Reynolds number,

   

showed the usual subsurface-void type of attack

to a depth of 10 mils after 500 hr, while loops 619
and 703 showed similar attack to depths of 13
and 15 mils after 1000 hr. These depths of attack

are similar to those obtained in the loops with the

- clamshell heaters, and thus the previous conclusion

’rhatljthe depfh ‘of attack is not affected by the
hea ng. method |s conflrmed

 

it was shown in 'rhe forced-cn‘culahon 1oops
that the maximum loop wall temperature was a
more lmpor'ranf vorldble than was the max1mum

 
  

depth of “attack found in recently operated forced-' .
han_that found o

 

c:rculahon Ioops has been less

   
 

 
 
  
   

be as much as 1670°F. This 170°F differential
from the 1500°F bulk-fluoride-mixture temperature

101

_ally ‘measured by a
y hof Ieg abouf Jin,

T T |

 

 

TTT Ree

 
 

 

 

“ANP PROJECT PROGRESS REPORT

el uNCLAssmED]

T- 8062

 

COLD-LEG SECTION

 

ECONOMIZER SECTIONS

Fig. 5.7. Three Sections from the Economizer
and One Section from the Cold Leg of Type 316
Stainless Steel Loop 4951.7 Which Circulated
Sodium for 476 hr. Hot-leg temperature, 1500°F;
temperature differential, 300°F; Reynolds number,
15,000.

would be éhbugh to explain the greater depths of
afiack

Effecf of Apphed Electromotive Forces

A series ‘of thermal-convection loops ‘were
operated with small applied potentials to determine
_whether the corrosion mechanism is electro-
chemical in ncture and at the same time, to
""determlne whether any deleterious effects would
rbe found wnth sfray currents, Wires were attached
“to the hot and cold legs of a loop, and a potential
/fwas Opplled by a bcn‘tery charger. The current
i "How was small, averaging only 5 amp at 1 v.
"'_'_’All these Ioops were fabrlcafed of Inconel, and

 
 

    

 TABLE 5.5. EFFECT OF LOOP CLEANING
METHOD ON DEPTH OF ATTACK

 

 

Ldop A Maximum Attack
No. Method of Cleaning C (miis)
725 Fluoride mixture _ 8 |
726 Fluoride mixture 7 8
722 Nitric ond hydrofluoric 8
acids
737 Nitric and hydrofluoric 8
acids
723 Dry hydrogen 9
724 Dry hydrogen 8
727 No cleaning 7
728 No cleaning 7
732 Layer machined from 7

inner wall

 

they circulated NaF-ZrF ,-UF, (50-46-4 m.o[e %)
at 1500°F. Results of some of the short-time

tests were reported previously, but the data are.
repeated here to present a complete summary, The_

data from the completed series of ‘tests are
presented in Table 5.6.

The loop that was operated for 2000 hr with a
positive charge applied to the hot Ieg showed
only about one-half the depth of attack found in
the loop operated with a negative charge cpplied
to the hot leg. With a negative hot leg, the depth
of attack was about the same as that found
the control loop with no applied potenfial From
these few data, it does not appear that small

s'rray pofem‘ials will increase the attack. Alfhough B
it would be difficult to apply a potential in'a
system as complicated as a reactor, it appears'
that it would be possible to reduce the attack by

applying a positive potential.
Effect of Oxide Additions to Sodlum

A series of Inconel thermal-convectlon |oop5m

was operated with varying amounts “of Na 2,0,
added to the sodium being C|rcu|cn‘ed ’ro defermme

whether mass transfer in the system was. caused ) v

by oxide impurities in the sodium. ; :
The data obtained, presented in .‘.;Table 5.7,

show an increase in mass transfer with increased

oxide content.

 

Corresponding lléhg‘thg 7 werecut o

 

T

 

 

 

 
 

o

¥

 

.L‘

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 5.6, EFFECT OF APPLIED POTENTIAL ON DEPTH OF ATTACK IN THERMAL-CONVECTION
- LOOPS WHICH CIRCULATED NaF-ZrF -UF4 (50-46-4 mole %)

 

 

 

 

 

 
  
  
   
   

 

L ol Eluorid Time of Final Chromium  Maximum
o Originat Fluoride Hot-Leg Charge  Operation Concentration Attack Intensity of Attack
No. Mixture Batch No. .
_ ' (hr) (ppm) (mils)
540 188 PF-1 None 500 520 |8 Moderate to heavy
{control)
541 188 PF-1 Positive 500 550 7.5 Heavy
542 188 PF-7 Negative 500 675 7 Heavy
554 188 PF-7 Nore - 500 635 11 Moderate to heavy
(control)
- 614 248 PF-4 None 1000 875 15 Heavy
(control) '
615 248 PF-4 Positive 1500 720 12 Heavy
616 248 PF-4 Nggaflve 1500 1200 20 Heavy
617 248 PF-4 None 2000 950 19 Moderate to heavy
(control)
552 217 PF-5 Positive 2000 610 7 Moderate to heavy
553 217 PF-5 | Negative 2000 740 15 Moderate to heavy
TABLE 5.7. MASS TRANS FER IN INCONEL THERMAL-CONVECTION L.GOPS WHICH
. CIRCULATED SODIUM WITH ADDITIONS OF N0202
Loop N0202 Added Operating Time Hot-Leg Attack Relative Mass
No. (wt %) (hr) (mils) Transfer*
729 ' ane 500 1.5 Trace
70 Nene o s10% s Trace
731 “Neme 2000 | 0.06
' o 432**"" 2 Trace
e T 1000 R 2 0.03
‘ “0\-2/“ Ten e 3]3** S o 2 0.04
o2 ’"'1000_'_( '3 0.05
500 ‘ 3 ' 0.02
S8 05 w0 3 ~0as

 

*R

elative figures obtained by weighing the metal removed from a known length of the horizontal hot leg and con-

verting the value obtained to an average weight per inch,

**T

erminated by leak.

103

 
 

"ANP PROJECT PROGRESS REPORT

from the lower vertical cold leg and the horizontal
hot leg of each loop. The deposit in each section
was brushed out and weighed, and the weight
value was converted to an average weight per
_inch, These values, given in Table 5.7, are
‘relative and subject to considerable error. They
do show, however, increased mass transfer with
increased oxide content.

| The increased mass transfer was also reflected
in increased otfock in the hot legs. The depths
of attack varled from 1.5 mils in the control loops
‘to 4 mils in the high-oxide-content loops. The
cfl’ack was primarily intergranular, with, possibly,
some general surface removal. Some difficulty
was encountered with the loop supports during
_this series of tests, and operation of several
- loops had to be terminated because of leaks.
'VThIS dlffaculty now appears to have been corrected.

GENERAL CORROSION STUDIES

' E. E. Hoffman W. H. Cook

| C. F. Leitten, Jr,
Metallurgy Division

R, Carlander
Pratt & Whitney Aircraft

Hot-Pressed ietal-Bonded Tungsten
Carbide in NaF-ZrF ,-UF
W. H. Cook
| Metallurgy Division
Several Haynes Stellite Company, experimental,
hot-pressed, metal-bonded, tungsten carbide
specimens have been screen tested in NaF- ZrF UF

(53. 5-40-6.5 mole %) and in sodlum in the seesaw

apparatus at 4.25 cpm.
were at 1500 and 1200°F, respectively, Each
specimen was held in the hot zone of its capsule
during the 200-hr test period. The nominal
compositions of the materials tested are given in
Table 5.8.

Metallographic examination of the untested and
tested specimens did not reveal any measurable
attack on any of the tested specimens. It did

show that the structure was, in general, good,

The tungsten carbide particles were small, and
the specimens had little porosity, The tungsten
carbide and metal distributions were good, with
the exception that there were small isles of metal
that was free of tungsten carbide particles in all
specimens, These isles were few and small in
the specimens that had more than 20% metal.
Typical untested and tested 88% tungsten carbide—
12% Hastelloy C specimens are shown in Fig. 5.8,

Solid-Phase Bonding of Cermets

W. H. Cook
Metallurgy Division

Recheck tests have been made on several cermet
pairs in order to evaluate better the solld-phase
bondmg results obtained in previous screening
tests,> The cermets tested were manufactured
by Kennametal, Inc., under the trade name
Kentanium. The results of the recheck, given
in Table 5.9, are the same as those obtained
in the previous tests,

 

3E. E. Hoffman, W. H. Cook, and C. F. Leitten, Jr.,
ANP Quar, Prog. Rep. June 10, 1955, ORNL-1896, p 96.

'_TABL'E 5,8, NOMINAL COMPOSITIONS OF SEVERAL EXPERIMENTAL, HOT-PRESSED,
Ll TUNGSTEN CARBIDE CERMETS '

 

Nominal Compesition (wt %)

 

 

- A?im'l Brin’der . wC Co Ni Cr Mo W _ Fq . C
- ,Hasfenoy c 88 6.8 2.0 2.1 0.5 o 7002 -
 Haynes Alloy No. 31 88 67 13 3 09 o 2 Toes
90% Co-10%Cr 84 14.4 1.6 I
"Haynes Alloy No. 6 84 10.3 4.4 0.6 05 06
Haynes Alloy No. 6 76 15.5 6.6 10 07 024
" Haynes Alloy No. 6-8% Co 80 15.7 3.3 05 04 000
" Haynes Alloy No. 6-12% Co 76 19.7 3.3 0.5 04 000

 

104

The hot anci cold Jzon-es |

 

 

oL DR

 

 
PERIOD ENDING SEPTEMBER 10, 1955

 

 

 

 

 

 

UNCL ASS?F]ED
Y- 15450
4
3
%
oy

] Fig. 5.8, (a) The 88% Tung.'sten Curbi&e—"lZ%r Hastelloy C Cermet Before VTesfing. (b) The Same
ic ' Specimen After Exposure for 200 hr to NaF-ZrF -UF , (53.5-40-6.5 mole %) in the 1500°F Hot Zone of
' a Seesaw Apparatus in Which the Cold Zone Was at 1200°F. Specimens unetched. 1000X. Reduced 2%.

 

 

      
 

 

     

 

      
  
  

   

 

  

   

            

 

 

 

TABLE 5.9. RESULTS OF RECHECK TESTS OF 'SQLID-P‘HA‘SE‘BONDING OF SEVERAL CERMETS EXPOSED
TO NaoF-ZtF 4-1.;‘|=4 (53.5-40-6.5 male %)' AT 1500°F FOR 100 hr AT 50,000 psi*
Composlhons. K]50A (80 wt % TIC 'IO wf % NbTaTIC ~10 wt % NI)
. . g el KISIA (70 wt % TIC 10 wt % NbTaTnC ..20 wt % Ni)
]5?8 (64 wf % TlC-—6 wf % NbTaTlC -30 wt % N.)
* K]é?B (64 wf % T|C~—6 wi % NbTuTIC -25 wt % Nl--5 wt % Mo)
L K150A vs K152 D | Some
: 7 KISIAwsKIS2B " Nonme
KisBvskie2B " Nome
. K162B vs ‘ - T Some
*Cale u'[_ar'rea ccnta ct lp}ess't;rfié.' ' \
i
' 105

-

 
 

 

 

 

ANP PROJECT PROGRESS REPORT

Further evaluuhon of fhe best of the nonbondmg
cermefs, as determlned by the original and re-

checkmg tests, will be made in solid-phase

'bondlng tests of these materials in the form of
valve disks and seats, It is planned to study the
resistance to sohd-phase bonding for long periods
~ of time, 1000 hr or more; the effects, if any, of
“braze-joining the cermets to Inconel (cf. Sec. 6,
“‘Metallurgy and Ceramics’); and, possibly, the
o effec’rs of repeofed seahngs of fhe dlsks und seats,

Effects of Ruihemum on Physncul
PrOperhes of lnconel

“C.F. Lelh‘en, Jr.
MetaHurgy Dlv:smn

 

Slnce the examlnanons of vorlous sechons of

‘the ARE4 and the LITR fluoride-fuel loop5

revealed slight deposits of ruthenium metal on
the walls of the Inconel tubes, the effect of
ruthenium on the physical properties of Inconel
is being studied. A thin layer of ruthenium metal
was electrodeposited from a solution of ruthenium
nitrosochloride on an Inconel creep-test specimen,
The thickness of the ruthenium plate on the
Inconel was approximately 1.5 mils. The plated
creep-test specimen was annealed in an evacuated
Inconel capsule for 100 hr at 1500°F to allow the
ruthenium to diffuse into the specimen. The
specu'nen was then glven a creep-rupfure test in a
purified argon atmosphere at a stress of 3500 psi.
In calculating the stress, the area of the ruthenium
plate was included. The results of the creep-
rupture tests on the ruthenium-plated Inconel and
on an unplated, standard, Inconel specimen are
presented in Table 5.10, Since a difference in
rupture times was found, another test is now in
progress to check the results,

A metallographic examination of the strained
portions of the ruptured rutheniumeplated Inconel
specimen showed no difference in microstructure
in comparison with that of a ruptured, unplated,
Inconel specimen (Fig. 5.9). Creep-rupture tests

~are also to be conducted on ruthenium-plated

Inconel specimens in the fused-salt mixtures,

 

S 4M. T. Robmson, S A. Reynolds, and H, W, Wr:ght,
ANP Quar, Prog. Rep. Mar. 10, 1955, ORNL-1864, p 14.

SM. T. Robmson cmd T H. Handley, ANP Quar Prog.

o Rep ]une 10, 1955, ORNL-1896, p 167.

i

" TABLE 5.10. COMPARISON OF CREEP-RUPTURE
DATA ON RUTHENIUM-PLATED AND
o UNPLATEDINCONEL ‘ '

Stress: 3500 psi
Test temperature: 1500°F

Test environment: Argon

 

 

Time to Final Creep
Specimen Rupture Elongation Rate
o () L) (%/hr)
Plated 873 13 00145
Unplated 1467 12 0.0028

 

Brazing Alloys on Inconel and on Nickel in

-Sodium and in NaF-ZrF4-UF4

C. F. Leitten, Jr.
Metallurgy Division

Seesaw tests have been completed on a series
of brazing alloys on Inconel T-joints. These

 tests were conducted in sodium and in the fuel

mixture NaF- ZrF -UF, (53.5-40-6.5 mole %) for
100 hr at a hot-zone temperai‘ure of 1500°F. A
temperature differential of approximately 400°F
was maintained between the hot and cold zones
of the test container, The data obtained from
weight measurements and metallographic ex-
amination of the tested T-joints are given in
Table 5.11. The 82% Au-18% Ni brazing alloy
and copper were tested only in the fuel mixture,
because previous corrosion data had indicated
the poor corrosion resistance of these alloys in
sodium, The seesaw test on the Coast Metals
No. 52 brazing alloy was a retest; the results
agree with those previously reported.6

The Coast Metals No. 53 brazing alloy and
Electroless nickel showed good corrosionresistance
to both mediums, as indicated in Table 5.11. The
Inconel T-joints brazed with Coast Metals alloy
No. 53 are shown in Fig. 5.10 after exposure to
the fue! mixture NaF- ZrF UF (53 5-40-6.5 mole %)
in a seesaw test for 'IOO hr af a hot-zone tempera-
ture of 1500°F and after exposure to sodium under
the same conditions. Only slight attack can be
seen along the surface of the braze fillet exposed
to the fuel mixture; the fillet exposed to sodium
showed similar slight surface attack.

 

6E, E. Hoffman, W. H. Cook, and C. F. Leitten, Jr.,

ANP Quar, Prog. Rep. June 10, 1955, ORNL-1896, P 98.

 

 

 

 
 

e

"

 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

" UNCLASSIFIED
o Y-15992

*

£
i
|

The strfiinéd Portions of (a) a Ruthenivm-Plated and (b) an Unplated Inconel Test Specimen

Fig. 5.9.
Following Rppibrg in o Creep Test. Cathodic etch, 250X,

 

 

UNCLASSI — 1 . UNCLASSIFIED
¥-1522 o Y-15228

 

 

 

 

 

 

 

 

         

- Flg. 5.10. Inconel T-Joints Brazed with Coast Metals Alloy No. 53 After Seesaw Testing for 100 hr
in (a) Fuel Mi‘xturé NOF-Z?F#-UF4 (53.5-40-6.5 mole %) and (b) Sodium at a Hot-Zone Temperature of
1500°F, E’rch.ed'wi'rh aqua regia. 100X. Reduced 20%.
107

 

Ll iiods

 

 
ik

ANP PROJECT PROGRESS REPORT

TABLE 5 11. RESULTS OF SEESAW TESTS OF BRAZED INCONEL T-JOINTS TESTED IN SODIUM
AND IN THE FUEL MIXTURE NaF-ZrF, -UF (53.5-40-6.5 mole %) FOR 100 hr ATA

 

 

 

1 - LT  HOT-ZONE TEMPERATURE OF 1500°F - .
. -'_- o ies Welght Change | - ’7;-
Brazing é!loy* Composnhon Bath —m™8 ¥ —— Metallographic Notes b
Cwew (9) (%)
- Coasf Metals No. 52, Fuel o 0-‘- V Umform surface attack along f1||e1’ to a depth T
o 89 Nl-—-5 SI-4 B-—-?. Fe of 0.5 mil
_ .Copper ’ Fuel ~0.0602 -—0.026 ‘ Surfoce attack along fillet to a dej;fl'l o? 70.5 mil _ | ' o
o Coqst Metals No. 53, Fuel —~0.0011 "-0.092 Nonumform surface attack along f:i!ef To a depth'
g S ' 81 Nl—4 5!-4 B 8 T of 1.5 mils ' W E
: ’ Erratic attack along surface of f:llef fo a depth 4 ’

c.-_3 Fe Sodium  —0.0009 —0.071

of 1 mil

 

No attack along surface of fillet

Electroless nickel, Fuel  —0.0004 —0.041
90 N:-IO_P“ |

   

Sodium  —0.0044 —0.50 _ 'Surface attack along f:llet toa depth of 'I.5 mlls .

©Coast Metals No. NP,  Fuel  —0.0009 —0.092

Unlform attack along surfuce of hiie'r to a depth )

 

 

50 Ni~12 Si—-28 Fe—4 of 1.5 mils 7 _ _
Mo—4.5 P=1 Mn=0.5 Cr  Sodium  —0.0069 —0.622 Uniform attack along surface of fillet to a depth
of 2.5 mils '
Genera! Electric No. 81, Fuel -0.0008 ~0.067 Attack along surface of braze fillet to o depth of
66 Ni=19 Cr=10 Si—4 3.5 mils .
Fe~1 Mn | Soedium  —0.0018 —0.0163 Uniform attack along surface of fillet to a depth
of 3 mils
82 Au-18 Ni Fuel 0.0011 0.12 Nonuniform attack along fillet to a depth of

4 mils

 

*Brazing alloys listed in order of decreasing corrosion resistance to both test mediums.

 

o G

G e e e o L e e e R G ;

 

The Electroless nickel (90% Ni-10% P) alloy
was unattacked by the fuel mixture; however,
there were several microscopic cracks present
in the fillet. These cracks indicated that this

alloy was brittle. It would therefore not be

sahsfactory for use in radiator fabrication, The

82% AU—TS% Ni brazing alloy will be retested in

‘_'rhe fuel mlxture because there is considerable
dlscrepancy ‘between the results of this seesaw

test and the results of the static test.”
The 75% Ni-25% Ge brazing alloy on nickel

T-joints has been tested in static sodium and in

~ the fuel mixture NaF-ZrF 4-UF , (53.5-40-6.5 mole %)
S for ]00 hr at 1500°F Thls alloy showed non-

 

7E E Hoffmcm, ANP Quar Prog. Rep. Dec. 10, 1954,

L 'I'.ORNL 1816, p 83.

 

uniform attack to a depth of 2 mils in sodium and
no attack in the fuel mixture. This alloy will be
seesaw tested in sodium and in the fuel mixture,

Brazing Alloys on Inconel and on
Stainless Steel in Lithium

C. F. Leitten, Jr.
Metallurgy Division

Corrosion tests have been completed on three
brazing alloys — Nicrobraz, 73.5% Ni—10% Si—16.5%
Cr, and 71% Ni-16.5% Cr-10% Si-2.5% Mn — on
type 316 stainless steel and Inconel T-joints in
static fithium at 1500°F for 100 hr to verify
corrosion results previously obtained,

The results obtained from weight measurements
and metallographic examination of the tested

 
 

 

‘:,::Nicrobraz,

  
  
 
 

specimens are presented in Table 5.12. All
the brazing alloys exhibited poor corrosion
resistance to lithium, and the attack was more
severe on the brazing alloys on type 316 stainless
steel. This is probably due to the difference in
nickel concentrations in type 316 stainless steel
(10 to 14%) and Inconel (75 to 80%). Since nickel
is preferentially attacked by lithium, the attack
on the Inconel T-joints would be uniform on both
the base material and the brazing alloys, which
in these tests contained approximately the same
nickel concentrations as Inconel. However, the
attack would be more concentrated on the brazing
alloys on type 316 stainless steel T-joints, be-
cause most of the nickel would have to be leached
from the brazing alloy in order to reach the
solubility limit of nickel in lithium. The results
of these tests corroborated the results of the
previous tests; that 'is, the nickel-base brazing
alloys have very poor corrosion resistance to
lithium, especially when used to braze iron-base
alloys.

Hastelloy B-Inconel in NaF-ZrF,-UF,

R. Carlander
Pratt & Whitney Aircraft

A static test of a Hastelloy B specimen in an
Inconel capsule containing a fuel mixture was
petrformed to determine whether dissimilar-metal

PERIOD ENDING SEPTEMBER 10, 1955

mass transfer occurred in such an isothermal
system. The fuel mixture was NaF-ZrF,-UF,
(50-46-4 mole %), and the system was held for 100
hr at a temperature of 1600°F. The Hasteiloy B
specimen showed a negligible weight loss of
0.0003 g, but no attack could be detected
metallographically,  Spectrographic analysis re-
vealed that no molybdenum had transferred to the
Inconel capsule wall.,

Two 5-mil cuts were machined from the surface
of the Hastelloy specimen and analyzed spectre-
graphically to determine whether chromium had
been picked up from the Inconel tube. Since
there was no detectable difference in the chromium
content of the two cuts, it appears that no ap-

~ preciable quantity of chromium had transferred.

The surfaces of the Hastelloy B specimen and of
the [nconel container are shown in Fig, 5.11.
Inconel is normally attacked to a depth of 2 mils
in an all-lnconel static test system under similar
test conditions. The presence of Hastelloy B
in the system increased the observed attack on the
Inconel to a depth of 8 mils.

Boiling Sodium in Inconel

E. E. Hoffman
Metallurgy Division

Tests, to date, at ORNL and at other laboratories
indicate that the oxygen content of sodium in

TABLE 5.12. BRAZING ALLOYS ON TYPE 316 STAINLESS STEEL AND INCONEL TESTED
IN STATIC LITHIUM FOR 100 hr AT 1500°F

 

Brazing-Alloy Composition
Cwt o

Base Material

@

Weight Change*
Metallographic Notes

 

 

Type 316

 

70 Nl—]4 Cr-6 Fe-—-S B 4 Sl—] C

< Incone[

U 73SNIS10SiZ165C 0 Typedls

stainless steel

71N'"]0 S$i—16,5 Cr—2.5 Mn - Type 31:6:}

   

Inconel

 
 

stclniess sfeel

 

.0 'stainless steel

“ 17 Joint failed during testing

 

- =0,16  Joint o'fl"aék'eawno'fiuniformly to a
posl 0 maximum depth of 9 mils, with

- uniform attack to a depth of 4
‘ 'miis over enhre fnllet ,

20,0235 7 L1297 | VJomf falled durlng feshng

   

Braze h”ef complefe!y aflocked‘ '

 

-0.0949 --1 1.3 7Brcze f|l|ef comp]e.fely afl'acked

 
     

0.0312  3.66  Unifomn attack in form of sub-

surface voids along fillet

 

 *Weight-change data for brazing alloy and base material of joint.

 

109

T o

-

 
RN

 

. Sl .. Wl i e o il S 3 N RS G e, Sl B b, AL . b e et s ot S e S . st o s i

 

" ANP PROJECT PROGRESS REPORT

 

 

 

 

 

 

 

 

 

 

 

Fig. 5.11. Surfaces of (a) the Enconel Container and (b) the Hastelloy B Specimén Followmg Ekp.bsure

to NoF-ZrF4-UF4(50-46-4 mole %) for 100 hr at 1600°F

duced 7%.

nickel or in iron-base alloy systems has a con-
siderable effect on the amount of mass transfer
that occurs in the system. Therefore, in order to
study mass transfer in a system in which the
oxygen content could be held to a very low level,
a boiling-sodium loop test was run. The loop is
shown in Fig. 5.12. The temperatures around the
loop during the 400-hr test are indicated. There
are two traps for liquid sodium in the condenser
leg of the loop. The first trap is filled with hot,

freshly condensed sodium during the test. This
.is an ideal location for metal solution to occur.
The second trap catches the overflow from the
first trap and operates at a considerably lower
‘temperature.” This is an ideal location for metal
‘deposition to occur. Three nickel cooling coils

are located on the condenser leg, and air flow

- “through these coils is regulated to maintain the
‘desired temperatures. The boiler and the small
_return line to the boiler are the only areas where
heat is applied. The first test was terminated
~ after 400 hr, when a small leak was detected i
the coolest sechon of the bottom return |1ne.

X-ray, mccroscoplc, and microscopic examination

o

Etched with 10% oxalic acid. 500X. Re-

UNCLASSIFIED
ORNL~LR-DWG 8363

CONDENSER LEG

RECEIVER
420°C
SODIUM LEVEL

DIFFUSION COLD
TRAP

 

10 2 4
bt

INCHES

Fig. 5.12. lInconel-Boiling-Sodium Loofi.

of the traps in the condenser line revealed no
mass fransfer (Fig. 5.13). Heavy intergranular
cracks were found in the condenser tube wall, and,
as yet, no satisfactory explanation has been

found. The extent of this intergranular cracking

-y

 

 

 

 

 

 

i

 
1o

e/

o

 

of fh:s experlmem rs' now under way.

"lncone_l fubes loaded wit 1
lithium mixtures in which the lithium content was
‘varied from 2 to 30_wt ‘%.” The ratio of sodium to

 

PERIOD ENDING SEPTEMBER 10, 1955

 

 

Fig. 5.13. |hcbhél:-‘rB\-bi'li‘ng--Sodi'um Sys;temVSpecirhens from (a) Hot Trap and (b) Cold Trap of Inconel
Condenser Tube Wall. Note heavy intergranular cracks and absence of mass-transfer crystals. 100X,

Reduced 11.5%.

varied from zero in the hottest section of the  in the presence of 5% lithium. It is not yet under-
condenser tube to a depth of 3 mils in the coolest  stood why the heavier attack occurred in the
section, with the heaviest cracking, to a depfh presence of the lower amount of lithium. Additional
of 50 mlls, midway between the two traps  specimens cut from the hot zone of the tube
'(Flg 5, 13) A secbnd tesf' for chec|<mg the resulfs confirmed the results. In none of the tests could

   

 

mass fransfer be detected in the cold zones of the
‘tubes, The results of meta!lographlc examination
. of the hot and cold zones of each test container
are presented in Table 5.13, and the hot zones of
the tubes tested with 2 and 30% lithium mixtures
'cre shown in Flg 5 14.

Mblybdenum, Yanadium, and Niobium in
Static Lithium
E. E. Hoffman

Metallurgy Division

 
   

Flve seesaw corrosion tesfs were conduc'red on
sodlum-potasswm-

        
  

    

  

       

potasswn? in each test was 56:44, The heawésf

attack in each test was found in the hot section  Corrosion tests have been conducted on molyb-
of the Inconel seesaw tube, and it varied from denum, vanadium, and niobium exposed to static
0.5 mil in the presence of 10% lithium to 2 to 3 mils fithium for 100 hr at 1500°F. The test specimens

111

T

i

i

 

 

 

 
i

G ool S

 

 

G R

 

ANP PROJECT PROGRESS REPORT

TABLE 5.13. RESULTS OF TESTS OF INCONEL SEESAW TUBES EXPOSED TO
| SODIUM-POTASSIUM-L!THIUM MIXTURES FOR '!00 hr o

 

 

 

 

 

 

Hof-zone temperoture.: 15000F 4 ’ ’ - » o 3
Cold-zone temperqture. 1100°F
Seesaw Test ~ Bath H Attack Cold.z Attack
Number o (wt %) ot-Zone Attac old-Zone Attac
L 7265 2Li-55Na-43 K Intergranular veids in scattered No attack
T SRR areas to a depth of 1 mil
266 5Li-53 Na—-42K Héavy attack in fHe forrfi of sub- ' No attack | ) *
. . , " surface voids to a depth of 2 E
to 3 mils; small, very hard par-
ticles of an unidentified phase
located in areas where the ‘ v
, S - - cflock ‘was heavsest _
270 10 Li-50 Na—40 K Small subsurfcce vmds to a No attack
_ i sl _ depth of 0.5 mil
271 | 20 Li—45 Na—35 K Small subsurface voids to a " Attack in scaflefed areas to
_ ~ depth of 0.5 to 1 mil a depfh of 0.5 mil
272 30 Li—40 Na—30K Subsurface voids to a depth of lntergranu[ur pene'h'aflon to a
: o S 1to 2 mils depth ofVO 25 to 0.5 mil

 

 

0.001

 

10.002

 

0003

 

0004L“"

INC

 

0.005

 

0.006

 

 

 

 

0.007

 

 

 

 

~ Fig. 5.14. Surfaces of Hot-Zone Specimens of Inconel Seesaw Tubes Exposéd to Sodium-P.dt;cjissiuwrfi-‘-’-
Lithium Mixtures for 100 hr ot a Cold-Zone Temperature of 1075°F. Tube (a) contained 55% Na~43% K-
2% Li; tube (b) contained 40% Na-30% K-30% Li. 500X. Reduced 3%.

112

 
a

 

 

and containers were of the same material in each
test. The results of weight-change determinations
and metallographic examination of the specimens
from these tests are given in Table 5.14, The
exposed surfaces of the niobium and the molyb-
denum specimens are shown in Fig. 5,15, The
niobium was only slightly attacked, and the molyb-
denum was unattacked,

-c

Analysis of Metal _Cl"ysfais.' from Inconel-Sodium
 Thermal-Convection L.oop
E. E. Hoffman
Metallurgy Division .
The metal crystals found in the cold leg of an
Inconel-sodium  thermal-convection loop® were
analyzed and found to be:

Content
Element
| , . (wt %)
. ' Ni 85.0.
' - Cr ‘ 11.2

Mn - 14

 

BE. El Hoffman, W. H' Cook, Gn(i C.F. Leifl‘en, .Jro,.- -

' ' ANP Quar, Prog. Rep. June 10, 1955, ORNL-]‘896, p 101.

   

NCLASSIFIED .
Y.15552

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 5.14. RESULTS OF EXPOSURE OF MOLYB-
DENUM, VANADIUM, AND NIOBIUM TO STATIC
LITHIUM FOR 100 hr AT 1500°F

 

Weight Ch
Material °19 ange

 

i hic R
(g/imz) Metallographic Results
Molybdenum 0 No attack
Vanadium 0.0084 Grain-boundary penetra-
tion by unidentified
phase to a depth of
2 mils
No attack

Niohivm 0.0006

 

  
   

 

 

 

.ok

 

 

 

 

 

for 100 hr ot 1500°F.

 

 

FUNDAMENTAL CORROSION RESEARCH

G. P. Smith
Metallurgy Division

Film Formation on Metals

J. V. Cathcart
Metallurgy Division

Ok‘idufion of Sodium. The investigation of the

oxidation characteristics of sodium has been con-

 

 

 

 

 

 

 

Fig. 5.15. Surfaces of (a) Niobium and (b) Molybdenum Specimens After Exposure to Static Lithium

113

 

 

 
 

Gkiii

ik ki

 

 

 

 

ANP PROJECT PROGRESS REPORT

tinved. Previous results? emphasized, contrary
to generally accepted oxidation theory,}0 that
sodium forms a highly protective oxide film in dry
oxygen. The research has, also, in some measure,
filled the gap which exists in the experimental
ox;dcmqn data for the alkali ond alkaline-earth
metals. Finally, the results indicate a consider-
able similarity between the oxidation of sodium
and other, more extensively studied, metals, such
as copper and aluminum. In this respect, addi-
tional evidence has been obtained in support of the
idea expressed by Cabrera and Mott!! that all
metals have comparable oxidation mechanisms at
relatively low temperatures,

OXidotion-fate measurements for sodium were
made manometrically. A film of sodium was
evaporated onto the walls of a small glass bulb
attached to one arm of a sensitive differential
manometer, The course of the reaction could
then be followed by measuring the decrease in
pressure in the bulb as the oxidation proceeded.

- The details of the experimental apparatus have

already been described.?
During the period covered by this report, most of

-the experimental work has been devoted to the

measurement of oxidation rates at 25 and 48°C
over extended periods of time. At these two tem-
peratures and also at 35°C, experiments lasting
between 15,000 and 20,000 min have been com-
pleted. |

Such prolonged experiments were considered
necessary for two reasons. First, they are re-
quired if a realistic determination is to be made of
the analytical form of the rate curves, Because
there may be a gradual change in the reaction
mechanism as oxidation proceeds and on account
of unavmdable errors in experimental measure-
mem‘s,' a misleading impression of an oxidation

curve can easily be obtained when data for only
‘the first portions of the curve are available.

A second reason for such extended tests was
that some metals (for example, columblum) undergo

‘abrupt changes in oxidation mechanism after o

relatively thick oxide film has developed on them.

 

 

9). V. Cathcart, ANP Quar. Prog. Rep. June 10, 1955,

-ORNL.-1896, p 105.

- 10y, B, Pilling and R. E. Bedworth, ], Inst, Metals
29, 529 (1923); see also U. R. Evans, Memllzc Corrosion,
Passivity, and Protection, p 102, Longmans, Green and
Co., New York, 1948.

1IN, Cabrera and N. F. Mott, Repts, Progr. in Phys.
12, 163 (1949).

114

It was important to determine whether or not sod:um
‘is subject to this phenomenon,

Seven rate determinations at 25°C and four at
48°C, all lasting for 17,000 to 20,000 min, have
been completed. Data from these, as well as the
measurements obtained at ~79, =20, and 35°C,
are shown in Fig. 5.16. The change in pressure
in the system, which is a function of the oxide-
film thickness, is plotted against the time of
oxidation on a log-log plot. Each unit of the
vertical scale is equivalent to approximately 10 A

of oxide; therefore after 20,000 min at 48°C an

oxide thickness of almost 1400 A was attained.

Each curve of Fig. 5.16 represents the com-
posite of data from three or more experiments,
Deviations of as much as 10% in the total pressure
change were observed between readings at the
same time and temperature in different experiments,
An error of this magnitude was not considered
serious; a basic assumption in comparing the re-
sults of different experiments was that the areas
of the initial sodium films were the same. A new
oxidation bulb was used for each experiment, and
it was impossible to avoid slight variations in the
dimensions of these bulbs. Therefore the assump-
tion that the apparent surface area of the sodium
films was identical for all experiments was true
only to a first approximation, Perhaps even more
important was the possibility that the surface
roughness (and, therefore, the true surface area) of
the sodium films differed slightly from experiment
to experiment. These two factors could easily
account for the observed variations in the experi-
mental results,

Mathematical analyses of the rate curves show
that none of the conventional oxidation-rate laws
was obeyed by sodium; that is, the data could not

- be fitted to rate equations of the form x™ =kt (where

x is the oxide thickness after a time ¢, & is the
rate constant, and » = 1, 2, or 3) nor to the loga-
rithmic, x = & log ¢, or the inverse logarithmic,

1/x = A - B In ¢, rate equations. The absence

of any abrupt changes in the oxidation curves
indicates that there were no sudden changes in
the mechanism of oxidation as the oxidation pro-
ceeded,

One interesting feature of the results was that
the oxidation proceeded at an ever-decreasing
rate with time, even at 48°C., Thus highly pro-
tective oxide films were observed under all con-
ditions investigated., There was no indication of

 

 

 
 

 

 

Ap (mm OCTOIL)

PERIOD ENDING SEPTEMBER 10, 1955

UNCLASSIFIED
ORNL-LR-DWG 9294

 

103 2 5 104 2 5 10°
TIME {min)

'Fig. 5. 16. oxiaafion' of Sodium at 79, =20, 25, 35, and 48°C.

any fendency toward crack formcmon in the ox:de.
~ Far from obeymg a linear rate. |cxw, as has been
' ‘predlcied the‘orehcozl_ly,lo the oxldchon charoc-

    
  
  
  
  
  
 
        
 

   

and of the nucleahu
‘ 'Vsodlum.

 

125, T. qufhmey, F. W. Young, Jr,, and J, V.
Cathcart, Acta Metallurgia, in press.

"rhe Temperature ronge -—79 to

 of the oxide films formed on

The oxidation of co-
lumblum follows a linear rate “equation above
‘400°C Below fh:s femperofure the rate curves
_are approxumateiy parabohc. n a narrow ’rempem-
' ture band cem‘ered around 400°C the rate curve

Oxidcfl.lon of Columbium.

tS _u_uhqlly purabollc, but 11 become_s almost

13

Th’e'only theorehcol bas:c for c llnear omdchon-
rute equchon is that glven by
However, the resulfs obtamed for

dequucy "of this fheory An mveshgohon
“of tfie”oxlldahon of columblum at temperatures in
‘the wcmn‘y of 400°C was undertaken in the belief
. that a careful sfudy of ’rhe s’rructure composnhon,

 

134, Inouye, Scaling of Columbium in Air, ORNL-1565
(Sept. 1, 1953),

115

Pilling and

 
 

 

 

"ANP PROJECT PROGRESS REPORT

‘and behavior of the oxide films during the transition

from the parabolic to the linear rate would give
some mdlcchon of the criteria for linear oxidation
rates,

The first phase of the work consisted in a

careful redetermination. of the oxidation-rate
curves for columbium at 400°C. A manometric
technique, very similar to that used for the
“oxidation of sodlum,9 was found to be satisfactory
for the rate measuremenis, The oxidation speci-
mens cons:sted of 1 x 2 e¢m strips cut from a
0.020-in.-thick sheet of columbium. The specimens
were electropolished mechanically with 4/0 emery
paper and then electropolished in a 10% HF-90%
H SO bath. One such strip was placed in a
100-cm glass bulb attached to one arm of a
dlfferenhal manometer,  Octoil-S diffusion-pump
oil was used as the manometric fluid in order to
' obtain the desired pressure sensitivity. The
oxidation and reference bulbs of the manometer
were evacuated to a pressure of about 10=6 mm Hg
and then immersed, along with an oxygen reservoir,
in a molten-salt, constant-temperature bath. The
oxidation and reference bulbs were then filled
simultaneously with oxygen to identical pressures.
The absolute oxygen pressure was about 700 mm
Hg. The rate of oxidation was then determined by
measuring, with the manometer, the decrease in
pressure in the oxidation bulb. The use of an
accurately controlled constant-temperature bath
was considered necessary to avoid spurious
pressure changes arising from temperature differ-
ences in the oxidation and reference buibs,

Four oxidation experiments have been completed
at or near 400°C. In each case the oxidation
curve was initially approximately parabolic and
then became linear as tHe time of oxidation
increased. The data from one of these experiments
is shown in Fig. 5.17. The change of oxygen
pressure is plotted against the time of oxidation,
For reasons as yet unknown, the time at which
‘the transition from the parabolic to the linear

portions of the rate curves occurred varied some-

what from experiment to experiment. However, the
amount of oxygen which had been consumed by the
‘time the transition region was reached was found
to be relatively constant.

The oxide layers formed during the early stages
of the reaction exhibited the interference tints
typical of thin, transparent, oxide films. However,
as the reaction progressed, patches of a white,

116

" UNCLASSIFIED
ORNL—LR—DWG 9292

240
220
200

Ap {mm OCTOIL)
» O B B g ®
S 538883588

 

<

0 100 200 300 400 500 GOO_ 700
TIME (min)

Fig. 5.17. Oxidation of Coldrfi-b.i.u;t.r ut- 4006C

opaque material developed on the surface of the
columbium specimens, These white areas ap-
peared under an optical microscope to be rough,
and it was assumed that they represented relatively
porous regions in the oxide film, X-ray ex-
amination of the oxidized specimens indicated
that both the white areas and those parts of the
oxide film that still showed interference colors
consisted of columbium pentoxide (Cb 205).  No
other oxides of columbium were detected,

Although the evidence is still not complete,
the occurrence of the regions of white oxide
seemed to be associated with the deviation of the
rate curves from their initial parabolic form. For
this reason it has been tentatively assumed that
the white portions of the oxide film represent the
type of oxide film which produces a linear reaction

rate. In order to study the nucleation and growth

of this white oxide, a hot-stage microscope has
been constructed and is being tested. This

apparatus should allow direct observation of the’

oxide films as they are formed.

An attempt is being made to express the over-all
oxidation-rate equation for columbium as the sum
of parabolic and linear terms approximately
weighted to take into account the fraction of the
total surface area oxidizing according to each of
the two mechanisms, Complete formulation and
testing of this equation awaits further data
concerning the growth characteristics of the

 

 

 

 

 

 

 
 

0

”

 

  
 
   
  

areas of white oxide. It is hoped that this infor-
mation can be obtained from observations of the
oxidation under the hot-stage microscope.

High-Temperature S'-peci'rophcromeiry in
Fused Hydroxides

C. R. Boston
Mefullurgy Division

Work is continuing on the improvement of
techniques for measuring absorption spectra at
high temperatures. The major components have
been fabricated and assembled for a new cell
arrangement, These include the MgO cell ('“-'7/ in,
deep with parallel flat sides); the stainless sfee!
cell holder, tube header, etc.; a quartz tube with
optically flat windows; a glass vacuum system;
and a furnace with light-path openings. A controlled
atmosphere can now be maintained over the melt
during spectral measurements, and, in addition,
the cell can be dry-box-loaded with previously
dehydrated hydroxide. This prevents etching of
the polished MgQ cell faces by water vapor, as
was experienced with the previous apparatus.
It was thought that the deeper MgO cell would
reduce the creep problem, However, with the cell
less than half full, the hydroxide was found to
creep over the top of the cell and down the sides
in a matter of minutes after melting. A still-
deeper MgO cell (~1 I/ in.} is now being cut.

At the present hme, a Coleman model 14
spectrophotometer modified for high-temperature
work is being used. This instrument is, of course,
inadequate for accurate work, and a Cary model 11
instrument is being acquired and will be rebuilt,

- The h:ghes’r purlfy sod:um hydroxlde commercrolly
o avcrlable s Bc_lker s reogent specrcl” grczde
~ which contains

.. NH, OH “and trace amounts of o'rher lmpurmes.

\ThlS product has been further purn‘red” by
L. G Overho[ser to remoy r |

but |f has been confenilna'red_ in 1he”process with

 

1.5% H,0, 0.5% Na,CO,, 0.02%

  

 

robqb Iy a nlc ke I-oxyg

s “on the‘surface 'of 'rhe sohd

q"sodlum hydroxnde ‘and, on meltmg, can be seen « —
on the quuld surfcce. Therefore_wans

ade to prepdre rehcnbly pure:s‘od‘l'um

\hydroxrde by the reaction of pure sodium metal

with water, followed by dehydration in a relatively

" inert ceramic container. The technique employed

‘ “and carboncn‘e, h

  
   

 

PERIOD ENDING SEPTEMBER 10, 1955

was to pass water-saturated (at room temperature)
argon over metallic sodium contained in a Morganite
(Al,0 ;) crucible, After all the sodium had reacted,
the system was evacuated, and the hydroxide
was dehydrated by heating to 500°C ot a pressure
of 10 p. A small amount of hydroxide prepared in
this way from reagent-grade sodium is now being
used in the thermal-decomposition studies. In
these studies the small amount of ammonia present
in Baker’s ‘‘reagent special’’ grade sodium
hydroxide might give misleading results,

Mass Transfer and Corrosion in Fused Hydroxides

M. E. Steidlitz
Metallurgy Division

Studies of the corrosion and mass: transfer of
nickel and Inconel by fused sodium hydroxide are
continuing in the “‘cold finger' apparatus.'4 In
addition, two new metals, copper and Monel, have
been checked under hydrogen, and the results

‘indicate that both are attacked to approximately
the same extent as nickel., The mass transfer on

the Monel cold finger is interesting in that both
nickel- and copper-colored crystals are deposited.

[t has been reported that additions of chromium
in fine particles eliminate mass transfer of
nickel.!5 One addition of chromium as a single
lump, which was the only form available, was
made at 650°C under hydrogen. No change in
amount of mass transfer was noted. Addition of
a manganese sheet to the hydroxide at 650°C
resulted in transfer of manganese to the cold
finger.

The effect of electrlcal contact between the

Teold fmger “and ‘the bucket was investigated.

Under hydrogen, electrically lnsulctlng the two

“Components produced little, if any, reduction in

mdss transfer over the temperature range 600 to

700°C With a helium blanket, however, mass
:"f‘rd'h'safer “at 650°C was substantially increased
~upon breakmg ‘electrical contact. This unexpected

res'ult was rechecked and conflrmed The electrl-

- cal effec'r is |Hustra'red in Fig. 5.18.

ddaifiefimof W&fet \}dpor""f’d the hydrogen

| ond 'ro the hellum reduced mass transfer of nlcke[

 

 

 

iy ”M. E. Steldhtz, ANP Quar. Prog. Rep ]une 10,
1955, ORNL 1896, p 110

]SA F. Foresheri and R. A. Lad Tbe Use of Metallic

Inbibitors for Eliminating Mass Transfer and Corrosion
in Nickel and Nickel Alloys by Molten Sodium Hy-
droxide, NACA-RM-E54L.13 (March 2, 1955).

117

 

|
E

 

 

 
 

 

 

 

"ANP PROJECT PROGRESS REPORT

UNCLASSIFIED
Y-15861

 

- Fig. 5.18. Nickel Cold Fingers Tested in Sodium
Hydroxide with the Bucket at 650°C and a Temperar
ture leferenhol of 100°C and with a Helium
Blunkei‘. Number 95 was electricafly connected
to the bucket, and No. 93 was insulated from the
bucket.

at 650°C. No reduction in the amount of deposited
nickel was noted with water vapor and hydrogen
at 750°C.

The corrosion of Inconel by sodium hydroxide
at 800°C produces an interesting microscopic
structure. The unetched sample shown in Fig. 5.19
has little, if any, grain-boundary attack, the
depth of penetration being marked by laminar
formations within the grains, Samples of 25-mil
strip showed a tendency to split into two thinner
sheets after 50 hr of immersion. A series of these
strips, Nos. 0 to 4, which were tested at 800°C
for 0, 7, 28, 54, and 100 hr, are shown in Fig. 5.20.
The results of spectrographic analyses of these
strips are presented in Table 5.15. Further work
on identification of the material deposited in the
lnconel is bemg planned.

Thermal D|ssocmt|on of Sodium Hydrox1de

‘M. E. Steidlitz
Metallurgy Division

Observahons of oxidation corrosion of metals
plus color chcmges]6 in fused sodium hydroxide

18

" TABLE 5.15. RESULTS OF SPECTROGRAPHIC
ANALYSES OF INCONEL STRIPS TESTED AT
800°C IN SODIUM HYDROXIDE UNDER A
HYDROGEN ATMOSPHERE

 

 

 

Strip Exposure Analysis {wt %)
Time :
No. Ni Cr Fe Mn Residve*
(hr)
0 0 72.6 159 7.0 0.4 4.1
1 7 69.0 17.0 6.6 0.3 7.1
2 28 66.6 16.1 6.3 0.3 107
3 54 68.7 15.8 6.3 0.3 8.9
4

100 63.0 162 6.0 03 145

 

*Unidentified material deposited in Inconel.

have caused some speculation as to the specnes

 of ions present at elevated temperatures. An

investigation of the thermal dissociation is one
of several studies being conducted to identify
these species.

The apparatus consists of an Alphatron pressure-
gage tube and a crucible machined from a single
crystal of magnesium oxide and connected through
a water-cooled copper block to a vacuum and a
gas-collection system. The hydroxide (B and A
reagent grade, carbonate-free, or prepared from
sodium metal and distilled water) is dehydrated
in situ at 400°C under vacuum for periods ranging
from several hours to several days. The temper-
ature is raised to 800°C, and then, after equilibrium
has been established, gas samples are taken for
mass spectrometric analysis, The residual
hydroxide has been analyzed, in some cases, by
wet chemical methods and, in one case, by x-ray
diffraction, |

The gas-analysis results have been somewhat
erratic; they showed 2 to 89% water and 10 to 82%
hydrogen at total pressures of 4 to 5 mm Hg.
However, in every case, both gases have been
present, and their sum ranges from 74 to 99%, the
balance being contaminants, such as nitrogen,
which probably arise from air leaks. Wet chemical
analysis sheds no light on the residual hydroxide
content, but x-ray diffraction has identified sodium

 

~ 16¢, R, Boston and M. E. Steldhtz, ANP Quar. Prog
Rep. June 10, 1954, ORNL-1729, p 85.

 

 

a3

 

 

 

 
 

» e

 

 

 

 

'temperatures, by fhe reaction

 

PERIOD ENDING SEPTEMBER 10, 1955

. UNCLASSIFIED
Y.15218

 

G000
7000
€000

 

 

 

INCHES

¢00'0
000

 

 

 

 

Fig. 5.19. Inconel Exposed to Sodium Hydroxide at 800°C,

oxide and, based on an x-ray line which is some-
what in doubt, sodium peroxide. It has long been
thought that sodium hydroxide should decompose,
as do the ulkalme-earth hydrox.ldes, by the

reuchon

(1) 2Nc:0H——-> Na, o + H 0

 

_ nd ‘H’HS__ work is the first |de_m‘1f|caf|on of Na o
as a decompos mon producf ‘

   

 

     

 

Recent fhermodyncmlc culculc:’nons‘I7 md:cuted
that, in addition to the reaction of Eq. 1, dis-
sociation is possnble, smuitaneously, ot hlgh

 

  

@) 2NaOH—=> Na,0, + H,

   

 

Theamoum‘ of hydrogen observed in these experi-

 

]7G. P. Smith and C. R. Boston, ANP Quar. Prog,
Rep. Sept. 10, 1954, ORNL-1771, p 102,

  

oo Fy Kertesz

 

 

  

ments is in excess of that predicted from the
calculations. It would appear, qualitatively,
that at 800°C o significant proportion of the

decomposition product is in the peroxide form.

CHEMICAL STUDIES OF CORROSION

H. J. Buttram
Materials Chemlsfry DIVISIOI‘I

_.Inconel in NaF-LiF- ZrF -UF
H. J. Buttram R. E. Meudows

- Materials Chemistry Division

A series of studies in which the four-componem‘

 

“fuel ‘NaF-LiF-ZrF 4-UF (22-37.5-35.5-5 mole %)

was exposed for ]00 11r in sealed capsules of
Inconel in the standard rocking furnace has
indicated that this mixture may be less corrosive
than others under consideration. The data ob-

tained for this mixture and for three other previ-

119

RO

AT

ot

Ry

 
 

 

ORERIE

 

ANP PROJECT PROGRESS REPORT

Fig. 5.20. Inconel Strips (25 mils thick) Tested
at 800°C in Sodium Hydroxide. Strips 1 to 4 were
exposed 7, 28, 54, and 100 hr, respectively.
Number 0, a control strip, was unexposed.

ously tested materials run as standards are shown

in Table 5.16.

Stability of UF; in NaF-KF-LiF in inconel
H. J. Buttram R. E. Meadows

Materials Chemistry Division

A series of mixtures of the NaF-KF-LiF eutectic
containing added UF; and UF, has been exposed
in sealed capsules of InconeT for times ranging
from 3 to 240 hr in the rocking furnace. The
capsules were heated and maintained at a hot-end
temperature of 800°C and a cold-end temperature
of 650°C. The rocker was operated at 4 cpm.
The total uranium contents of the mixtures studied
were 10 to 20 wt %. After exposure the capsules
were opened and unloaded in a helium-filled dry
box, and the cooled melts were ground and

- analyzed chemically for UF, and UF,.

As the data in Table 5.17 indicate, UF, is quite
unstable under these conditions, regardless of the
orjgino] UF,/UF, ratio. It appears that con-
siderable disproportionation of UF, must be
"expected in this system under all circumstances.

120

 

Effect of Chromium on the Mass Transfer of

Nickel in NaOH
H. J. Buttram R. E. Meadows
F. A. Knox

Materials Chemistry Division

The effect of chromium on the mass transfer of
nickel by NaOH was tested in a nickel rocking
furnace. The exposures, at 4 cpm, were for 100 hr,
with the hot ends of the capsules at 775°C and
with a 100°C temperature drop along the capsules,

When 1 wit % Cr® was added to the purified
NaOH, the mass transfer of nickel was reduced
by about a factor of 2. However, in no case was
the mass transfer eliminated, Analyses of the
caustic revealed the presence of about 4000 ppm
Ni in the tests without chromium added to the
NaOH, while about 2500 ppm Ni and 3200 ppm Cr
were found in the caustic to which 1T wt % Cr® was

added,

This decrease in mass transfer presumably can
be explained by the reaction of Cr® with NaOH to
yield hydrogen gas, which tends to suppress the
reaction of NaOH with Ni®. Since hydrogen diffuses
readily through nickel metal at the test tempera-
ture, these experiments were repeated with the
nickel capsules enclosed in evacuated, sealed
quartz jackets. In addition, two capsules were
sealed in quartz envelopes containing hydrogen
at 540 mm Hg pressure. After 100 hr in the
rocking furnace, no mass transfer was found in
any of the capsules, except one for which the
quartz envelope had cracked. This lack of mass
transfer was evident, however, even in the control
capsules containing no chromium, Furthermore,
analyses of the caustic after the tests showed
quite low (less than 100 ppm) concentrations of
nickel and chromium. Pressures inside the quartz
envelopes, obtained by breaking the quartz inside
an evacuated standard volume, were about 350 mm
Hg (590 for the one initially set at 540 mm Hg).
The gas was not analyzed, but it is presumed that
it was hydrogen.

From these studies it is not apparent that

chromium is particularly beneficial. It does
appear, however, that some mechanism to prevent
loss of hydrogen from the system — perhaps

cladding the nickel with some metal impervious to
hydrogen — might be quite beneficial,

o

 

 

 

 

 
 

 

Y

PERIOD ENDING SEPTEMBER 10, 1955

TABLE 5,16, CORROSION OF INCONEL CAPSULES EXPOSED IN ROCKING-FURNACE
TESTS FOR 100 hr TO SEVERAL FUEL MIXTURES

 

Chromium Found in Fuel Mixture Depth of Attack on

 

Fuel Mixture Composifion' After Test Inconel Capsule
(ppm) (mil}
NaF-LiF-KF 410 | | 1
(11.5+46.5-42 mole %)
NaF-Zr F4 ‘ 760 1
(53-47 mole %) '
' NcF-ZrF4-UF4 1370 ' 1
{53.5-40-6.5 mole %)
NaF-LiF-ZrF,-UF, 220 0.5

(22-37.5-35.5+5 mole %)

 

TABLE 5.17. DISPROPORTIONATION OF UF, IN NaF-KF-LiF EUTECTIC IN INCONEL CAPSULES
|  TESTED IN THE ROCKING-FURNACE APPARATUS

Hot-zone temperature: 800°C
Cold-zone temperature: 650°C

 

Uranium Compounds Found (wt %)

Uranium Compounds

 

 

 

 

Added (wt %) After 3 hr After 27 hr After 243 hr

UF, UF, UF, UF, UF, UF, UFy UF,
12.4 0 2.0 5.98 3.6 7.6 1.8 9.3
6.4 6.4 2.3 9.2 2.4 9.2 1.8 10.8
1.3 11.8 0.1 12.7 0.1 o127 0.3 12.5
o | 2.9 0.2 13.1
180 57 214
23.0 3.4 20.6

270 0 © 255

263 o1 2560

 

 

 

121

 
 

ANP PROJECT PROGRESS REPORT

6. METALLURGY AND CERAMICS

W. D. Manly

J. M. Warde

Metallurgy DlVlSlon

Stress-rupture tests are being mode of /-m.
Inconel tubing in argon and in fused salts in order
to evalucte the effects of the biaxial stresses
‘present in pressurized tubing. The results of tube-
‘burst tests are compared with those of tensile
'creep-rUpfure tests of 0,060-in.-thick sheet in
argon ‘and in fused salts at 1500°F. Data from
creep-rupture tests of 0,020- and 0.060-in.-thick
‘Inconel sheet are presented which show that rup-
ture life for a given stress is shorter for the
fhlhner'sh'ee'rw An evaluation of a welded Inconel
bellows for use in fused salts at high temperatures
and the resulfs of a study of the interaction be-
tween Inconel and beryllium under pressure in an
:merf environment are presented.

Additional oxidation and fabrication studies of
mckel-molybdenum alloys are discussed, and de-
~sign curves for solution-annealed Hastelloy B
tested in fused salts at 1500 and 1650°F are
given. The influence of aging on the creep-rupture
properties of Hastelloy B is being studied, and
data are presented for creep-rupture tests of aged
and solution-annealed material in argon at 1500
and 1800°F and for short-time fensile tests of ma-
terial aged at high temperatures.

The results of static and cyclic oxidation tests
at 1700°F on several brazing alloys are presented.
Several components containing cermet-to-Inconel
joints were successfully brazed, and techniques of
brazing boron carbide are being investigated.

The lnconel core-shell assembly for the high-
fempero'rure “critical experiments was fabricated
“after eXper:mentcl techniques for minimizing dis-
“tortion were developed. Methods of producing

- ';j'_fquanfmes of Coclsf Metals No. 52 presintered

- brazing clloy rings were developed in preparation
~ for the impending construction of another large
'V{_NGK-fo-alr racllafor. A second sodiumsto-air radia-
" tor was completed for the Corell Aeronautical
: 'Laboratory, ‘and two fuel-to«NaK heat exchanger

o ff’rube bundles are being fabricated.

"~ The bas;c‘ concep’rs of eddy-current testing as
 applied to low-conductivity alloy tubing are dis-
¢ussed, The merits of the impedance analysis
" method are presented, along with problems of flaw

detection with both encircling and probestype
'7":c0|ls._ An ultrasomc method for the inspection of

e

small-d(cmefer tubing is descnbed thaf is suf-

ficiently sensitive to detect the types of flows
encountered to date.

The investigation of corrosion-erosion in the
graphite-hydrogen system at high temperatures was
continued, Two control rod assemblies of rare-
earth oxides were fabricated, as well as calcium
fluoride and alumina spacers. The optimum press-
ing conditions for pelletizing fluoride fuels are
being studied, and an investigation of the feasi-

bility of synthesizing Mo,B. and B,C is under

275
way. Dysprosium oxide disks and ~europium oxide

wafers were fabricated.

The special materials studies reported include
the results of studies of diffusion barriers for use
between Inconel and columbium and a discussion
of further attempts to prepare two- and three-ply
tubing. The results of oxidation tests of several
commercial aluminum bronzes being considered for

oxidation protection of copper radiator fins are

presented.  Attempts to prepare creep-resistant
lead-calcium alloys for use as shielding material
are described, as well as experimental work under

~way for preparing B,C.Cy neutron shielding ma-

terial for the ART,

MECHANICAL PROPERTY TESTS OF INCONEL

D. A. Douglas J. R. Weir
J. H. DeVan J. W. Woods
Metallurgy Division

C R. Kennedy, Pratt & Whitney Aircraft

Stress-Rupture Tests

J. H. DeVan
Metallurgy Division

Stress-rupture testing of %-in. inconel tubing in
argon and in fused salts at 1300, 1500, and 1650°F
is now in progress., Although data for Inconel
tubing at 1300 and 1650°F are as yet incomplete,
the results of tests of Inconel tubing with wall
thicknesses of 0.010, 0.020, 0.040, and 0.060 in.
at 1500°F in argon and in fused salts are suffici-
ent to show important trends.

In studying the biaxial stress system set up in
the tubes, the question arises as to whether failure

is related to the effect of combined stresses or

 

 

 

 
»

R A

 

 

 

      

whether, 1he maximum stress component (in this
case, hoop stress) singularly controls the time to
failure, A comparison of the results of tensile
creep-rupture tests and tubesburst tests is pre-

sented in Fig, 6.1 for tests in argon and in Fig.

6.2 for tests in NaF-ZrF 4~UF, (50-46-4 mole %).
It may be seen that there is good agreement be-
tween rupture times for 0.060-in.-wall tubmg and
0.060-in.-thick sheet in both environments. Thus,
the presence of an axial stress in the tubing does
not appear to affect the time to failure,

It may be noted, however, that the specimens

“with thinner walls, 0,010 to 0.020 in., ruptured in
‘much shorter times than those observed for 0.060-

in.~thick sheet specimens at comparable stresses,
While this was originally thought to be due to the
effect of the combined stress system acting in the
tubing, it now appears from the results of tests on
0.020-in.-thick sheet that the poor rupture propet-
ties are associated with the smaller section thick-
ness.

PERIOD ENDING SEPTEMBER 10, 1955

The results of creep-rupture tests of 0.020-in,-
thick Inconel sheet in argon and in hydrogen at
1500 and 1650°F are compared with those of tests
of 0.060-in.-thick sheet in Table 6.1. As may be
seen, a substantial reduction in rupture life ac-
companies the reduction in section size,

Bellows Test

J. H. DeVan
Metallurgy Division

An evaluation of a welded-diaphragm bellows
for use as a seal in a shut-off valve for controlling
the circulation of fused salts was conducted by
utilizing a modified creep-rupture machine to simu-
late operating conditions. [n this equipment the
bellows was deformed while the outside and inside
surfaces were in contact with the fused salt mix-
fure NGF-ZrF4-UF4 (53.5-40-6.5 mole %). Tests
were made at 1300 and 1500°F, each for 100 hr, A

leak check following each test failed to indicate

UNCLASSIFIED
ORNL-LR-DWG 8936

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10,000 —
LT
- FINE-GRAIN, 0.060-in.- THICK SHEET
™
.
\:\\
. \
= 5000 s ,k\
o } \ N
w \ A N
2 O A
& \ |
’_
‘e SRR
= SETN . \
o= -
2 = -
= Al N
e © 34-in.~ID, 0.060 -in.-WALL TUBING NN \
9 ® 3-in-10, 0040 in.- WALL TUBING \\ \
2000 ——— 1,1in.-1D, 0.020- in.- ~WALL TUBING " SN
4 Ypin-ID, 0.010-in“WALL TUBING N
7800 1000 2000 5000

Fig. 6.1.-

 

10,000

Comparison of Tube-Burst and Stress-Rupture Tests of Inconel in Argon at 1500°F.

123

 
 

 

 

 

 

ANP PROJECT PROGRESS REPORT

ORNL-LR-DWG 8937

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10,000
L FINE- GRAIN, 0.060-in.- THICK SHEET
™
<\
\\

5000 A \
= \
2 \ \
9} A \\ \\= };\
& \ NN
5 N ON|
oy N RN
= , \
g O 3,-in.-iD, 0.060-in.-WALL TUBING N \
" ® 3,-in.-ID, 0.040-in.-WALL TUBING \\ \\ N
2 & Yy-in.-ID, 0.020-in.- WALL TUBING *\
O
S oo A Yo-in.~1D, 0.010-in-WALL TUBING \ | N
\\ |

]\
\\
\\
1000 AL , '
{ 2 5 10 20 50 100 200 500 1000 2000 5000 10,000

RUPTURE TIME (hr)

Fig. 6.2, Comparison of Tube-Burst and Stress-Rupture Tests of Inconel in NaF-ZrF4-UF4 (50-46-4

mole %) at 1500°F.

TABLE 6.1. COMPARISON OF RESULTS OF STRESS-RUPTURE TESTS ON 0.060- AND
' 0.020-in.-THICK INCONEL SHEET

 

Time to Rupture (hr)

Total Elongation (%)

 

 

 

Temperature Stress
(°F) (psi) Atmosphere 0.020-in.- 0.060-in.- 0.020-in 0.060-in.-
Thick Sheet Thick Sheet Thick Sheet Thick Sheet
1500 3500 Argon 270 1467 8 13
Hydrogen 350 618 n 8
1650 2000 Argon 740 1125 14 30
Hydrogen 200 385 12 N

 

any flaws in the bellows. Metallographic examina-

tion showed only normal corrosive aftack in the

weld areas, and no cracks could be detected as a
direct result of the flexing of the bellows.
_-_Interq_;fipn Bgtween Inconel and Beryllium

A programfordefermmmg the extent of inter-

action between beryllium and Inconel at elevated
temperatures in an inert environment was'reCentIy
completed to establish design limits for the high-
temperature critical experiment.  The - modified
creep-rupture equipment was used for the tests so
that a conirolled atmosphere of argon could be
maintained and a specified load could be trans-

 

 

 

 

 
 

"

wh

 

mitted to the specimens of beryllium and Inconel
at 1350°F, The results of the three tests are
summarized in Table 6.2, |
The diffusion area in the Inconel specimen from
the 100-hr test is shown in Fig. 6.3. The extreme
hardness, as shown by Knoop indentations, and
the sharp interface of the diffusion area indicate
TABLE 6.2. RESULTS OF TESTS OF THE INTERAC-
TION BETWEEN INCONEL AND BERYLLIUM UNDER
'DIFFERENT PRESSURES FOR CONTACT TIMES
OF 40 AND 100 hr '

 

 

Time in Contact Depth of Alloyed
Contact Pressure Region in
(hr) (psi) tnconel (in.)
100 500 | 0.015
40 - 100 | 0.005

40 50 0.003

 

 

PERIOD ENDING SEPTEMBER 10, 1955

that an intermetallic layer was formed by the inter-
action of the beryllium and the Inconel. This layer
is, undoubtedly, detrimental to the load-carrying
ability of the Inconel, although it represents an
extremely small percentage of the cross section
of the specimen tested.

DEVELOPMENT OF NICKEL-MOL YBDENUM

ALLOYS
J. H. Coobs J. P. Page
H. Inouye T. K. Roche

. Metallurgy Division
M. R. D'Amore, Pratt & Whitney Aircraft

N | Oxidution‘Studies
Experimental data obtained on the oxidation in
static air at 1500°F of several nickel-molybdenum
alloys are presented in Table 6.3. The alloy with
the nominal composition 10% Mo-10% Fe-6% Cr—

74% Ni was tested previously in a fluoride mixture

 

 

- Fig. 6.3.- Diffusfon Area in k|ncone| After Contact with Beryllium for 100 hr at 500 psi.v 500X. Reduced

5.5%.

125

 
 

 

 

ANP PROJECT PROGRESS REPORT

TABLE 6.3. OXIDATION RATES OF NICKEL-BASE
ALLOYS IN'STATIC AIR AT 1500°F FOR 167 hr

 

Ce e Wéight.gain
) Al.l‘lox o (g/cmz) Remarks

 

10% Mo—-10% Fe—-6% 0.0005 Ac”;erent oxide
Cr --?4% N.i formed

”H"as;teu;'y B,  0.0010

vacuum melfed

Oxide spalled
upon cooling

| 2.5% Be-97. 5% N|  0.0004 Adherent oxide
i , o formedr _
2.5% Be--S% - 0,0014 Adherent oxide

Cb-‘9_2.5% Ni 7 7 forr_ned

 

and was found to be promising. Its oxidation rate

was between that observed for a 7% Cr—-20% Mo~-
73% Ni alloy and that for a 10% Cr-20% Mo—70%
Ni alloy. Further, it is considered to be heat
resistant, and it forms a nonspalling oxide upon
oxidation. The samples from a vacuum melt of
Hastelloy B supplied by the Haynes Stellite Com-
pany showed no significant differences from com.
mercial grades of Hastelloy B, The 2.5% Be-
97.5% Ni alloy also forms a nonspalling oxide and
is considered to be heat resistant, The addition
of 5% Cb, however, increased the oxidation rate to
that observed for Hastelloy B.

Fabrication Studies

[t was reported previously! that a deleterious
effect in nickel-molybdenum alloys containing
titanium, aluminum, vanadium, zirconium, colum-
bium, or chromium was noted after long heat treat-
ments in hydrogen. Additional similar tests were

 

]J. H. Coobs, H. Inouye, and M. R. D'Amore, ANP
Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 120.

TABLE 6.4, ROOM—TEMPERATURE TENSILE TESTS OF 0 065-m SHEET SPECIMENS OF

made to ensure that the effect was au'e'fo hydrogen .

and not to the aging heat treatment, The check

__:_‘tesfs were made on a 5% Cb-20% Mo-75% Nl‘_ .
“alloy. The results of the tests, presented in
Table 6.4, verified that low strength and duc-

tility resulted from long exposure in hydrogen at
elevated temperatures. The tests also indicated
that the effect of hydrogen could be removed by
vacuum annealmg.

The possibility that a hydride forms during
aging has been suggested, but, if such is the

“case, embrittlement in alloys containing chromium

and aluminum cannot be explcined The embrittie-

““ment effect was not detected in tests of Hastelloy

B and of the binary nickel-molybdenum alloys, and

it was not found in all the elevated-temperature -

tensile tests of the ternary alloys that have been
studied.

Twelve extrusions of Hastelloy B were attempted
by the Babcock & Wilcox Co. to determine the
commercial feasibility of making tube blanks for
seamless tubing. This effort was supported by
the Haynes Stellite Company, Each of the twelve
extrusion attempts were unsuccessful because of
severe cracking of the tube blanks or failure to
extrude (‘‘stickers’’). The extrusion temperatures

used were between 2000 and 2250°F, Based upon

experience at ORNL, it is concluded that improper
temperatures, socking time, and press capacity
were the main causes of failure,

Since this effort was made because successful
extrusions had been made at ORNL, additional ex-
periments were performed here to verify the recom-
mended extrusion conditions, to verify, with a
different heat, the superior redrawing properties of

the exfruded tube blonks, cmd fo de‘rennine meth_cr)d&s N

5% Cb—20% Mo-75% Ni ALLOY

'Specimens annealed 1 hr at 21QOOF and aged at 1650°F

 

Yield Point,

Reducflon o e

 

: by 2hr in vacuum

Condifiofi of Tést Specimen 0.2% Offset Tensile Strength Elongation in Area
- Aged 285 b in vacuum | 78,500 153,700 35
A Aged 285 hr in hydrogen | 77,600 113,600 h 13
_ Aged 285 hr in hydrogen followed 78,300 ]51,000 30

 

126

 

 

 

 

 

 

 
S

o e b

S

 

 

 

 

for reducing. the pressure required for the extru-
sion. Two uncanned billets of Hastelloy B were
successfully extruded into tube blanks at an ex-
~trusion ratio of 5.5/1, The recommended condi-
_ tions were verified: namely, temperature, 2000°F;
‘heating time, 45 min per inch of billet thickness;
and extrusion pressure, 80 to 90 tsi. These tube
blanks are to be redrawn to small-diameter tubing.

A Hastelloy B blllef conned in / ¢-inethick
Inconel was extruded at 2100°F at an ex'rrus:on
~ ratio of 5.5/1, The pressure required for this ex-
" trusion was 55 tsi, as compared with 80 tsi for the
uncanned billets, A second canned extrusion was
made at a ratio of 9/1 at 2100°F with fair success,
At these temperatures, the first billet, if uncanned,
would have cracked severely. The high extrusion
ratio used for the second extrusion would have
caused the blHef, if uncanned to be a sticker,

The use of Inconel as cannmg material for
Hastelloy B is justified, since it serves a three-
fold purpose.. Besides acting as a lubricant, the
Inconel provides a thermally bonded oxidation-
resistant tayer for the alloy. Also, in a high-
temperafure air stream, the Inconel would inhibit
formation of the corrosive MoO vapor, and thus
there would be no possibility of catastrophic oxi-
dcmon occurring in the structural material, More-
over, the clad Hastelloy B could probably be re-
duced more successfully than the unclad material,
because Inconel is not sub|ect to ‘‘heat cracking,’’
that is, surface cracking durmg cnneahng of cold

work that is aftributable to poor themal conduc-.

tivity and a high rate of work hcrrdemng.

 
 
 
 
 
 
 

PERIOD ENDING SEPTEMBER 10, 1955

STRESS-RUPTURE TESTS OF HASTELLOY B

D. A, Douglas
Metallurgy Division

C. R. Kennedy
Pratt & Whitney Aircraft

A series of creep-rupture tests has been com.
pleted on solution-annealed Hastelloy B at 1500
and 1650°F in the fluoride mixture NaF-ZrF -UF
(50-46-4 mole %). The results are summanzed m
design curves presented in Figs. 6.4 and 6.5. A
similar series of tests is nearly completed for
solution-annealed Hastelloy B at 1300 and 1800°F
in argon and in the fluoride mixture.

The influence of aging heat treatments on the
creep-rupture properties of Hastelloy B is being
studied at several temperatures ranging from 1300
to 1800°F. In general, aged specimens have
shown marked reductions in creep rate and tensile
elongations compared with specimens in the solu-
tion-annealed condition. Results obtained from
creep-rupture tests on solution-annealed and aged
specimens at 1500 and 1800°F in argon are pre-
sented in Table 6.5,

The creep-rupture properties at 1800°F are not
substantially affected by aging, since at this tem-
perature a single-phase alloy exists. However, at
1500°F, a second phase that appears upon aging
produces a marked change in the creep-rupture
properties of the alloy. It may be noted that for
stresses which produce a rupture life in excess of
100 hr for the solution-annealed material at 1500°F
the difference in rupture life of the solution~

This is

annealed and aged specimens is small.

. understandable, since, during testing at this tem-

~Two additional tube b!anks of_q\]57 M'o.-ss% N

peroture, the solunon-annealed specimens actually
.. age.” During testing at higher stresses, the solu-
V__:_:hon-c‘:nnealed materlul does not age appreciably

before rupture, and thus the rupture times for the

aged specimens are opprecmbly longer than those

for the squhon-anneoled Spec:mens.

 

PHYSlCAL PROPERTIES AND

MICROSTRUCTURE STUDIES OF HASTELLOY B

  
 

 

afier aglng,
modifications
o _VHClstelloy B. Furf

loy W,

  

d fhe""effect of shqht composmon
e mechar 'cal prOpen‘les of
,""cvresearch program parailel-”‘f"w_
‘ing that of Hastelloy B is com‘emplafed for Hasfel-\-

   

P chfrlarca R. E. Clausing
' Metallurgy Division

Short-tume 'rens:le tests conducted after long-time
" aging of Hasfe]loy B at elevufed temperatures in-
" dicate that residual stresses not removed affer a

127

 
ANP PROJECT PROGRESS REPORT

’ ORNL—I!R—DWG 8938

20000

=

 

ELONGATION 1% RUPTURE

15,000

12,000

STRESS (psi)

10,000

 

 

 

6000
1 2 5 10 20 50 100 200 500 1000 2060 5000 10,000

TIME (hr)

Fig. 6.4. Design Data for Solution-Annealed (2100°F), 0.060-in.-Thick Hastelloy B Sheet Tested at
1500°F in NoF-ZrF -UF , (50-46-4 mole %). ‘

TABLE 6.5. RESULTS OF CREEP-RUPTURE TESTS ON AGED AND SOLUTION-ANNEALED
HASTELLOY B AT 1500 AND 1800°F IN ARGON

 

 

 

Test Final
Temperature Condition Stress Rupture Time Elongation
(°F) (psi) (h) (%)
1500 Solution annealed 12,000 | 1180 20

Aged 40 hr at 1500°F 12,000 1450 16
~ Aged 70 hr at 1500°F 12,000 1070 8.7
Solution annealed 13,500 170 16
Aged 100 hr at 1300°F 13,500 670 57
Solution annealed 15,000 160 49
Aged 100 br at 1300°F 15,000 330 n
1800 Solution annealed 5,000 112 12
Aged 100 hr at 1300°F 5,000 - 90 10
Solution annealed | 3,500 410 7.3
Aged 100 hr at 1300 3,500 300 59

 

S8

 

 
 

 

 

PERIOD ENDING SEPTEMBER 10, 1955

ORNL—-LR-DWG 8939

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

14,000 '
12,000
4 ELONGATION {% (2% 5 % 10 % RUPTURE
10,000 \ =
\ \\\ \\ . N
N ™ \
\ \\ ™ N
z ™~ NN
s ™\ L NN
& \ N N
W \ N ‘ NNEEANY
= ™~ N
0 k \\ \\ \
\ ™
. ™ u \ \
SN ~ \
ny \ §
5000 \\ NG N
4000 :
f 2 - 10 20 50 100 200 500 1000 2000 5000 10,000
- ' . TIME (hr) |

Fig. 6.5. De5|gn Data tor Solutlon-Annealed (2100° F), 0 060-|n.-Th|ck Haste!loy B Sheet Tested at
1650°F in NaF Z:F -UF (50-46-4 mole %)

 

celd-WOrl(ing opereti"efi‘er]e ‘qu'ite.detr'imentct to the 1500°F ate-:c':eriksid\erd'l")ly grecter.' than the ductili-
Tensile specimens, which t:es of those aged at 1400°F. This may be due to
‘a reduced amount of beta phose, which may be

aB_Ie dt 1500°Fm The gamma phase appears to be

A

~ ductility of this alloy.
- haod been cast from sheet that hadw been cold-

 

  
    
   
 

 

 

o AR

“from 1100 to. 1500°F in i

R e

nd teeted at_ both roomvand,“elevute-

SR e A G reRR

in_ Table 6.6 show__.consflémb,‘xfijower

o ERATIAR G

d'uctthty for the cold-wo

 

 

 

  

= B i

”"“'noted that the ductliltles 3\‘ the specxn';enaswoged at the onalyses of the ef'Fectsf 'ot heat treatment .

 

  

129

   

 

  
 
 
 
 
 

will cf.d consnderably In A

 
  

‘ANP PROJECT PROGRESS REPORT

 
 

' TABLE 6.6. RESULTS OF SHORT-TIME TENSILE TESTS OF HASTELLOY B )
AGED AT HIGH TEMPERATURES

 

 

Teh.sile :Yie!\d:i

 

 

 

 

BRAZING ALLOY DEVELOPMENT AND
TESTING

P. Patriarca R. E. Clausing
Metallurgy Division

, Developments Tests

The OXIdelOI’I re5|stance of high-temperature
brazing alloys under static conditions at 1500 and
1700°F and in cyclic service at 1500°F has been
reported 2,3 The results of cyclic tests at 1700°F
are repon‘ed in Table 6.7. It can be seen that the

cycled from room temperofure to 1700°F is greater
than thot of [omfs ‘tested at a constant tempera-

~ ture. The Coast Mefals No. 52 alloy, which is
" ysed exfens:vely in brazmg h:gh-conducflwty-fin
V‘radrctors, was atfacked severely in these tests,
f_jjwhtle fhe chroml‘um-bearmg Coosf Metals No. 53

 
  
    

 

2P Po'rnarca et al. ANP Quar Prog. Rep March 10,
955, ORNL-1864, p 121.

3P Patriarca et al., ANP Quar. Prog. Rep. June 10,
955, ORNL- 1896, P 130.

    

extent of ox;dahon of brazed joints thermally'

alloy was not. The No. 53 alloy will therefore be
seriously considered for this fabrication applica-
tion, even though a higher quality hydrogen atmos-
phere may be required to obtain flow with this
chromium-containing alloy.

Melting point studies are being made on the

high-temperature brazing a“oys by usmg “sintered,

conical samples. The cones are smtered from flnew
powder in a Lavite mold and heated at various tem-T T}
peratures on a thin, nickel sheet in a dry-hydrogen'_'* T
atmosphere. The results of studies on a ’rypuca] .
brazing alloy, Cocsf Metals No. 52 ore shown in o E

F;g. 6 b.

Brazing of Boron Carbide

 

: Agmg - Aging Tesf
""Tem‘;‘)erd;c;e ~ Time Temperature Ductility Strength Point
S CFY (hr) R (%) (psi) (psi)
Cold Reduced 20% Before Aglng _
500 Room o 3.0 181,300 ,f.] 154,600
500 1100 2.5 130,500 115,500
500 1100 3.5 122,900 104,400
200 Room 0.8 173,500 150,000
200 Room 0.5 191,600 148,000
| 500 Room 3.8 164,000 88,200
500 1400 6.3 84500 T
500 Room 20.0 163,000 100,000
500 ‘ 1500 30.0 64,600
Fu“y Annealed Before Aging _
500 Room 14.0 126,000  zgoco T g
500 1100 14.0 91,100 58,600
; 500 Room 6.0 92,600 6,000
: 500 1300 12.5 140,600 90,000
500 Room 20.0 128,000 73,500
500 1400 10.0 76,500 ' '
500 Room 12.5 109,800 54,900
500 1500 37.5 63,100 )

The incorporation of a boron carbide radmhon: e

 

shield into a fused-salt pump assembly req

that the boron carbide compact be brazed to cn it
Inconel envelope to provide sufficient thermal con-

ductivity to remove the heat induced by radiation.
The service temperature may be as high as 1700°F,
with helium as a protective atmosphere. Although

 

  
 
 
 

 

 

 

 

 
 

' 'TABLE 6.7, OXIDATION RESISTANCE OF DRY-
HYDROGEN BRAZED INCONEL T-JOINTS
TESTED FOR 500 hr AT 1700°F

 

- _ -’; _Oxid'c'rion with
Oxidation 220 Air Cools

 

Brazing Alioy* ip Static to Room
Airr Temperature**
Wall Colmonoy Nicrobraz .Slight Slight
Coast Metals No. 53 . Slight Slight
Wall Colmonoy lowsmelting Slight Moderate
Nicrobraz
Coast Metals No. 50 Slight Moderate
Pd-Ni (60-40 wt %) Slight Moderate
Pd«Al (92-8 wt %) ~ Slight Moderate
Coast Metals No. 51 Slight Severe
Coast Metals No. 52 Slight Severe
Coast Metals NP Mdderafe Moderate
G-E No. 81 MQJerote Moderate
. : Ni-Ge (75-25 wt %} " Moderate Moderate
Ni-Ge-Cr (65-25-10 wt %)  Moderate Moderate
N i-Mo-Ge (50-25-25 wt %) Moderate Moderate
: AuNi (82-18 wt %) Moderate Severe
Au-Co (90-10 wt %) ' Severe Severe
, Pd-Ge (90-10 wt %) o :“C_ompr|efe Complete
' Ni-Sn (68-32 wt %) 'Com.plefe " Complete
Ag-Pd-Mn (64-33-3 wt %) Complete Complete
Ni-Mn (40-60 wt %) " Complete  Complete
Ni=Mn-Cr (35-55-10 wt %) Complete Complete
Au-Cu (80-20 wt %) ~ Complete Complete
Ni-Mn (40-60 wt %) Complete Complete
Copper Complete Complete

 

*Alloys listed in order of decreasing oxidation re-

*

mstonce. .

**Very shght Iess than 1 mal of penefrahon sl:ghf

_ 1to 2 mils of penetration; moderate, 2 to 5 mils of pene-
“tration; severe, greater than 5 mils of penetrahon, coms-

PERIOD ENDING SEPTEMBER 10, 1955

a large number of high-temperature brazing alioys
have been tested, none investigated to date have
proved entirely satisfactory, Most of the brazing
alloys which wet the boron carbide also react with
it to form brittle bonds that crack upon cooling.
The most promising alloys tested thus far are
those that either contain zirconium as a component
or require an application of zirconium hydride
powder to the boron carbide surface prior to place-
ment of the brazing alloy on the same surface,
Experimental alloys are now being prepared to
permit a more thorough investigation,

Brazing of Cermet Valve Seats to
Inconel Components

Three assemblies incorporating Kentanium cermet
valve parts have been successfully brazed by using
the technique described previously.* An exploded
view of an assembly for determining the self-
welding characteristics of these cermet valve parts
is shown in Fig. 6.7. This assembly consists of
an Inconel structural piece, two copper-foil disks
that supply brazing alloy for the lnconel-to-nickel
joint, @ I/4--1in.-thi<:l-< nickel block to dissipate ther-
mal stresses resulting from the different coeffici~
ents of thermal expansion of Inconel and Kentani-
um, a copper-foil disk to supply supplementary
brazing alloy for the nickel-to-Kentanium joint, and
the Kentanium seat plated with a 0.0001-in, layer
of nickel-phosphorus followed by a 0.003-in, layer
of copper on the surface to be brazed,

‘Figure 6.8 shows an actual valve subassembly,
which is similar to the assembly for self-welding
studies except for the omission of several copper-
foil disks, Copper powder, which was substituted
for the disks, was placed in especially provided

 

 

 

. pie*er f||[ef completely destroyed.

   

*1bid., p 145.

 

1000°C 1020°C

Fig. 6.6. Melting-Point Determinations of Coast Metals No. 52 Brazing Alloy Made by Using Sintered
Cones Heated on Thin, Nickel Sheet in a Dry-Hydrogen Atmosphere.

"

 

131

 

 

 

 
 

 

 

 

ANP PROJECT PROGRESS REPORT

   

'UNCLASSIFED © receésses. The complete valve is shown after

          

     
    
      
   

  

Y-16173 . L x
brazing, but prior to finish machining, in Fig. 6.9, o
}... ¢.
i w FABRICATION OF TEST COMPONENTS f
J Components for High-Temperature
y \/ "** . T Critical Experiment
g x - “ P. Patriarca
tg & S g Metallurgy Division
z =
; L ' /U\ + S The fabrication of the inconel core shell for the
X S i high-temperature critical experiment required ex-
’ tremely close control of the distortion associated :
A x with production of the girth weld. Preliminary
& % a experiments were therefore conducted to determine,

U

T YREUKE D
YA

 

SEAT DiSK
{£) BRAZED ASSEMBLIES

Fig. 6.7. Assemblies for Self-Welding Studies of Fig. 6;9. Brazed Valve Aésembiy Pribr‘ to
a TiC-Ni Valve Seat and Disk. Machining.

 

Fig. 6.8. V.olvéISeat and DlSk Subussembhes. "

 

132

 
 

o

”w

 

 

~dnd the oufe cc;re “shell T »
“inner shell, The complefed Ieakhght' umt |s shown
‘in Fig..6 12, N

qucnhtahve!y, the extent of distortion to be exs
pected. The first test consisted in the deposition
of a butt weld between two cylindrical segments
4 in, in depth, Each segment was machined from
a nominal, 12-in.~ID, sched-40 pipe, in which the
inside diameter was bored to 12,250 in. to improve
the roundness and the ou'rsrde diameter was tumed
down to 12,500 in. to provide a 0,125.in. wall for
3 in. of the 4-in. depth. A stainless steel face
plate was tack-welded at 60-deg intervals to the
l-in, shoulder of each segment to provide a degree
of restraint intended to approximate that provided
by the tapered reduction in diameter of the island
sheath. Micrometer caliper measurements at 48
scribed positions indicated that a maximum change
of 0.030 in. occurred in the nominal 12,5-in. out-
side diameter, As was expected, the frequency of
large changes increased wn‘h increasing proximity
to the weld. o

A second test was conducted to determine the
extent of distortion to be expected when butt-
welding the spun core segments with the aid of
mechanical support, Two 6-in.-deep Inconel cylin-
ders, 12,5 in. OD, ]/s-in. wall, were butt-welded
with an aluminum backing ring containing Lavite
inserts for support. Micrometer and dial-gage
measurements made at 96 positions on the weld-
ment, before and after welding, indicated a maxi-
mum shrinkage of 0,020 in. on the 12.5-in, outside
diameter, that is, an improvement of 0,010 in. over
the shrinkage obtained without mechanical support.

~The information gained in these experiments
was then utilized in the fabrication of the actual

core shells, The upper and lower halves of the
“inner shell were stress-reheved in dry hydrogen |
for 2 hr at 1500°F with heating and coolmg times

of 3 hr. They were then machined to the desired

length and beveled to fac:htate weldmg. Berylhum
blocks were fhen ‘assembled in the island inner =
‘ :_core, as shown in Flg. 6. ]0 The outside of the

:slund core shell_ is_shown in Fig. 6. ]] ofi*er com-

   

 

Two control rods were also fabricated for the
high-temperature critical experiment. Inserts of

PERIOD ENDING SEP TEMBER 10, 1955

 

Fig. 6.10. Lower Half of Inconel Island Core
Shell Showing Stacked-Beryllium«Block Inner As.
sembly Before Positioning of Upper Half of the
Core Shell.

rare-earth oxides were assembled in the control
rod housing and sealed by welding.

NaK-to-Air Radiator

P. Patriarca G. M. Slaughter
Metcllurgy Division

R L. Heestand |
Prcfi & Whlfney Aircraft

The fabr:ca’rlon of a third, 500-kw NcK-to-mr
rad:q’ror with type 310 stainless-steel-clad copper,
high-conductivity fins, is under way, Although
the design of this unit is essentially the same as
that of the radiators fabricated previously,® a few
modifications in the fabrication techniques will
be utilized. These modifications include the

 

SIbid., p 134.

133

 

 

 
 

 

 

 

 

ANP PROJECT PROGRESS REPORT

 

 

 

Fig. 6.11. Completed Island Core Shell After

Deposition of the Girth Weld.

construction of a 36-hole punching die to assure a
uniform fin-punching geometry, the construction of
new tube-bending dies for close control of the
tube-bending  variables, the development of
techniques to permit the assembly of headers in
such a manner that all the tubes will enter normal
to the curvature of the header at the point of
entrance, and the utilization of the presintered-
ring method of brazing-alloy preplacement.
Techniques for the preparation of presintered
rings of Coast Metals No. 52 brazing alloy were
developed by personnel of Pratt & Whitney
Aircraft.® The alloy is sintered in suitable
graphite molds in a dry-hydrogen atmosphere
to form rings that possess adequate strength and
rigidity. Molds of the type shown in Fig. 6.13
. were used to prepare approximately 45,000 rings,

 

SPratt & Whitney Aircraft, Nficlear Propulsiofi Proe
ram Engineering Prog. Rep. No. 15, PWAC 551, p 71

o (1955).

4

 

 

Flg. 6;12.. ’Completed. Core As‘semblly Shéfiing
Outer Core Shell and Upper and L.ower Hubs.

3/16 in. in inside diameter, Exper-imenfs conducted
on 12-in. lengths of vertically brazed tube-to-fin
joints have indicated that good flowability and
edge-protection of the copper on the punched fin
lips can be obtained with the use of these brazing-
alloy rings.

Intermediate Heat Exchanger No, 3

P. Patriarca G. M. Slaughter
Metallurgy Division
R. L. Heestand
Pratt & Whitney Aircraft

One 500-kw, fuel-to-NaK heat exchanger tube
bundle has been fabricated for the third series of
intermediate heat exchanger experiments, and the
second tube bundle is partially completed. Although

these bundles are similar to those incorporated?

 

7P. Patriarca et al,, ANP Quar. Prog, Rep. June 10,
1955, ORN|.-1896, p 131.

Ll

 

 

 

T T 7

 

 
 

CM 52 ALLOY

 

in intermediate heat exchanger No. 2, several
design changes have been made to promote better
operating characteristics and to aid in fabrication.
Larger diameter tubing, 0.250 in., is being used,
and the right-angle corners of the header components
of the previous heat exchanger have been eliminated

‘The new desngn is shown in Frgs 6.14 and 6.15.

    
   
 
 
 
    

 

  

is shown at

ishell design to faci l'rcn‘e fabrl ation.

Al

mixture and to sodium and since its flowability
s ~ characteristics are better than those of Nicrobraz,
The alloy is applied as presintered rings, which

 

to. prowde more favorable stress distribution,

. " The torch used in semloufomchcally weldmg 'rher
ltubeoto-header joints of one bundle is shown in
iFlg. 6.14, whlle the header, mcludmg the stronge

‘:backs used to mmlml'e"dxsforhon durmg weldmg,'

Icfer stage of fdbrlcuflon in Fig. 6.15. ’

A modlflcatlon was also ‘made in fhe pressure-'

ik ::::adequofe corrosion resistance to both the fluoride

PERIOD ENDING SEPTEMBER 10, 1955

UNCLASSIFIED
Y«16104

 

SINTERED RINGS

 

GRAPHITE MOLD

& Fig. 6.13. Graphite Mold Used to Produce Sintered Brazing Alloy Rings From As-Received Powder.

are shown in position on the underside of the heat
exchanger header in Fig. 6.16. These rings can
be secured to the header during assembly of the
tubes either by electric-resistance spot welding or
by using a volatile organic binder, Figure 6.16
also shows the diagonal comblike spacers
installed on these units to assure uniform spacing

 between the tubes and to increase the rigidity of

"rjh'e _bundle.
Cornell Radiator No, 2

P. Patriarca G. M. Slaughter
Me’ra”urgy DIVISIOn

~ R.L. Heestand
Prcfif & Whlfney Alrcraf'r

 

A second hqu:d ‘metal-to-air radlcltor, de5|gned

by the Cornell Aeronautical Laboratory, has been
fabricated, This construction problem was very

5|m|ldf to fha'r described prevmusiy, except that
one header was not attached to the side plates
and was thus free to ‘‘float” during thermal

 

81bid., p 139.

135

 
 

 

 

 

LR

 

ANP PROJECT PROGRESS REPORT

UNCLASSIFIED"
CYaTeezs

 

Fig. 6.14. One Header of the Intermediate Heat
Exchanger No, 3 Tube Bundle After Welding of the
Tube-to-Header Joints by Semiautomatic Inert-Arc
Techniques. Masking tape was used to ensure
adequate shielding-gas protection ot the roots of
the welds.

expansion and contraction of the assembly. The
radiator core, with the floating header at the
bottom, is shown in Fig. 6.17.

The side plates and remaining header sections
were welded to the core shown in Fig. 6.17, and
the tube-to-header joints were back-brazed with

~ Coast Metals No. 50 alloy. Leak testing with a

helium mass spectrometer indicated that all welded
and brazed joints were leaktight,

136

UNCLASSIFIED
Y-16061

   

Fig. 6.15. Improved Header Design of Inter
mediate Heat Exchanger No. 3. Strongbacks were
used to minimize distortion during welding.

NONDESTRUCTIVE TESTING
R. B. Oliver

J. W, Allen K. Reber
R. W. McClung
Metallurgy Division

The recent successes of several AEC instal-
lations in the use of eddy currents for testing
aluminum shapes suggested that the technique
might be applied to the inspection of Inconel and
stainless steel tubing. A comprehensive literature
and patent survey indicated that the eddy-current
flaw detection method had been successfully
applied to magnetic materials and to high-conduc-
tivity nonmagnetic materials but that little success

 

 

 

 

 

 

 

 
PERIOD ENDING SEPTEMBER 10, 1955

 

 

 

- w},“‘

 

 

 

  

: Fig. 6.!6:..' lnfermedxafe Heat Elkehan'gerrNo.Lg Tube Bundle Showing Coast Metals Alloy No. 52 Brazing
Alloy Presintered Rings for Back Brazing in Place on Under Side of the Header and the Combiike Spacers
for the Tubes.

-

   
 
 
 
 
  
    
  
 
  
  
 

  
 
 
  
 
   
 

  

 

 

hvfly nonmagneflc a“oys.

Laboratory"kwork with

md]c__a_'red “however,

      

LA R =

     

A

hrough an encnrcimg coil fh'an |t |s 'ro cause a

S __equipment has_
¥ i that there are good ‘pos'SIbllmi fa|r he sucaessful

se “of eddy-current fechmques

 

 

          

 

           

Tl

fuBe, the"inlhal emphasns has been placed on the

N

  
     
    

 

St e

2 nc1rc||ng coils.’
Since the‘magmtude\cf the eddy current in a

So ARy

fparhcuiar plece of ma'rer:al is dlfectly proporhonal

 

“of filncronel tubing. "

 

   
 
 

 
 

 

      

 

:;ln brief, thew Vrii‘éthad' consists ‘in exciting eday
‘ =t “_H_[nspec'fed by bringlng

f a éonl whlch

   

 

 

   

   

    

     

sensmv:fy “of any eddy-curren’r' instrument varies

“with the frequency\” 1t has also been found that to

T e

   

 

        
 

      

;obfaln adequate sensmwfy, flght coupllng between

and 'rhe tubmg must be malnfalned A

lmpedyance

e el

‘in the

 

      
   

  
  
 

 

 

    

         

h.—-“ - s 2

 

AT

tube that changes the magm’rude or fl;e d:strnbuhon

- - SRR o S et

of the eddy curren
the impedance of the coil and moy be measured.  currents in the tube wall, The relative magnitudes

Itl

i
i '  Coils that encircle the tube and very small “‘probe”  of the eddy currents at any depth x in a tube

 

 

 

   

 

 

n 4t"fi‘e sensitivity ‘

ts is reflected as a change

 

    

is the “skin effect” of the eddy

- 137

 

         

 

 

 

kL

 
 

 

 

ANP PROJECT PROGRESS REPORT

UNCL ASSIFIED
Y-15160

 

Fig. 6.17. Core of Cornell Radiator No. 2.

wdl_l may be expressed by:

i
x - e_x\f‘TTfp,)t ,
Isurfo;e
w‘_h'er'ef L
f = frequency

“permeability,
conduchwfy.

="
Ii

It

      
   
 

   

For Inconel,

e elaionship s

 

T

x.

surface -
o Inconel

-5 Hx\/_x 107

whlch mdncates thcn‘ 'rhe frequency must be limited
if the eddy currents are to penetrate to the inside
,wall of the ?ube._ The choice of frequency is thus
'a compromise tho'r must be made for each tube

(for x in mlls) :

138

 

 

 

 

 

 

 

 

 

 

   
 
 
 
    

 

 

 

  

 

 

 

 

 

 

 

‘size and material, The frequency selections for :
testing Inconel tubing are given in Table 6.8 .

An impedance diagrom for a tightly coupled :
coil encircling a Y%-in.dia Inconel rod is shown
in Fig. 6.18 (sohcf curve through open circles),
The frequency of the current (m kllocycles per ' E

TABLE 6.8. OPTIMUM FREQUENCIES FOR
EDDY-CURRENT FLAW DETECTION IN i
INCONEL TUBING b
Outside Diameter  Optimum Maximum Inspection
of Tubing Frequency Depth F
(in.) (ke) {in) .' 3
0.20 500 0.030 to 0.040 b
0.25 320 0.035 to 0.050 ‘
0.375 140 0.055 to 0,075
0.50 80 0.070 to 0.100 oy
1.00 20 0.150 10 0.200
UNCLASSIFIED
CRNL-LR—DWG 8240
1.0
0 'WLA;COIL IMPEDANCE ALONE
% = REACTIVE COMPONENT
0.9 %-RES!STIVE COMPONENT_
FILLING FACTOR=0.85
20
0.8
© INCONEL ROD
VARIABLE FREQUENCY
o IN KILOCYCLES {
53 I \ /
® INCONEL TUBING M dso !
WALL THICKNESS > EFFECT OF | #.,
IN MILS VARIABLE
07 — COUPLING
4 INCONEL TUBING '
mAbllTL%CCENTRIC]TY 0 \ Y
oo\, s o | 7h0EL 9
\\;7\20;
o 150 \ \\:-—5-2_
0.6 AN —=
T \
400
0.5
0 X o2 03 04
AR -
Wig

Fig. 6.18. Normullzed |mpedunce Plune for ‘/- -
in.-dia Inconel Rod or Tubing Obtumed wnth an_

Encnrclmg Coil.,

 
 

w

 

o SRR

 

- Selecfed (manufactured y Efecfro CII‘CUI‘I’S, Tne.y

   
   

second) is shown for each point. If a particular
frequency is chosen (80 kc/sec) and the effect of
varying the wall thickness of an Inconel tube is
investigated, the impedance curve will follow
hrough the pomts indicated by the open squares,
Hence, if / in,~dia, 65-mil-wall, Inconel tubing
were being exammed the change in the resistive

component would be a sensitive measure of a

change in wall thickness, However, changes in
eccentricity also produce changes in the resistive
component, as shown by the curves through the
triangles. Thus, in order to interpret the measure-
ments in terms of the nature of the tubing flaws,
both the changes in the resistive and the reactive
components must be measured,

There is no instrument available at the present
time which will moke an lmpedance type of
analy51s over the wide range of frequencies
indicated in Table 6.8. Work with experimental
equnpment has :ndlccted however, that such an
instrument is feasible but would require con-
snderable development work. As a result,emphasis
is now being placed upon the development of
su:table probe-type coils with which impedance
anulyses could be made without changing frequency
for each different size of tubing. In addition, it
is felt that a probe coil, because of its small size,
would be more sensmve to small flaws than would
an encnrclmg coil.

Since the eddy-current flaw-detection method

~ has many more unresolved problems than does the

method that employs ultrasonic equipment, the

" latter method has been chosen for the inspecflon

of smc“-dmmefer tubmg. The uh‘msomc equ:pmenf

- FATE SR

gives very" clecm flaw s:gnafs ondiflhlghfresoluhon

 

 

 

 

The~ ultrasonic_“equipment has

crmlluNiT

 

capable of detecting flaws which ore os “small” as

pin “holes and which penetrate no more than
0.003 in. below the outside surface. All defects
on the outer surface of the tubing that can be

 

PERIOD ENDING SEPTEMBER 10, 1955

located by dye-penetrant inspection have been
located with ultrasound. Alsa, subsurface defects
of equally small size haye been detected with
ultrasound and verified by metallographic exami-
nation. Experience indicates that defects located
on the inner surface will be similarly detected.
The conventional presentation of ultrasonic

~data is on a cathode-ray tube as the ‘*A* scan,

the vertical sweep being the signal strength and
the horizontal sweep being the distance (travel
time) from the crystal. This presentation is not
interpretable at high scanning speeds, since the
repetition rate of the flaw signals will be in
excess of 100 times per minute. Gated alarm
signals would also be unsatisfactory, since all
spurious signals, as well as flaw indications,
would pass through the gate, The less con-
ventional ‘‘B*" scan would present the defects as
a standing image of the tube cross section on a
17-in. cathodesray tube. In this case, the vertical
sweep would be a scan of the ‘A" type of
presentation, and the horizontal sweep would be
the rotation of the tube.

A ““B’* scan pattern from a tube having only

one defect in the particular cross section being

scanned is shown in Fig. 6.19. The defect, which
is represented by the diagonal line, was a non-
metallic stringer about 0.001 in. in diometer and
about 1/ in. long. This defect was longitudinal
and was located about 0.010 in. below the outer
surface of the tube and paralle! to the tube axis,

UNCLASSIFIED
" UYa16274

 

Fig. 6.19. A ''B’’ Scan Oscilloscope Trace of a
Tube.

139

 
 

 

 

Mechanical ‘equipment to rotate the tubmg cnd'”
to produce a relative linear motion between the
: tublng and the inspection device is being designed.
 This scanner will permit rapid inspection of
-fubmg, and, if it is found to be des:rable, eddy-
current and uh‘rasomc methods can be ‘used
's:multaneously. The use of the “B’ scan
~ requires an accurate signal from the rotation of
‘the tube, and one of the major problems in the
design of the scanning equipment is that of ob-
- tammg thxs s:gnal '

. C E Curhs J. R. Johnson

J. A_Grn‘fm AL Taylor
e Meta“urgy Division |
Gruph:te-Hydrogen Corros;on-Eros:on o
LR Inveshguhon '

 

Graphlfe-hydrogen ‘systems at tempercfures of
. the order of 2400°C or higher are of interest for

nuclear rocket applications. An investigation
was carried out to determine whether hydrogen
flowing at a velocity of Mach 0.15 in high-density
graphite tubing (]/ in, OD, 1 in. ID, 24 in, long)
at’ approx;mately 2400°C caused significant
corrosion-erosion of the graphite. Two types of
graphite specimens were used:  Graph-i-tite
(density, ~1.8 g/cm3), supplied by Graphite
Specialties Corporation, and commercial graphite
(density, ~1.55 g/cm3),

A schematic drawing of the apparatus used for
the high-temperature graphite-hydrogen experiment
is shown in Fig. 6,.20. The graphite specimen was
heated electrically to 2400°C, and a nitrogen or
helium atmosphere was maintained in the specimen
until equilibrium conditions were reached. The
high-velocity hydrogen was applied for 10 min,

~and then the hydrogen was replcced ‘with helium
“while 'rhe specimen was coohngr Data and resulfs
“of the experlmenfs cre given in Table 6.9.

The average corrosion-erosion for the tests on

Grcph-:-hte showed a loss of 0.0001 g/m. .sec,

It was noted that the we;gl-n‘ losses were about

~ the same wn‘h hel;um as wn‘h hydrogen, and I1‘ !S
-.’rprobable 'rha'r ‘the losses were due to water vapor

- _inthe gases, The bulk temperature of the hydrogen
- was believed to be about 1600°C, and the velocity
L was about 1900 fps, or approximately Mach 0.18.
‘It was observed that the weight losses of the

o commercial grcxphn‘e were somewhat greater that
. f"fthose of fhe Graph-l-hfe. ‘

 

Rcre-Earfh-Ox:de Confrol Rods R

Two confrol rod cssemblues, snmlar to those
prepared previously,? were fabricated for critical
experiments from a mixture of rare-ecr’rh oxides
containing 63.8 wt % Sm ,0, and 26.3 wt % Gd203,
the remainder being prlmarlly other rare-earth
oxides.,  The firing temperature of 1550°C for
these shapes was 50 deg higher than that previ-
ously used, The annular specimens were all
1 in. long with either 0.125- or 0.250-in. walls,
The 0.125-in.-wall specimens were 1,275 in, OD

CERAMl(.;. .R.IVESEJ;RCH L e gnd T.005 in, 1D, and the 0.250-in.-wall specimens o

were 1.275 in, OD and 0.775 in. ID. Figure 6.21

shows samples of the fmlshed shapes.

Calc:um Fluonde und Alumrnu Deiecfor Spucers

Flux detector spacers of calc:um fluonde and

aiummc were produced in 'rhe form of wafers

"0.190 in. thick and 0.800 in. in dlcmeter, ‘with an

indentation 0.12 in. deep and 0.747 in. in diameter.
Reagent-grade calcium fluoride was precalcined
at 1140°C and ground to pass a 200-mesh Tyler
screen; 2 wt % Carbowax 4000 was added as a
binder and lubricant. The wafers were pressed in
a hard, steel die at 50,000 psi and then sintered
at 700°C for 1 hr. The alumina wafers were
pressed similarly from fine-grained A|203, without
calcination, and then sintered at 1350°C for 1 hr,

Fluoride Fuel Pellets

The optimum pressing conditions for pelletizing
fluoride fuels in convenient shapes for loading
into a reactor are being studied. A pressure of
4000 psi was not sufficient to produce pellets of
NaF-ZrF ,-UF, (50-46-4 mole %) that had adequate
sfrength ngher pressures are bemg used in
experlmenfs now under way. .

Synthesis of Boron Compounds

Ah'invesf-igat‘ ion of the fecmblll'ryofsynthes:zmg T

Mo,B. and B,C has been started, These, com-
pounds are of interest as sh:eldmg “material.

Several samples of MogB have been prepared and 7Tk
are being analyzed. Samples of B

- 4
by heating a stoichiometric mixture of boron and

carbon to 1750°C. It is anticipated that about_
100 tb of B10 will be made into B C

 

1955, ORNL 1896 p 145.

9J A, Griffin et al.,, ANP Quar. Prog. Rep ]une IO

 

 

 

 

 

 

C were made |}

 
TABLE 6.9. RESULTS OF GRAPHITE-HYDROGEN CORROSION-EROSION EXPERIMENTS

 

 

 

 

Surface Temperature

Surface Temperature

LLocation of Thermocouple

Total Weight

 

 

 

 

 

bReoding_.obfained with optical pyrometer sighted down sight tube (see Fig. 6.20).

“Reading obtained with optical pyrometer sighted down exit tube (see Fig. 6.20).

 

 

 

 

Rund Gas in Gas Flow Time Between of Graphite Specimen of Graphite Specimen Maximum ¢ Which Maxi G | Moeximum Graphite Power Input L ¢
Nun Graphite Rate Temperature at Beginning of Timed at End of Timed Temperature of Gas Trom :c v::xlm;;: ) asd "E‘ipecimen Temperature to Specimen s os's °
o Specimen (ctm, STP) Readings (min) Period Period cC) emperature i as aine ! (°Cye (kw) pecimen
(OC)b (OC)b : (see Fig. 6.20) l (g9)
Specimen: Graph-istite (0 = 1,8 g/Cm3) |
2 Nitrogen 14 20 2245 ) 1 8.7
Helium 18.6 13 2110 1935 f |
3 Nitrogen 41 20 2260
i Helium 9.6 16 2255 2230 10.7 3.3
Hydrogen 16.2 10 2230 1830 '
4 Helium 10.8 33 20 2250 109 1.6
Hydrogen 12.2 10 2250 2220 _ * *
5 Helium 9.6 28 20 2205 } |
Hydrogen 11.2 10 2205 2090 . 11.0 1.8
6 Helium 8.8 10 2240 2035 E 11,2 1.4
Helium 4.7 17 20 1500 650 A ) ‘
- 1190 B
14.5 5 ISOQ 1860 590 A > 2450 ‘j‘_: 12.1 1.1
Thermocouple melted B o
' 9.4 23 1860 2250 840 A
Thermocouple melted B /
8 Helium 4.6 19 20 2155 1540 C 3 .
9.6 10 2155 2210 850 C
Hydrogen 2.8 5 2210 815 C > 2400 8.8 1.7
6.5 3 970 C W
4.9 2 2200 1120 C J
9 Helium 5.5 28 20 2160 1770 b
10.1 12 2160 2220 1770 F 2350 o 9.3 0.8
Hydrogen 13.3 9 2220 Thermocouple melted ’
9.8 1 2270
10 Helium 4,3 35 20 2020 2450
9.4 6 2380 2.3 4.1
: Hydrogen 11.7 10 2250 '
Specimen: Commercial Graphite (p = 1.55 g/Cm3)
" Helium 4.6 19 20 2210
11.3 12 2210 2285 2380 L
12.2 4 2285 9.3 3.3
Hydrogen 9.8 1
9.8 9 2025
12 Helium 5.6 15 20 2210
- 9.2 15 2210 2305 2530 10.8 6.0
Hydrogen 8.6 10 2260
s “Run 1 was for checking apparatus,

141
1

 

 

"w

 

ESE

Wi %

 

UNCLASSIFIED
ORNL~LR-DWG 8944
A,B,C,D,E AND F REPRESENT
THERMOCOUPLE POSITIONS FOR
DIFFERENT RUNS (SEE TABLE 6.9)

    
   

GRAPHITE EXIT
TUBE

-+ | FIREBRICK WALL

 

CARBON BLACK
FILLS 18~in. OF 20-in. DEPTH

 

GRAPHITE SIGHT TUBE

 

 

 

  
   

GRAPHITE

SPECIMEN WINDOW

- &
3,-in, WATER-COOLED A
STEEL PIPE COPPER BLOCK p2

Fig. 6.20. Apparatus for High-Terfipefuiure Graphi‘fe-l'lydr'ogen Corrbsion-Erosion Expetiment,

 

Fabrication of Dysprosium Oxide Disks

 
 
 
 
 
  
   
 
 
         
 
 
 
 
   

The fabrication of dysprosium oxide, which is
of interest for use in medsuring thermal flux in a
reactor, is being investigated by the American
Lava Corporchon under an ORNL subcontract.,
”Dlsks '0.242 in. in diameter and 0.011 in. thick

Tty T,

that contam approxlmately 1 fng of dysprosnum

'md T ¢ and the flller is electrlcu“y fused AlLO

  

and finer, and a purlty “of about 99.5%. The
principal contaminant is Si0,. These disks were
submitted to the Critical Experiment Facility for
evaluation

    

of léfire;.E‘.c:ri.B- ' Fabrication of 'Europium- 7.(')..xide Waifér-s‘

| Fié. 6.21, Annular Speéimens;
‘Oxides (Sm,0, and Gd,0;) for Control Rod As- A preliminary investigation has indicated that
semblies. wafers 3{1 in. in diameter and 0.030 in. thick,

143

oxnde per squcre cenhmeter of area have been -
prepared “The disks contain ‘an organic resin

e

 

WEEer g e

L

T e FFrmEY R

 

 
 

. to the MTR for abouf ]02]-nvt exposure.

 

ANP PROJECT PROGRESS REPORT

prepared from Eu,0, powder by preealein'ing at

1200°C for 1 hr, cold-pressing at 20,000 psi, and
sintering at 1500°C for 2 hr, are unstable in the
presence of moisture, Similar specimens prepared
from uncalcined material cold pressed as above
and smtered at 1200, 1300, 1400, and 1500°C
for 2 hr showed no signs of breakdown in the
"'presence of moisture. As soon as suitable speci-
“mens' have been prepared, they will be submlfled

4

SPECIAL MATERIALS STUDIES

~J. H. Coobs J. P. Page
- H. Inouye T. K. Roche
' Metallurgy Division
©r ot Mo R, D’Amore
= Pratt & Whitney Aircraft
B ~ Columbium Research
 The investigation of metal diffusion barriers
~ for use in the fabrication of Inconel-clad columbium
sheet was continued, »
‘ported, 0 the interdiffusion between Inconel and
columbium at elevated temperatures is extensive.
The reaction products, which appear to be nickel-
columbium compounds, are brittle, and, because of
the difference in the coefficients of thermal
expansion between the metals, separation occurs
af the interface when the composite is cooled to
room temperature, The problem is not that of
preventing the formation of intermetallic compounds,
although this would be desirable, but primarily
that of selecting a suitable combination of metals
which will remain thermally bonded in service.
This is mandatory, since the composite is being
considered for use as a heat transfer surface,

The barrier metals to be studied were selected
somewhat at random, since the phase relationships
between columbium, the possible barrier metals,
and Inconel have not been thoroughly determined,
especially as to the nature of the compounds
which are formed. Vanadium, titanium, molybde-
num, tantalum, and copper have been evaluated
as the barrier metals, _
Composites were fabricated by hot-rolling
capsules which had been evacuated at 1100°F to
at least T x 10~4 mm Hg. Rolling temperatures
between 1800 and 2100°F were used at reduction

 

 

oo _]0.1. H. Coobs, H. Incuye, and M. R. D’Amore, ANP

o .Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 140.

As was previously re-

schedules of 30 and 40% per pass. The total
reduction in thickness was about 5/1. No
particular effort was made to predetermine the
thickness of the columbium core or the Inconel
cladding. A final composite thickness of 0.020 in.
was evaluated. The barriers were calculated to
be 0.001 in. thick after being rolled, Cross
sections of strips that had been bent fo form an
involute of a circle were examined microscopically,
" The Inconelevanadium-columbium composite in

- the as-rolled condition withstood several 90-deg
'bending reversals over a sharp rudius. M:croscoplc
‘examination

showed a reaction layer at the

vanadium«|nconel interface that was less than

'0.0001 in, thick, The reaction layer was cracked
- perpendicular to a tangent to the curved surface.
‘Since no separation occurred at the interface, this
composite could probably be formed to a desired

shape by cold spinning. After 100 hr at 1650°F,
the composite separated at the Inconel-vanadium
interface upon bending, The reaction layer at
this interface had increased to about 0.0005 in.
in thickness, and the vanadium layer had decreased
in thickness. After 500 hr at 1650°F, the composite
separated during sectioning. Microscopic exami-
nation showed that the vanadium layer had been
completely transformed to an intermetallic com-
pound,

An |Inconel-titanium-columbium composite was
found to be unsatisfactory because of the formation
of a low-melting compound at the rolling tempera-
ture of 2100°F. The compound was brittle, and
separation of the composite occurred during
cooling. The experiment is to be repeated with a
lower rolling temperature,

An Inconel-tantalum-columbium composite was
prepared, but, since tantalum and columbium are
isomorphous in several aspecfs fhe composn‘e"
was not expected to be successful. Specifically,
this combination was tried to determine the effect
of the melting-point difference of the metals on
their diffusion rate with Inconel (columbium melts
at 4380°F; tantalum melts at 5425°F) It was -
also desired to compare the proper’rles of The
expecfed reaction products, ' -

Examination of the composite after rolling at
2100°F showed a 0.0001-in.-thick reuchon Iayer

at the I[nconel-tantalum lnferface, which ‘was

‘found to be brittle at room temperature. However,

separation did not occur after numerous 90-deg
bending reversals. In the unbent position, the

 

 

 

 

 

 
 

e

 

 

~ metallic compound

- under ’rempemrure cychng.

WL

o ?festmg the creep s'rrengrh of _ofh el ,
7" columbium.  This mformahon is needed for de-
“termining whether columblum may be wbshtutedimt
Wlncone[ in an ART-typeMW
- ussembly and have strengrh equal to or greater
“than that of !nconel alone. ~ Other tests will be
7 ‘which barrler 'ma'rerlcl copper””
: N_or fcmmlum, IS beh‘er su:ted for 'rhls apphcatlon.h_‘
‘Investigations of copper as “the bcmer materlai:""_
"'fwere reporfed prev10us|y 10 Tantal ‘ |
‘_dlsadvanfage of havr Gfron cro

oy

" made to determme

 

" width that will permit the welding of totally clad
++ columbium without serious heat damage to the
ST _mterfoce.

 

- composnfe showed no crccks in fhr= reachon layer.
During service tests at ]650°F fhe recchon layer
;between the tantalym cmd ‘the Inc |
""'w:th time, until in 500 hr abou__

   
 

mcreased

the rantalum had bee_n converted to cm inter-
The rantclum-columb:um
mferface showed no fendency to dn‘fuse at these

. rem percfures .

Bendlng-reversal 'res'rs ‘mode qfrer 500 hr at
1650°F resulted in no seporahon in the composne,“ '
even 'rhough ‘the reachon layer cracked perpendicu-

lar to a tangent to ‘the curved surface. No cracks
were observed in the unbenf portion of the com-
posn‘e.' It may be possuble 'rherefore to form ’rhe‘-f
|ncone|-tantaIum-columblum . composne “in the
'qs-rolled condmon and to mamtmn rhermal bonds‘:‘

Addmonal experlments are being desrgned for

  

 

for a large porhon of th

 

  

 

m. of

   

“and unclcdi R

 

PERIOD ENDING SEPTEMBER 10, 1955

In a preliminary investigation of Cb-UQ, fuel
elements the compatability of UO, and columbium

'wcs' studied at 1500 and 1832°F for 100 hr. The

specimens ‘were prepared by hot swaging 20 wt %

‘U02 with columbium powder at 1742°F. “Both
~steam-treated UO, and high-fired UO, were “used

for these tests. The mixtures were exommed in
the as- -swaged condition and after the 100-hr
heafmg period. Under all test conditions a third
phase, which was pinkish in color, was found in
the matrix., This phase appeared as discrete
particles or as a portion of a former UO, particle,
and the quantity of this phase increased with
heating time, Xeray patterns of the tested mixture
indicated that a solid solution of columbium-
uranium was present. The same phase was found

gs a network in an arc-melted button of a 5%

. Cb-95% Ualley.

-_.r.mcomposlte Tublng Fabrlcohon

   

Ex'rrusron experlmen'rs for preparmg two- and

.';rhree-piy seomiess tube blanks have been con-
_‘tmued As in prevnously reported experiments,
the layer compositions for these experiments

10

were carbon steel and type 316 stainless steel,

;The Two-ply extrusions were carbon steel clad on
. the outside with stainless steel, and the three-ply
_,extrusmns were corbon s’reel c!ad on both sides
_,.,.>_.“wn‘h stamless s'reel

o= When équal sforhng fhlcknesses were used in
i previous experlmenfs, the ihlckness of fhe extruded

  

qlso served as markers to outline the deformatlon
pattern.  In addition, the markers outlined the
]/2-in.-fhic|< disks and thus served to determine

145

  
  
   

 
 

 

 

o s bies i

ANP PROJECT PROGRESS REPORT

the final shape of the disks when théy were used

in three-ply exfrusmn.
Both flcn‘-nosed and 45-deg tapered dummy
blocks were used in front of equal thicknesses of

concentric rmgs.' In the two-ply extrusions,
devu:mons in thickness were about 15%. The

‘inner layer in both extrusions was thinner than

desired. In the three-ply extrusions, the inner

 cladding was thinner and the outer cladding
!’rhlcker by 15% than had been calculated. The
‘ "core of the extrusion showed no deviation from
the calculated thlckness.' The results obtained
. to date in attempts to extrude concentric layers
. _of equal fhlcknesses ore summarlzed in Tqble 6 ]0.

_"_‘:__i'-:were sechoned lengfhw:se and examined. The
" "'deformaflon across the tube wall was complex,

since both compression and tension were evident.

Tensile forces prevclled in the outer and inner

layers, while compression forces prevailed in the
zone in between. The original 0.010-in.-thick
molybdenum washers varied in thickness from
about 0.003 in. near the tube walls to 0.050 in.
at the middle of the tube wall,

The displacements of the mo[ybdenum washers
from their starting positions were noted. After
extrusion, the washers could be located as
coordinates of a parabola of revolution, with the
axis being the center line of the tube. The shapes
of the separate washers differed along the length
of the extrusion. It is evident that because of the
tensile forces near the tube walls the flaws in
this zone in the starting billet will result in flaws
in the extrusion,

TABLE 6.10. RESULTS OF ATTEMPTS TO EXTRUDE CONCENTRIC LAYERS OF EQUAL THICKNESSES o

 

Oxidation Tests of Alummum Brouze -

The aluminum bronzes are bemg considered as

cladding material for the oxidation protection of

copper fins for radiators, Therefore oxidation
tests of several commercial bronzes were conducted
in air at 850°C. The weight gains of the alloys
as a function of time are shown in Fig. 6.22.

Lead-Calcium Alloys

The chemical-analysis technique for evaluating
the creep resistance of lead-calcium alloys was
investigated. It was determined that chemical
analyses were not indicative of the creep per-
formance of these alloys, because the normal
scatter of the data on the calcium content frequently

exceeded the desired calcium content of between

0.03 and 006%. The strengthenmg effect of

calcium can be realized only if it is present in
solid solution or as a compound with lead; that
is, if a portion of the calcium is presem‘ as Ca0,
the beneficial effects of the calcium are lost,
Since the chemical analysis gives the total
calcium rather than the CaO content of the alloy,
the chemical analyses can serve only as a guide

in the evaluation of the creep performance. For

example, three alloys calculated to contain 0.06%
calcium on the basis of analyses of three master
alloys containing 2% total calcium were found by
chemical analysis to contain 0.043, 0.030, and
0.059% calcium, respectively. In cantilever creep
tests at room temperature and 750-psi stress, the
deflections of these alloys at the end of a 2-in.
beam were 0.87, 0.33, and 0.54 cm, respectively.
Since the chemical anclyses defermme only fhe

 

Tubing  Type of Dummy Block

~ Average Layer Thickness

Length of Extrusion

 

 

 

Containing Uniform Layers
Used Inside Layer Core Outside Layer (@) e o
Two-ply . None 0.087 0.132 T 45 T ; CLE
7 Flat nosed 0.109 0.140 - éa - -
h 45;d?g'-fapered ' 0.107 0.1 18" o Ce6 T - :
Three-ply Norne 0.050 0.073 0.111 s
St Flatnosed 0,070 0.074 0.103 Css
.. 45-deg tapered © 0.078 0.075 0.104 56

 

 

 

 

  

  

   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
    
 
 
  
   
  
  
 
 
  
  
  

 
.y

 

 

 

)

 

PERIOD ENDING SEPTEMBER 10, 1955

UNCLASSIFIED
ORNL-LR-DWG 7388

 

T T T T T T

 

 

 

 

10 |- MATERIAL COMPOSITION (wt %)
: e W W ALLOYS 5.4 Al—~BALANCE Cu
02 =10 712 ALLOY 4.33 Al-1.22 Si~BALANCE Cu

 

 

A DURONZE NO. 4 6.2 AlI-0.21 As~0.2{ Zn—BALANCE Cu

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

08 = [ A AMPCO NO. 8 9.49 Al-1.94 Fe—BALANCE C —
mO : - .__..—-—""'"'_——_
: o7 . .—-—-‘,—" /___.—"O
—_ ' §
< o0s ' e | AT
s “ P ——
<Z“E -, 4=
= ‘ A
g 04 _,// R ‘/
oo A /O/' C A
3 03 P / ] K P

> / /%’ M“'"&-

A
0 :
0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 10 20 30 40 50 60 YO 80 90

100 110 120 130 440 150 160 170 4180 190 200 210

ELAPSED TIME (hr)

Fig. 6.22. Oxidation of Commercial Aluminum Bronzes in Air ot 850°C.

total calcium, the master alloy must be made
under conditions in which the oxidation of the
calcium will be held to o minimum, It might be
possible to improve the accuracy of the alloyed
calcium determinations by filtering the master
alloy at about 1475°F through a quartz filter
prior to analysis. ,

About 150 Ib of a lead-calcium alloy containing

0.05% calcium was‘.sppphed to WADC for elevated-

ifemperature U epw' fes'rmg." Chemical enalyses
e alloys ~
“reduce the neutron flux through it about 90%.
Since such a relahvely thin layer would meet
"»'“eoctor spec:f:cahons, if therma”y bonded to the
~Inconel, experiments for depositing boron by
_‘ electrophores:s  and electroplatmg of boron

slurrles are being contemplated Unbonded tiles

ot sfrtps “of boron compounds in a metalllc matrix
" would then be used to fill the remaining space of
+ the ¥%-in. annulus provided.

  
  
  
 
 

 

N “contenfrs of t

 

T material is bemg studied, A material contommg a"

minimum of 1 g of boron per cubic centimeter of a
metallic matrix is to be bonded to a l/4-in.-th'icl<
Inconel hemisphere.  The initial experiments

were prepared from
“had” been profected from .

   
 
 
 
 
  
 
 
  
  

were attempts to obtain thermal bonds by flame-
spraying mixtures of Cu-B,C. The results, to
date, do not appear to be promising, since a large
proportion of the B, C is lost during spraying.

in other experiments, attempts were made to wet
B,C with copper alloys and to cast the resulting
slurry. The oxidation of the B,C to B,0, during
melting appears to be a ma|or problem in this
opprouch The experiments are fo be continued
in an inert ‘atmosphere.

A |ayer equwalent to 0.008 in. of boron will

Adisk and a ring were fabricated from a boron

'_ eerblde—copper mixture for use as neutron poison
“in the high-temperature critical assembly. The

disk was to be 63/ in. in diameter and 0.500 in.
thick; the ring was to have an outside diameter of
]2]4 in., and a cross section of /16 X ’/‘, in.

147

 

 

 
 

 

 

ANP PROJECT PROGRESS REPORT

The disk was formed by pressing a mixture of
B,C and copper powder in a graphite mold at
900°C, and it was found to be 6]/2 in. in diameter
“and 0.447 in. in average thickness; it contained
0.367 g of boron per cubic centimeter, or 97.4%

of the theoretical density. The disk was machined

_to 6% in. in diameter with diamond tools, and
diffusion barriers of copper and stainless steel
were applied by flame-spraying. ~ The barrier
_layer thicknesses were 6 and 15 mils, respectively,
'~ Attempts to fabricate the material for the
]2Z-in.-0D ring by extruding a billet of compacted
B ,C-Cu powders canned with copper were unsuc-
cessful. Severe edge cracking of the material

| _}48_. -

occurred during extrusion and was probably due
to either too high an extrusion temperature or too
thin a copper can, After the unsuccessful
extrusions, the ring was fabricated by drawing a

round copper tube filled with B,C powder into a

rectangular tube with a % . x % in. cross section,
The tube was then roll-formed to a IZZ-in.—OD
ring and split into two equal segments for ease
of assembly., The open ends of the tube were
closed with boron carbide—copper plugs. The

ring contained approximately 2.25 g of boron per

inch of length. A 7-mil layer of type 430 stainless
steel was flame-sprayed onto the ring to act as a
diffusion barrier,

 

 

 
PERIOD ENDING SEPTEMBER 10, 1955

7. HEAT TRANSFER AND PHYS!CAL PROPERTlES

o & A s
0

H F Poppend[ek

Rec:cfor Experlmental Engmeermg Dlvns:on

 

{ - Forced-convectlon heat trqnsfer expenments ('l] 2—4]-45325 mole %) flowmg in a heated ‘
' wn‘h the fuel mixture ‘NaF- KF LIF UF, flowing in  Incone! tube gave results which were 40% below ..~
S a heafed tube were conhnued and heat transfer  the general turbulent-flow heat transfer correlation.

f " and friction charactenshcs of a full-scale ART These results (Fig. 7]) have been duplicated :
' , _hea'r exchanger were determined. Flow pcrh‘erns in recent experlmenfs in whxch a type 316 stainless E
. . in models of the 18- and 21-in. ART cores have steel tube was used. It was not possible to :
been determmed for rofoflonai and axial flow, as  examine the inside surfoce of the stainless steel 3
well as various en’rronce condlhons, ‘and the  tube, because the experiment was terminated by ‘
. results are summarized. The ’remperature distri-  the melting of the test section, However, it is F
{ . butions within fluids flowmg through converging believed that there were probably no surface
and dlverglng channels “were determ:ned experi- depos:fs on the stamless steel tube, which was 1
mem‘a”y in the volume-heat-source system. the same case in earlier experiments with :
The em‘halples and heat capacmes of LiF-KF NaF-KF-LiF (11.5-42-46.5 mole %). Thus, the
- S (50-50 mole %) were determined, and the vuscosmes' reduction in heat transfer obtained with this E
o T of elght fluoride mixtures were measured fluoride mixture is probably due to something 1
{ FUSED SALT HEAT TRANSFER =+ other than the formation of surface deposits.
UL S S Some evidence exists which indicates that the
N _ H W. Hoffman P E. Stover NaF-KF-LiF-UF | composition contained particulate E
s Reactor Experlmental Englneermg Drv;s:on B matter that made it a dilute slurry. However,
s Prev:ous forced-convechon heat h'cnsfer experl- before any conclusions can be reached as to the
'y '-“">","'-""ments with the fuel mixture NaF-KF- LIF-UF4 effect of this feature on heat and momentum

 

 

}  ooi0 A e o o ORNL-LR-DWG 8942
o] B G NoF-KF-LiF —UF, IN INCONEL

S ® NoF-KF—LiF—UF, IN TYPE 316 STAINLESS STEEL
A4 Hy0 IN TYPE 3{6 STAINLESS STEEL

 

 

 

 

 

 

 

 

 

 
   
   
 

g " |- GENERAL CORRELATION

L i » =T T 2 /=0.023 Mg 02
LT AT ‘s

7 | S L A T
, Lo EL A

3

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i
4"‘,' L o ; _1‘ ) :F.ig.r 71 Compunson of Heat Trunsfer Meusurements on NoF KF-LiF- UF, (11, 2-41-45 3 2.5 mole%) and
l AR Wm‘er w;th the General Correluflon for Ordmury Fiuxds.

 

 

149

    
 

 

 

" ANP PROJECT PROGRESS REPORT

transfer, further experiments must be made. Heat
transfer studies in which a Hostelloy B test
sechon is used are now in progress.

As a check on the over-all performance of the

_"experlmen’rol cppcra’rus used for these studies,
the system was operated with water as the heat

’rransfer medlum. The water data, shown in

‘Fig. 7.1, are in ‘good agreement with the general
,:,‘_,"turbulent-flow _correlation’ and  indicate  that
~the low heat transfer values obtained with

"‘NaF-KF-LlF UF are real.
- consnderable scah‘er in the water results has not
ye’r been determined.

Assembly of a heat transfer system that includes

S g pump is in progress ‘and will be completed in
~ ‘the near future, The system will be opera'red with
"fwc'rer‘prlor to “studies  with the salt ‘mixture

NoF-ZrF UF (50-46-4 mole %).

ART FUEL TO NaK HEAT EXCHANGER

J L. Wantland
Reactor Experimental Engineering Division

For the recent experiments with the ART fuel-to-
NaK heat exchanger! the tube bundle was heated
by passing an electric current through it; water
was circulated outside the tubes, but there was
no fluid flow Through the tubes. The fuel heat
transfer coefficients were determined by measuring
tube-wall and mixed-mean-fluid temperatures, as
well as heat transfer rates. As is shown in
Fig. 7.2, the data obtained corroborated the heat
transfer coefficients previdusly determined by
using the apparatus as a water-sto-water heat
exchanger.

Also, the heat transfer and isothermal friction
characteristics of the fuel side of the heat
exchanger were determined with all the tube
spacers removed from the tube bundie, except for
one horizontal and one vertical spacer at opposite
ends_.' In Figs. 7.2 and 7.3 the data are compared
with previous data taken with all the spacers in
the heat exchanger. As was to be expected, both
the heat ftransfer coefficients and the friction
factors decreased upon removal of the tube
spacers,  The difference is partially due to the
spacers creating form drag and inducing some

__additional turbulence, However, when the spacers
' were removed the tubes were not rrgldly held, and

 

15, L. Wantland, ANP Quar. Prog. Rep. June 10, 1955,

 
 

© ORNL-1896, p 149,

The cause of the

“*channeling’” occurred in the flow pattern. The

flow channeling increases the effective equivalent
diameter (and hence the Reynolds modulus) and

ORNL—-LR—DWG 8943
100

Nu/Pr%= 0.023 Re

50

20

Nu/Pro?
S

SPACERS

e WATER-TO-WATER HEAT EXCHANGER

2 4 RESISTANCE-HEATED TUBE BUNDLE
WITH NO FLUID FLOW THROUGH TUBES

 

{000 2000 5000 10000 20,000 50,000
REYNOLDS NUMBER, Mg,

Fig. 7.2. Heat Transfer Characteristics of the
Fuel Side of the ART Fuel-to-NaK Heat Exchanger.

ORNL-LR-DWG 89844

0.5

 

&
S 0.2
g .
b SPACERS
8
P 0.4
Q
n
lL-
*0.05 T SPACERS
LAMINAR FLOW IN
A SQUARE DUCT.
L= o 25
0.02 TURBULENT FLOW
IN A SMOOTH DUCT
0.01

100 200 500 1000 2000 5000 10,000
REYNOLDS NUMBER, Wg,

Fig. 7.3. 1sothermal Friction Characteristics of

'fhe Fuel Side of the ART Fuel-tosNaK Heat Ex-

changer.

 

 
 

 

 

 

decreases the efféctive surface area for heat
transfer. Since the cmoun’r of channeling was not
known, it was lmposs:ble to determme the effect

of the presence of the spacers on fhe hecn‘ fransfer '

and frzchon characterlshcs. 7

 

   

ART CORE"HYDRODYNAMICS
U FUE. Lynch o
Recctor Experlmentcl Engmeermg Dw:snon

‘ G L. Mu”er
Pra’rf & Whl'mey Aucrcuft

 

 

Sfudies of fhe fiow feafures of a serles of one-_

quarter-scale ART core models have been under
way for the past six months. The specrflc fypes
of cores and flow condmons studied, as well as

‘the kinds of entrance condmons, are llsted in

Table 7.1, The hydrodynamlc s'rruc’rures were

- studied wnh quahfchve, as well as quanhfotwe,
'7.'_techn‘|ques over the Reynolds modulus range
3,000 to 40,000. ln all cases, some flow sepa-

L rohon, flow sfognc’non, or transnen’r flow conditions
" were observed. In general, rotational flow yielded
. flow separohon on the |sland wall, and axial How

gave flow seporahon on “the outer wall vanes

usually produced transient flow.

    

PERIOD ENDING SEPTEMBER 10, 1955

Recently, two cores with constant spacing
between the inner and outer walls were studied.
One of these, in which the ratio of the flow cross-
sectional area at the equator to the flow cross-
sectional area at the inlet was low (1.44), was
characterized by uniform and steady flow,

Research on a variable-geometry diffuser has
been initiated. A plastic housing for the flow
system has been completed, and the templates
for the channel are being designed. The flexible
walls for the divergent channel are being fabri-
cated,

REACTOR CORE HEAT TRANSFER

N. D. Greene H. F. Poppendiek
L.D. Palmer

Reactor Experimental Engineering Division

The temperufure structures within fluids flowing
through short, converging and diverging, plastic
flow channels were experimentally determined in

‘the volume-heat-source system (Fig. 7.4). The

heat sources were generated electrically by a
high-voltage power supply. The channei walls,
which were not cooled, were instrumented with
thermocouples that were located about 30 mils

_TABLE 7.1. SUMMARY GF ART CORE HYDRODYNAMICS STUDIES

 

Core Diameter

. | Type of Flow
(in.)

Entrance Conditions

Flow Features

 

   
 
   
   
 
     

 

18

Rotational (45 deg}

 

_Rotational

No vanes or screens

    

Wislicenus vanes

 

" Flow Separotion on outer core shell

Flow separahon on lslond

 

 

 

Flow separation on island and flow

transients

 

~

 

 

151

 

‘Flow sepqranon on outer core’

 

e it

 

 

 
 

 

— . o

o,

W

7 Sl

O e
IR

=

g

S

¢

E-'l‘.

<.

BX:
BRI,

KOOI
PO ..

a

 

 

W X) .
S A x
a’ ANPANPS o
~ %,

 

 

 

 

UNCL ASSIFIED
PHOTO 24441

 

 

 

1¥0dIY SSIUO0YUd LIFr0¥d dNV

 
v M

 

 

 

" conditions of the experlmenf
- profiles that make vp each of these flgures show

below fl"ler_’s;d-rche. Some typical fluid and wall

temperature distributions for the 16-deg diverging

and converging chunnels cre shown in Figs, 7.5
and 7.6, which also give some of the specific
The two sets of

some typlcal femperature distribution extremes,
It was possflole to make the following observations
about the nature of fhe thermal structure in the
uncooled chcnnels (1) “axial wall temperafure
profiles exhnbn‘ed large degrees of asymmetry;
(2) the wall temperatures fluctuated s:gnn‘:contly

‘with time, while the mixed-mean-fluid temperatures
into and out of ’rhe test section, as well as fluid
flow rates, were constant with time; and (3) the

radial temperature differences for the convergent

PERIOD ENDING SEPTEMBER 10, 1955

flow channel were greater than corresponding

differences for the divergent channel. The first
two observations were expected because corre-

sponding velocity variations were observed in the .

hydrodynamic field. The third observation was in
agreement with previous temperature calculations

‘that had been made for converging and diverging

channel systems.

A one-half-scale model of the ART core is being
designed for insertion in the volume-heat-source
system. An electrode system has been devised
whereby nearly uniform or sinusoidal electric flux
fields can be generated in the flowing e[ec’rrolyte.
Wall temperatures for the uncooled walls dre to be
measured, In this way the influence of the complex
hydrodynamic features of the flow on the thermal

UNCLASSIFIED
ORNL—LR—-DWG B945

 

 

 

 

46
. O O

45 : # /

‘ | } P
/ d

44 Of‘ /I o

, /7 1° / /
. / / /o

43 / P [/ P

! / / ] 7 7
. O / I
/| / / /’ /
42 /

 

 

Ld /o
/
. ) / / /
. . /
41 7 /
’/
‘/ / /
40 - A/

 

 

 

 

 

 

S
§ MIXED MEAN FLUID
,P d E ,l/f
W _ /ly ——=— ONE WALL /9
2 39 : L / — = mw QOTHER WALL 0’/,/ /
& /
i , N ;?\,O/ / ‘ Re= {8,000 Pr=8 _ 5 /
L Y LNO WALLCOOLING /( /

 

 

 

 

 

 

 

™
AN
N

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fig. 7.5. Experimental Temperature Structure in a Divergent Channel Huving'u Uniform Volume Heat

Source.

 

153

 

 
 

 

 

'ANP PROJECT PROGRESS REPORT

 

 

 

ST yNGLASsIRED
" ieliliiere ORNL—LR—DWG 8946
S g e _ iz ORNL= > 89
e
o \\
e \

 

 

(%‘
ks
‘l ~
A
b
1
1’(5
O
L

 

 

N\ | A
& v/ \o._\ ‘ \)
\ A,
‘\ . ‘\
’ N\
\ \
\ \\

 

 

 

 

 

\ \‘o\ -

— v v——

d

 

 

TEMPERATURE {°C)
W
fl
/ i

: \

MIXED MEAN FLUID
ONE WALL \

OTHER WALL ™

Re= 18,000 Pr=8 \
NO WALL COOLING —

 

 

 

 

 

 

 

 

34 \

TIME, 11:28 \

TIME, 14:28:30 \

 

‘

 

32 \

 

 

 

 

 

 

 

 

 

 

 

 

 

 

) W«—/ 1
30 ' L

 

 

 

 

 

 

 

 

 

POSITION

Fig. 7.6. Experimental Temperature Structure in a Convergent Channel Having « Uniform Volume Heat

Source.

structure can be determined for a series of entrance
conditions. '

A report has been prepared which describes
applications to more general convection problems

of previously developed mathematical temperature
" solutions for forced-convection systems having
‘volume heat sources within the fluids. Convection
" solutions are tabulated so that it is possible to

determine the detailed radial temperature structure

~within a fluid having @ uniform velume heat source

and being uniformly cooled at the duct wall; the
detailed temperature profile of a specific system

- is presented. The derivations of equations
S dé‘sérrri_bing the temperature structure and heat
.. transfer rates in a duct system in which the wall

  

POSITION

is nonuniformly cooled are also given, and the
temperature structure of a specific heat exchange

system is presented.

HEAT CAPACITY
W. D. Powers

Reactor Experimental Engineering Division

The enfhalpy- and the heat capocny of LiF-KF

(50-50 mole %) were determined by using the
copper-block calorimeter, The results are:

Solid (107 to 466°C)
Hp - Hyo o = -9.38 + 0.2817T + (3.82 x 10~5)72
c, = 0.282 + (7.64 x 10=5)T

 

 

FETY T

 

 
L)

 

i LR AR, i i e 5 enmial+ e e sk s e o

U

MLEARAC || G s, AN L i sbmin L St ...wum‘ - el

 

N »

"

 

_L'iqu'id (532 to 893°C)

) HT"H25°C = ~30.85 + 0.'583“-)"1“';(]0.28 o 10‘5)T2

= 0584 (20 56 x 10~ ")T2
f*- 93 Cfl' 4920(:

" n \‘hese expressnons

25%¢ = enthalpy in cai/g,

“p
AH], = heat of fusion in cal/g,

' = heat capacity in cal/g-°C,

T = temperature in °C,

The facility for defermining heat capacities has
been completely renovated, and it is now being put
back into operation. Three copper-block calorimeters
will be used in conjunction with four furnaces.
Preliminary designs have been made for a furnace
and o calorimeter to be used with beryllium-
containing materials, Complete protection against
exposure to beryllium will be provided.

A report is being prepared that lists all the
heat capacities and enthalpies that 'have been
determined for the fluoride mixtures. General
equations have been developed so that enthalpy
and heat capacity predictions can be made.

VISCOSITY
S. 1. Cohen

Reactor Experimental Engineering Division

Viscosity studies were carried out on eight
fluoride mixtures. The results are presented in
Table 7.2 and in Fig. 7.7. Most of the data can

be expressed in fhe form

 
 

  

" cen pmses Cand T in °K7N

 

where p. 15

. 4equahon :s hsted |»n Toble 7 2 f?F Sdlf C becau53\‘"‘:‘“;

from an ‘avercge llne” through the= “data did nof""" ‘
'exceed +12%) Sahs d, e [ and & which con-

" “tained BeF., were studted in q separate dry box
“Used only for BeF2 mixtures, Measurements were

mcde on ’rhese salts with two capillary viscometers

"to furnish checks. Salt g g appcrenfly had very high

surface tension and nonwettmg properties and thus
would flow through the capillary; consequently,

5 VISCOSITY (cp)

PERIOD ENDING SEPTEMBER 10, 1955

no results were obtained on this mixture. It did
not have the turbid appearance characteristic of
other BeF, mixtures but was clear, glassy, and
full of air bubbles. lts viscosity appeared to be
high.

The low viscosity values obtained for salt &
suggested the possibility that research with

. systems containing RbF might produce a satis-

factory fuel with a viscosity lower than that of
the zirconium-base fuel now being considered for
the ART. Salt b, which is, approximately, a
prototype of the proposed ART fuel, with RbF
substituted for NaF, was therefore investigated,
It was found to have a kinematic viscosity about
20% lower than that of the corresponding NaF
mixture, Further research and the use of higher
purity RbF might produce fuels with even lower
viscosities.

The results of examinations of seven fluoride
mixtures  containing BeF, are presented in

Table 7.3. Salts 4, e, f, and g were studied at

UNCLASSIFIED
ORNL-LR-DWG 8947

TEMPERATURE (°K)
700 800 200 1000 1100 1200

 

© 400 500 600 700 800 900 1000
TEMPERATURE (°C)

Fig. 7.7. Viscosities of Fluoride Mixtures.
Compositions of the mixtures are given in Table

7.2,

155

 

 

|

 

 
" ANP PROJECT PROGRESS REPORT

 

1

9 o . :
4 ~ORNL, and salts 7, j, and & were investigated at  mixtures ‘previously studied. Salt i appears to be

] Mound Laboratory. Tabulated with the compositions  the most favorable BeF, mixture, from a viscosity

; are the viscosities at '700°C and the BeF . content  standpoint, that has been found thus far. |ts :
j _in weight per cem‘ Figure 7.8 shows a plot of  kinematic viscosity at 700°C is 1.84 centistokes, :
3 the viscosities vs the corresponding weight  which is about equal to that of the NaF-ZrF -UF,

j | percentage of BeF,, as well as the data for  mixture currently planned for use in the ART.

" TABLE 7.2. SUMMARY OF CURRENT VISCOSITY STUDIES

 

 

Yiscosity

 

 

   
 
 

 

 

 

Composmon '(rh'ol'.e'..%r) o 1 Reference
NaF KE-LIF-UF,  areoc, 5.8 0.0319 455877 (@
o (112-41-45 3-2. 5) . Ar750°C, 27
5 o ‘_'""':"RbF LiIE  Arso°C, 9.0 0.0223463/T )
SRR (57-43) - - At 650°C, 3.4 - ' ’ ‘ o
3 &  LiF-NeF-ZeF SUF,  ALS00°C, 20.0 | e
(35:32-29-4) At 800°C, 4.6
d  NaF-LiF-BeF, At 575°C, 8.2 0.0784 ¢3944/T @
(64-5-31) At 850°C, 2.65
e KBeF, At 570°C, 20.0 0.00443 <7096/T (e) o "
At 800°C, 3.3 |
/ NaBeF, At 600°C, 15.0 0.0411 &5148/7 ()
At 800°C, 5.0
g ' LiF-BeF2
(50-50)
b RbF-Z+F -UF, At 550°C, 9.5 0.113 3648/T n
(48-48-4) At 850°C, 3.1

 

 

Previously unpublished data,
Previously unpublished data. Composition shown here is the nominal composition. Because of the KF present in
the RbF used 'ro.prepare the mixture, the actual composition is RbF-KF-LiF (45.6-11.4-43 mole %).
S, I. Cohen and T. N. Jones, Measurement of the Viscosity of Composition 86, ORNL CF-55-7-33 (July 7, 1955).
dS l. Cohen and T. N. Jones, Measurement of the Viscosity of Composition 88, ORNL CF-55-8-21 (Aug. 15, 1955)
€S. l. Cohen and T. N. Jones, Measurement of the Viscosities of NaBeI:"3 and KBeF g and Some Observatzons on
(LzF-BeF2 50-50 Mol %), ORNL CF-SS 8-22 (Aug. 15, 1955). ' 4
Prevrously unpubhshed data, Composition shown here is the nominal composmon. Because of the KF present in
the RbF used, the actual composmon is RbF-KF-ZrF -UF (38 4-9,6-484 mole %).

   

 
 

 

j 7 : _ PERIOD ENDING SEPTEMBER 10, 1955
?
1

i, | TABLE 7,3_. ] CURRENT VISCOSITY STUDIES ON MIXTURES CONTAINING BeF,
e . . ‘- . ) . . C ; n
; | - . ) _ Mixture ACOmposifion (mole %) VISCOSIZPT 700°C BEF&,::;TB ! Reference
4 NaF-LiF-BeF, 4.55 34.1 (@)
(64-5-31)
| o - e KBeF, ” 44.7 %)
;o NaBeF, ' 8.1 52.8 ()
g LiF-BeF, 64.5 ' (b)
5 (50-50)
: : NoF-BeF, 3.7 32.6 ()
(89, 8-30 2)
A O ""NaF-L.F-BeFZ' - 4.4 46,7 ()
. FEEE S - T (27-35-38)
' B NaF-LiF-BeF jUF, 4.9 | 38.6 ()

- {26.3-34.1-37.1-2.5)

 

s, 1, Cohen and T. N. Jones, Measurement of the Viscosity of Composition 88, ORNL CF-55-8-21 (Aug. 15, 1955).
S. |. Cohen and T. N. Jones, Measurement of the Viscosities of NaBer and KBeFq and Some Observations on
(LzF—BeF2 50-30 Mol %), ORNL CF-55-8-22 (Aug. 15, 1955).
“Personal communication from J. F. Eichelberger, Mound Laboratory, to W. R. Grimes, CF-55-5-3, May 2, 1955.

Personal communication from J. F. Eichelberger, Mound Laberatory, Aug. 1, 1955.

. ) 7 UNCLASSIFIED THERMAL C_ONDUCT!VITY
. : Tl ORNL-LR-~DWG 8948
] 20 _ W. D. Powers

" ® Reactor Experimental Engineering Division

 

B

N The thermal conductivities of molten fluoride
" mixtures NaF-ZrF -UF, (50-46-4 mole %) and
NaF-KF-LiF (11.5-42.0-46.5 mole %) were measured
o with an alternate type of thermal-conductivity
device, namely, a constant-gap cell. The results
® were in agreement with values previously de-
termined by the variable-gap apparatus,

e The thermal "conductivity of RbF- ZrF -UF,
0d - o/ | (48-48-4 mole %) is currently being meusured
Also, the thermal-conductivity apparatus which
will be used to study solid lithium hydride has
been completed.

sty o i,

 

 

 

 

 

 

S e o Rk i MM s W B
O

 

 

- VISCOSITY (cp)
®
-

 

 
 
 
 
 
 
 
     

 

 

 

 

O MIXTURES RECENTLY’ STUDIED , -
2 @ MIXTURES PREVIOUSLY REPORTED 5 ELECTRICAL CONDUCTIWTY )

(ANP QUAR PROG REP, JUNE 10, 1955, ORNL 1898, . 159) , N. D. Greene ,
) ' Reactor Experimental Engmeering Division

The construction of the experlmenfa] platinum
conductivity cell has been completed, The
0 4 . s0' e 70  inclusion of the fourth electrode should permit
éF, CONTENT OF MIXTURE (wt %) L re lsi'wn‘y measurements to be made mdependently
f the effecfs of polarization at electrode surfaces.

 

 

 

 

   

 
 

i Fig. 7.8. Viscosities at 700°C vs BeF, Content After testing and standardization of the cell,
. of Fluoride Mixtures Containing BeF,. Composi- measurements will be made of several fluoride
tions of the mixtures are given in Table 7.3. salt mixtures,

157

 

 

i v o

 

 

 

 

 
 

 

 

 

 

 

“ANP PROJECT PROGRESS REPORT

8 RADI AT!ON DAMAGE

D S Bl”lngton

J. B, Trice

Sohd S’ra're Division

The results of metallographlc examlnahon of fhe\
'fluoride-fuef |oop recem‘ly operated in the LITR

horizontal beam’ hole are presented. Corrosion of
the Inconel tubmg by the c1rcu|atmg-erl mixture

was, found to ‘be lqw. The mlnlofure in-pile loop
" wads operaied for a short time in a vertical hole
of the LITR, but the experlmenf was terminated |

because of difficulties caused by failure of the

- -7",:'pump motor. Modlflcchons to be mcde in the loop |
" are descri bed. N

Spec1men “assemblies bemg prepared for stress-

 ;:-:.;'_f_,;'_corrosnon experlmen’rs m the LITR are described,
-~ as well as bench tests of an apparatus for creep
tesfs in fhe MTR Flux measurements in the MTR
. -_f-_f_’are ‘feported, and the results of analyses of ir-
" “radiated reactor-grode beryllium are gwen.

LITR HORIZONTAL-BEAM-HOLE
FLUORIDE-FUEL LOOP

C. Ellis
J. G. Morgan

0. Sisman

C. D. Baumann

W. E. Brundage M. T. Morgan

R. M. Carroll A. S. Olson

' W. W. Parkinson
. Solid State Division
The previously described! loop in which fluoride
fuel was circulated in a horizontal beam hole
(HB-2) in the LITR was examined visually, metal-
lographically, and chemically for effects of the
irradiation. During operation of the loop the maxi-
mum fuel temperatures were 1450 to 1500°F; the
fuel Reynolds number was 4500 to 6200 or about
5000 for most of the operation; the power gener-
ation in the fuel was about 2.8 kw; the power
density was 0.4 kw/;m3; and the duration of oper-
afion was 645 hr, with the reactor at full power
(3 Mw) for 475 hr.
From flux measurements in the empty beam hole,

it was eshmm‘ed that the power generated in the
loop should be con5|derab|y higher than the 2.8

kw obtained from heat balance measurements, and
BE V,_fi__fherefore an oddlhonal experiment was carried
out ’ro meosure the flux depresswn m ’rhe fuel

 

 
 

B O Snsmcm et al ANP Quar. Prog. Rep. ]une 10,
T 1955. ORNL 1896 P 163

 

e

'in a simu!rated loop nosepiece. An assembly that

duplicated the irradiated portion of the loop was
made up and was filled with a 2% Cd-98% Pb
alloy, which had the same macroscopic absorption
cross section as the fuel salt. This assembly
was inserted in a water-cooled jacket in hole
HB-2, The neutron activation of the cadmium-
lead alloy was measured, and the relation between
the activation and the neutron flux was established
by calibrating the data for the alloy with cobalt-
foil data. Because of uncertainty regarding the
position of the fluoride fuel loop in the hole, it

was not considered that this experiment gave a

reliable measurement of the flux, however, the

flux distribution found did agree ‘with that cal-

culated from the heat balance data and that de.
termined subsequently from the cobalt foils that
monitored the fuel loop.

After disassembly of the loop in the hot cells,
a measurement of neutron flux was obtained from
the activity of cobalt monitor foils which were
at various points in a copper tube attached to
the nose of the loop. Since these foils were of
necessity outside the fuel tube, it was also neces-
sary to measure the activity of the Inconel fuel
tube. The neutron flux was calculated from the
data by using the relation, measured by Bopp,?
between the activity of the Inconel and the flux.
The neutron flux data obtained from the activities
of the cobalt foils and the Inconel are plotted
as functions of distance along the loop in Fig. 8.1.
The fission power calculated from the Inconel
activation was 2.8 kw, and the maximum power
density was 0.4 kw/em?3.

The fission power generated in the fuel was
also determined by measuring the activities of
fission products in the samples taken for chemicadl
analyses., The octivity of the fuel was obtained
with a gamma-ray spectrometer, and the height
of the zirconium-niobium peak was compared with

that of a standard. The zirconium-niobium concen-

tration corresponded to a fission power of 2.0 kw,
The zirconium fission product and the cesium
flssmn product were qlso chemlcolly separc’red

 

C D. Bopp, Gamma Radzatzon Induced in Engzneermg 7

Materials, ORNL-1371 (April 16 1953) B

 

e

 

 

 
) ay

A

 

 

NEUTRON FLUX (neutrons/om- sec)

2 3 cnd 2. 5 kw.'_

ORNL~LR~DWG 83834

e COBALT-FOIL ACTIVATION
4 INCONEL-TUBE ACTIVATION

 

2° 4 & 8 10 12 14 46 18 20
DISTANCE"FRO'M' EN'D"OF Loop (in.) :

 

Flg. 8. 1

from two of fhe scmples, ‘and the flssmn rate
wds eshmated frorn fhelr CIC‘I’IVIi‘IG'S. The cesmm
samples produced’ estimates of 1.5 and 2.1 kw,

 which were considerably lower than the determina-

tions by other methods. The power indicated by
the C5137 content could be expe=c'red to be low,
since it is known that the gaseous parent, Xel37,

"esccpes from the fuel mixture. The two separated

zirconium samples mdlca’red fission powers of

 

 
  
     
 
   

Thermul-Neutron Flux 'Traverse Along h
, LITR Horlzonful-Beam-Hole Fluorlde-FUel LOOp, )

. UF, (47 wt %), 62.5 mole % NaF, and 12.5 mole %
_A"ZrF and the enrichment was about 93%.
be seen that the preliminary flushing with NaF-

|n comparlng the resulfs of the |

| :“'cofed by the elecfrlcal measuremenfs and' by ‘the
' flux defermmchons, |f should be kept m mmd

7 swn reuchons. 7

E’xamznazzon ‘

of
‘Number 4695-1, ORNL CF-55-3-179 (March 28, 1955).
""":“"‘4G M. Adamson and R. §. Crouse, Examination of

PERIOD ENDING SEPTEMBER 10, 1955

the end of the loop was about 0.4 kw/cm?®.

The tube sections cut for chemical sampling
were taken from both the in-pile and out-of-pile
parts of the loop to ensure representative sampling.
Fuel samples were taken from the ends of the
tube sections by drilling out solidified fuel with
carbide-tipped bits. Clean samples were obtained
by collecting only ‘the borings from a small bit
after a large bit had been used to drill through
the surfoce material.
been circulated during operation of the loop were
obtained in the same manner from the portion of
the loop filling line that was outside the pump
shield. The samples were’ analyzed for uranium,
zwcomum, and the constituents of Inconel (nickel,
chromium, and iron). The results of the chemical
analyses are presented in Table 8.1. The nominadl
composition of the original fuel was 25 mole %

It can

ZrF, for cleaning and testing diluted the uranium
content of the operating charge of the loop. The
analytical results show that the uranium concen-
tration remained unchanged during the run, that
the chromium content increased from corrosion
of the Inconel tubing, and that the nickel content

‘probqbly ‘decreased during the course of the corro-

These changes in the concentra-

"vhon “of |ncone! components in the fuel and the

corrosive attack dlscussed below are consistent

"';wnih observatlons on umrradmted Ioops.3 4 The

 

3G M Adomson, R S Crouse, and P. G. Sml’th
‘Inconei-Fluoride 30-D Pump Loop

Fluoride Pump Loops 4930-A and 4935-1, ORNL CF-

55-4-181 (April 26, 1955).

Samples of fuel which had -

 

 

: in-pile

"Fuel from

Fu of-pile
* portion of loop

   
  

*The deviation listed is the maximum variation from the meqh of any but obviously contaminated samples.

 

. 159

 

w

 
 

 

 

 

 

'plcflng out of RU

 

"ANP PROJECT PROGRESS REPORT

103 Nb95

loop was described prev1ously
The tube sections cut from the loop for metal-

‘lographic examination were cleaned of fuel to
facilitate pohshmg and etching, They were placed
'_verhcc”y in an inert-atmosphere furnace and held

at ‘a femperature above 750°C until the fuel had
completely drained from the tubes. Control speci-

;_Vmens (pieces of as-received tubing cut from the
“ends of fhe fubes used to fabricate the loop) were

examined | for comparison with the samples from

the loop. The methods of examination and the
results have been reported by M. J. Feldman and

hns <:o-workers.‘S

In general, ’fhe chcnges in the Inconel s were those
expected in specimens subjected to 'rhe heat treat-
ment imposed by the operation of the loop. The
corrosion averaged 0.5 mil of penetrcmon. The
penétration and the points from which specimens
were taken are indicated in Fig. 8.2. No deposits
of mass-transferred mate_ri'qi were observed. Also
shown in Fig. 8.2 are the locations from which
the specimens shown in Figs. 8.3 through 8.8
were cut. Figure 8.3 presents a typical sample

 

M. T. 'Rrobins‘on, S. A, Reynolds, and H. W, Wright,
The Fate of Certain Fission Products in the ARE,
ORNL CF-55-2-36 (Feb. 7, 1955).

" SM. J. Feldman et al,, Metallographic Analyses of
Fuel Loop II, ORNL. CF-55-6-22 (June 21, 1955).

onto the walls of the

from the short out-of-pile section of the loop, the
only section unheated during actual irradiation.
This section was in the outlet leg of The !oop,
and it was maintained ot about the femperature
of the rest of the loop by the molten salt. Samples

from this section of the loop showed the minimum

corrosion attack, the average penetrahon being
less than 0.5 mil. A sample that is typical of the
entire unirradiated part of the loop is shown in
Fig. 8.4; the specimen was taken from the inlet

" leg of the loop within the LITR shleld The corro-

sion penetration averaged 0.5 mil, and the maximum
penetration was 1 mil.

The maximum corrosion in the loop was found
in the unirradiated portion of 'rhe ou’rle'r leg, Fig.
8.5, where the average penetration was ] m|1 cnd
the maximum was 2.5 mils. The control specimen
taken from the unused end of the tubing in the
outlet leg of the loop is shown in Fig. 8.6. Surface
cracks on the inner surface of the tubing were
observed to be common in the control specimens.
The cracks were probably respon5|b|e for occa-
sional voids extending to depths of 3 or 4 mils
in the loop specimen. A specimen from the tip
of the irradiated loop nosepiece, which was closest
to the LITR lattice, is shown in Fig. 8.7, The
corrosion was about the same as the average
corrosion found elsewhere, without the occasional
deep voids shown in Fig. 8.5, The thick-walled
tubing for the irradiated section was taken from

 

 

 

 

 

 

: S50-B8-1257
: . -QRNL-LR-DWG-8384
—-—-,%.
I WELDED JOINT BE.TWEEN LENGTHS OF TUBING
' NUMBERS REPRESENT PENETRATION IN mils
SPECIMEN 19 SPECIMEN 11  SPECIMEN 2—
CONTROL 1B CUT
FROM THIS END
0.5 av
1.0 av 0.5 av C.5av >0.5av
3.0 max 0.8 mox 0.8 max 0.5 max 0.8 max
i A) 7N
T Te— AL z_/
N — 1
S A 1
A L ri
0.5 av 0.5 av 0.5 av lflZ\ 0.5 av
1.0 maox 1.0 max 1.0 max 0.5av 1.0 max
1.0 max ) _
7 SPECIMEN 8 o

 

0 2 4 6

s Flg. 8.2 Locahons and Average and Mux|mum Depfhs of Corrosmn Aflack of LITR Honzontal Beam-Hole
S F!uorlde-Fu .IflLoop.

 

 

8 10 - . 14 ‘
DISTANCE FROM END OF LOOP (f) - ' '

 

 

 

 

"

 

 
 

 

L4

# 2y

» 4

 

PERIOD ENDING SEPTEMBER 10, 1955

a different lot than that used for the umrmdlcted However, fhe Iarger grain structure observed may

 

sechons. Anofher specimen from the lrradlated have been due to hlgher femperqfures res_ulhng
section, taken from a lower flux region in the * from better contact with the tubular heaters.
outlet leg, Fig. 8.8, exhibited the same degree =~ No increases in corrosion attack because of ir-

radlctlon and no other unusual effects were found

 

of corrosion afiack as thm‘ shown in Flg 8.7.

   
 
     

 

UNCLASSIFIED
_RMG 1154

UNCLASSIFIED
RMG 1112

Fig. 8.4, Sfiééimefi 8: Sécfi;n of Inconel ‘Tub-in'g. Taken ‘fr..Om Unirrudiofed Porfibn of Inlet Legofl.oop

Within the LITR Shield.

161

    

 

 
.ANP PR OJECT PROGRESS REPORT ‘ : !

   

 

 

 

3
4
N

 

   

 

 

 

Fig. 8.5. Specimen 11: Section of Inconel Tubing Taken from Unirradiated Portion of Ouflef Leg of Loop
Within fhe LITR Shield. 250X. Reduced 3.5%.

 

 

 

 

' } Flg. 8.6. Control Specimen: Section of Inconel Tubing Taken from Unused End of Tubmg Used for Out-
* “let Leg of Loop from Which Specimen 11 (Fig. 8.5) Was Taken. 250 X. Reduced 3.5%.

 

162

 

 

 
.z

 

" ¥

3 N

 

PERIOD ENDING SEPTEMBER 10, 1955

'UNCLASSIFIED E

 

NR;MG 1125;
!
i
i
Fig. 8.7. Specimen 17: Section of Inconel Tubing Taken from In-Pile End of L.oop Nosepiece. 250X. E

Reduced 4.5%.

UNCLASSIFIED
RMG 1134

 

 

Fig. 8.8. Specimen 19: Secfion of Inconel Tubing Taken from Outlet Leg of Loop Nosepiece. 250X,
Reduced 4.5%. | o

163

 
 

 

i

A

oo ol dihes 3

 

ANP PROJECT PROGRESS REPORT

in the loop. The results of this exper:ment are
thus in agreement with the findings in some of
the capsule irradiations.”’ The exper(menfal work
on thls loop has been comple'red and a topical
reporf is belng prepared :

  
 
 

MIVNIAT_U R{E lN PIL E LOOP
_”G w Kenlholtz |
o H E Robertson

 

C. C. Webs’rer
_ W. R. WI”IS
Solld Sfcfe D:vusuon
D E. Guss

o Unlted Sfcn‘es Alr Force

 

The mlmafure fluorlde-fuel loop, descrlbed pre-
v:ously, was operafed in vertical position C-48

~of the L!T‘R( Because of faulty behavior of the

pump~ motor, the fuel flow velocity could not be
maintained at @ steady value, and the test was
incomplete as a corrosion study. [t was possible,
however, to make a fairly thorough study of in-
pile characteristics of the loop, which operated
for a total of 30 hr at full power, with a fuel
Reynolds number of 3200 to 5000, The operation
throughout this period was complicated by inter-
mittent pump motor trouble, and the experiment
was terminated when the motor stopped completely.

The instruments controlled the temperature satis-
focforily to within £5°F ond did not allow ex-
cessive excursions in temperature when the fuel
velocity chqnged rapid[y Several scram situations
occurred during the short run because of low pump
speed, high activity in the off-gas stream, and
a high temperature indication that resulted from
a thermocouple failure. All these situations were
adequctely hondled in ’rlme to prevent high-tempera-
ture excursmns _

Threg c_omponents of the loop were faulty: the
pu{fiip""""""mb‘fpf, the pressure transmitters on the
venturi ’rube, and the header box in the off-gas
system.  No complete explanation for the erratic
behdvior of the motor can be given, as yet, but

it appecrs that ’rhe combination of high temperature
~and very dry helium atmosphere caused brush
’fcnlure. An mduchon motor with a canned rotor
“will be used on the next loop.

 

7W E Brownm Solid State Semiann., Prog. Rep.

: .'Aug. 30, 1954, OR [- 1762, p 39.

8w. R. Willis et al., ANP Quar. Prog. Rep. Man 10,

1955, ORNL-1864, p 147,

164

“The pressure tronsmtfiers were unre[:cble, but

it does not appear that any changes which could

be made in a reasonable time would be of much

‘help. Therefore they are to be removed from the

system. Removal of the pressure cells will permit

“-“more of the loop to be lowered into the reactor.
- This will increase the total power generation and

will place the pomt of maximum temperature near

' fhe maximum neutron flux.

During the short irradiation of the first loop a

- gasket was blown out of the off-gas header box

located at the face of the reactor. This allowed
some leakage of activity into the midriff area of
the LITR. The header box is to be redesigned to
withstand the fuli, avculcble, air pressure without
leaking.

Since the flowmeter was moperoble, another
method of obtaining fuel velocities was devised.
The method used was to find the pressure de-
veloped at a given pump speed and to find the
change in pressure vs flow for the loop. These
values were measured byusing water as the pumped
fluid in exact duplicates of the pump and the loop.
Figure 8.9 shows a plot of pump speed vs Reynolds
number of the fuel.

in Fig. 8.10 experimental and calculated curves
of temperature differential in the loop as a function

S
SSD-A-1253
ORNL-LR-DWG-8380

104

PUMP SPEED (rpm)

103

 

{0 REYNOLDS NUMBER 10

Fig. 8.9. Pump Speed vs Reynolds Number of
the Fuel in the Miniature In-Pile Loop.

 

. e T

 
"

e

 

W

“t

-

-

TEMPERATURE DIFFERENCE (°F)

_ of Reynolds number are compared.

  

S ated in the loop,' |
7 due tq_uhe loop was greater than estimated or
. if the spatial distribution of the flux was greatly:
“: altered, a lower‘ffemperofure differenhcfl than thut»___ '
.gfi'f'_-calculcfecl would have resul ted. “This poss:blhty"fl“'
 will be mveshgated when the values of the flux,

" measured by a monitor in the loop, are known.

T
$SD-A-1254
ORNL-LR-DWG-8381

CALCULATED CURVE

EXPERIMENTAL CURVE

 

REYNOLDS NUMBER
Fig. 8.10. Comparison of Calculated and Ex-
perimental Temperature Differentials vs Reynolds
Number in the Miniature In-Pile Loop.

mental temperature differentials shown in Fig. 8.10

- are about 40% below the calculated values. This
s cons:dered to ‘be a’ sohsfcc'rol’)' Check, e
the assumpflons made in performmg the deslgn"“" '_
qlcuiahons vtended fo predtcf tempera'rure differ.

  

"an d

 

 

The next loop will be inserted farther into the

PERIOD ENDING SEPTEMBER 10, 1955

active lattice to obtain a higher flux and a higher
temperature differential.

CREEP AND STRESS-CORROSION TESTS

J. C. Wilson

N. E. Hinkle
J. C. Zukas
Solid State Division

W. W. Davis

The Reactor Experiment Review Committee ap-
proved the insertion of the pressurized stress-
corrosion apparatus in HB-3 of the LITR. The
apparatus, which was described previously,? is
shown in Fig. 8.11. A thin-walled tubular speci-
men, surrounded by an annulus containing approxi-
mately 2.5 g of enriched fuel, is internally stressed
by helium gas through the pressure chamber while
being maintained at a temperature of 1500°F. The
heat generated by the fissioning fuel is conducted
through fins to a water cooling coil. A resistance

 

The .ékpéri-'

   
   

, thus, if the flux depressaon'”x'k

9W. E. Davis et al,, ANP Quar. Prog. Rep. June 10,
1955, ORNL-1896, p 170.

UNCLASSIFIED
PHROTG 14344

 
   
 
 
 
 
 
 
 
 
 
 
   

PRESSURE CHAMBER

PRESSURE TUBE

L o Ll

  

{ for fuel annulus )}

1 SPECIMEN TUBE

- FUEL TUBE

 

COPPER SLEEVE
(solders to water
cooling coil }

 

   

COOLING FINS

e

- i FueL TuBE PLUG

Fig. 8.11, Stress-Corrosion Apparatus for Use
in the LITR. '

165

 

 

T N T P " T e o r

 

 

 
 

ANP PROJECT PROGRESS REPORT

heater spaced between the fins is provided for
additional control of the specimen temperature.
The control thermocouple on the inside of the
specimen tube should give very nearly the maxi-
mum temperature of the system at any time. Heat
transfer calculations predict that the fuel will
freeze at the outer walls of the fuel chamber so
that the fuel will be contained in a cup of solid
fuel; thus the surface-to-volume ratio of the test
system will be considerably lowered. |t does
not appear likely, at this time, that the tempera-
ture of the specimen can be maintained at the
control level during reactor shutdown perlods,
however, fuels with a power density of 1000 w/cm?®
are dlfficuh‘ to simulate in bench tests of this
type of expenmenf. The gas space above the
fuel is connected through a capillary tube to an
expansion chamber to decrease the back pressure
on the specimen walls when the specimen is
brought up to temperature. The first group of
specimen assemblies is now being filled with
fuel for tests in the LITR. Engineering of the

 plug design to adapt the apparatus for insertion

in the MTR is under way.

A simplified version of the apparatus to permit
creep data to be obtained in any combination of
two nonfuel atmospheres is undergoing bench
tests prior to irradiation. Measurements of the
total strain in the specimen tubes will be ac-
complished with a pneumatic measuring gage.
Inside-diameter measurements of the specimen tube
can be made with an accuracy of about 0.00005 in.
with the pneumatic gage, and the manipulative
operations are simple enough to be done in a hot
cell.

The operahon of a number of pressure-reducing
valves, with pressure ranges from 200 to 1000 psi,
has been checked over a period of about a week.
Of the two valves presently being tested, one
appears to be capuble of holdmg set pressures
to within 1% over the range of room temperatures

‘encountered in the Iaboratory

A 1:1 rcmo pressure-volume transformer for iso-
lahng the gas inside the spec:men tube from the
pressure supply has been built and tested. Be
cause of the controlled limited volume of gas
avmlable to the specimen tube with this apparatus,
leukage or rup?ure results in only a very small
pressure rise in the whole apparatus. Protection
against pressure rises in the water-cooled irradi-

166

ation &R Will make possible a large reduction
in the weight of the irradiation can.

The MTR (tensile) creep test apparatus was
irradiated for two cycles in hole HB-3 of the MTR
at 1500°F and 1500 psi. For some unknown reason,
two of the four Bourdon tube extensometers did
not operate after insertion in the reactor. There-
fore total creep measurements and a determination
of the effects of radiation on the calibration fac-
tors of the Bourdon tube extensometers will be
made when the apparatus is returned to Oak Ridge.
The apparatus has been cut from the irradiation
plug at the MTR site. The bench test equivalent
of the inspile rig in time, temperature, and siress
has been assembled and testing has started.

FLUX MEASUREMENTS IN THE MTR

J. B, Trice H. V. Klaus
Solid State Division

F. J. Muckenthaler
Applied Nuclear Physics Division

T.L. Trent

Engineering and Mechanical Division

P. M. Uthe, United States Air Force
J. F. Krause, Pratt & Whitney Aircraft
R. H. Lewis, Phillips Petroleum Company
F. W. Smith
Consolidated Vultee Aircraft Corporation

Neutron flux measurements have been in prog-
ress for some time in hole HB.3 of the MTR.10 A

- schematic diagram of the location of the HB-3

beam hole in relation to the lattice is shown in
Fig, 8.12. High-energy flux data were obtained
because of a need to know the fast-neutron flux
intensity and distribution available for tests of
structural properties of ART materials, For such
tests to be realistic, both a high-energy flux and a
high, integrated fast-neutron dosage were required.
Thermal-neutron flux data were also obtained for
estimating power generation in uranium-bearing
fluoride fuel loops.

The results obtained include a crude spectral
analysis of the neutron energies in HB-3, made

 

105, B, Trice and P. M. Uthe, Solid State Semiann.
Prog. Rep, Aug, 31, 1953, VORNL-160_6, p 23.

 

 

 

 

e

 
D)

S sk ki

 

.

N

w

HB-3 beam hole.

‘:"??‘”fure bcsed on the

 

UNCLASSIFIED
ORNL-LR~DWG 8B36A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

B SHIM RODS
3 6

o

Y re
12

~INCHES

Fig. 8.12, ‘Sch'érr;‘ufic-Difig.rum c;f MTR Lattice
Showing Position of HB«3 Beam Hole,

with threshold detectors and resonance detecs

tors,'! thermal-neutron unperturbed flux traverses,
and traverses made with mockup materials which
simulated, more or less, the true neutron-absorbing

- characteristics of in-pile fuel loops, 7
ngh-energy flux d:sfnbuflons are shown m Flg.
8,13, in integral form, for three positions in the
The experimental points shown

_ ‘Iear__reqchons listed in_ 'chle'fi
8 2 The‘dofled lines are fissu_on spectrvdk ndrmal-m

   
 
  
  

 

fi'i___\ufrons decreases with mc‘reasmg\ dfstance oway“:”_
cafes 'rh“t the hlghesf -

 
 
 
   
 

 

 

”For method, see S VM. Dancoff et al., Activation
Cross Sections by Boron Absorption, CP-3781 (May 6,
1947), p 17.

 

 

  

PERIOD ENDING SEPTEMBER 10, 1955

UNCL ASSIFIED
ORNL-LR-DWG 8644A

— —FISSION SPECTRUM
® £ ' SPECTRUM TO 2.4 Mev

O Rf11 POSITION, 2.8in. FROM BEAM HOLE END
4 R1{4 POSITION, 8.8 in. FROM BEAM HOLE END
D R32 POSITION, 20.8in. FROM BEAM HOLE END

~

HIGH- ENERGY NEUTRON FLUX ABOVE THRESHOLD (neutrons /cm?2- sec)

 

0 2 4 6 8 O
THRESHOLD ENERGY (Mev)

Flg. 8 '|3 ngh-Energy Neufron ‘Flux Distribu-
‘tions in Three Posnhons in Hole HB-3 of the
MTR.

“of the nurnber of neufrons in the specirum above

0.1 Mev, based on measurements of the number of

"“neutrons in the spectrum ‘above 2.4 Mev, as meas-
““ured with the reaction P31(z, p)Sn“ The ratio of

the total number of fast neutrons in the spectrum
to the fhermal-neutron flux, as shown in Flg. 8.14

':':'":m‘ ‘the same posmon, is 0.135. At position HB-2
he  in the LITR, the rcmo]2 is 0.80. This means
‘that 67 days ‘at a flux of 3,175 103 fast neu-
'trons/cm 'sec are required for the estimated total

 

 

12y B. Trice et al,, Solid State Quar. Prog., Rep.
Aug, 10, 1952, ORNL-1359, p 12,

167

 
 

 

 

 

* ANP PROJECT PROGRESS REPORT

fast neutron exposure of 1.8 x 102° 1o be reached,
~which is the estimated dosage for certain struc-
tural members in the ART for an operating period

UNCLASSIFIED
ORNL-LR-DWG 8825A

FLUX MEASURED WITH BARE COBALT

-sec)”
o

" THERMAL-NEUTRON FLUX {neutrons/cm?

 

10
3 7 M 45 9 23 27 31 35
DISTANGE FROM BEAM-HOLE END (in)
Fig. 8.14. Thermal-Neutron Fiux in Hole HB-3
of the MTR.

of 500 hr.13 Since the flux g"radiént is very steep,

as shown in Fig. 8.15, such a test would have to
be made in a limited regioh in the beam hole. The

" magnitude of the maximum instantaneous fasteneu-

tron flux available, 3.1 x 1013, is less than the
flux expected in the ART, 4 x 10'4, by about a
factor of 10 and misses some aircraft reactor de-
sign fluxes, for example, ~2 x 10'5, by about a
factor of 70. Further analyses are being made in
order to substantiate or refute these fast-neutron
flux measurements,

A series of measurements of flux depression in a
mockup of a fluoride~fuel in-pile loop section as a
function of amount of fuel present, wall thickness
of the container, and position of the container were
made in order to serve as a basis for design of a
loop for operation in HB-3 of the MTR. The values
of effective thermal-neutron flux obtained from this
experiment were used to calculate the total power
generation and power density to be expected in
the loop.

The apparatus for the series of measurements
consisted of two major pieces. The outer piece
was an Inconel cylinder 7 in. in length and hol-
lowed to contain a solid-cylinder core of the neu-
tron absorber used to simulate the fuel. The core

 

13y, K. Ergen, private -communication, Aug. 19, 1955.

TABLE 8.2, CHARACTERISTICS OF DETECTORS USED FOR FLUX MEASUREMENTS |
IN HOLE HB-3 OF THE MTR

 

Threshold or

Cross Section of

 

Reaction Resonance Energy Resonance Integral Half Life
Co%%(n,y)Cot? 120 ev 34b 5.2 years
Na23(n,y)Na24 1710 ev 0.27 b* 149w
C137(n,3)C138 1800 ev 0.348 b* 38 min
V5 n,) V52 3000 ev 2.0 b* 374min
AI27(n,)A128 9100 ev 0.14 bxrxx C23min
P31 (n,p)si3" 2.4 Mev 75.0 mb*+ 26hr
A2 (5, p M2 4.6 Mev 39,5 mb** 98 min

28(n pIAIZ8 5.5 Mev 79.8 mb** 2.3 min
' Ma24(n,0)Na?4 6.3 Mev 47.6 mb** 149 b
A127 ()N o24 8.1 Mev 111 mb#* 149 he

 

- '*S ‘M ‘Dancoff et al., Activation Cross Sections by Boron A bsorption, CP-3781 (May 6, 1947).
**R. F. Taschek Radzoactwe Tbresbold Detectors for Neutrons, LADC-135 (also, MDDC-360,

17 1946)

1 68 |

 

 

Declassified Sept.

 
 

"%

L

.w

v?‘\‘ :

 

",

-

UNCLASSIFIEb
ORNL-~LR—DWG 8645A

POSITION IN PNEUMATIC RABBIT
112 13 14 21 22 23 24 31 32 33 34 41 42 43 44

2
101‘!
5
2 RUN NUMBER
DIA
10'° D1B

DiC

DiD
SPECIAL RUN

o
0

a

n

o

 

o

4 6 8 10 12 14 16 18
DISTANCE FROM BEAM-HOLE END (in)

HIGH-ENERGY NEUTRON FLUX ABOVE EFFECTIVE THRESHOLD (neutrons/cm®-sec)
o

~ Fig. 8.‘157.7‘ _rFusf.-Né't-mon'Fl'uxr""l'ruverse of Hole
HB-3 as Measured with the Threshold Reaction
A127(n,c)Na?4 (8.1 Mev and Above).

was assembled from two half cylinders slotted at
intervals along the leng?h in such a way as to

'allow cobult f0||5 to be posmoned along the axis

7 ate  conti _iwuth ?he two ‘halves of the
""cylmdér. Three ‘dlffefen’r core composmons “were o

- namely, chmlum, boron, qnd”z

  
 

in inti

“Used. in k’rhe
“Tithiom,

»s_ts‘

   

 

    
 

Measuremenfs made with each of the three cores
gave the same effective neutron flux within ex-
perimental error limits. Flux traverses through the

 

 
 

The cadmlum was\qlfoyed ‘with ‘mcgneslum.-;ondr
S f_OM Yein. rods; the boron wds mixed, as'-, _
B C wn‘H a!ummum _by a pressmg and hIgh f..—,ngh-

  

PERIOD ENDING SEPTEMBER 10, 1955

three cores with the reactor operating at 5 Mw are
shown in Fig. 8.16.

An operating power of 5 Mw, rather than the usual
30 Mw, was chosen because the gammaeray heat
at 30 Mw would have raised the temperatures of
the cores to above their melting points. The ther-
mal-neutron flux as a function of power was ob-
tained at two positions, as shown in Fig, 8.17, so
that the flux-depression data obtained in the S<Mw
runs could be scaled to 30 Mw. As may be seen,
the flux vs power curves are nonlinear between 20
and 30 Mw,

A calculated estimate of the power to be ex-
pected in the first MTR in-pile loop,'4 based on
these data and on the results of irradiation of a
simulated loop, 13 was 31 to 36 kw for the fluoride
fuel mixture NGF-ZrF4-UF4 (53.5-40-6.5 mole %).

ANALYSIS OF REACTOR-GRADE BERYLLIUM

J. B, Trice
Solid State Division

R. H. Lewis T. Passel
Phillips Petroleum Company

Two reactor-grade samples of beryllium metal
chips obtained from The Brush Beryllium Co. were
irradiated and then analyzed to secure information
needed in designing procedures for handling and
processing the large amounts of beryllium which
will have been irradiated for hundreds of hours in
aircraft-type reactors, Chemical analyses with a
limit of detection of 200 ppm showed the following
amounts of impurities:

Impurity Amount (wt %)
Fe and Mg 0.03
Si and Al 0.02
Ni and Cr 0.006

. Mn 7 0,003

[ 1 0.0002
Ca - 0.0001
Se ' None

- Analyses with a sod|um—|od|de scmhllohon spec-

'"frometer revecled two gammc-roy peaks, one at

0.90 Mev and one at 1,12 Mev, as shown in Fig.

8 ]8 Flgure '8.19 shows 1rhe gamma-ray spectrum
”ifor machated Sc:“6 As ‘may be seen, the peaks

 

MJ B Trlce, MTR Loo Deszgn Memo SSD-] ORNL
CF-55-5-138 (May 18, 1955).

15p, M. Haines, ANP Quar. Prog. Rep. June 10,
1955, ORNL-1896, p 26.

169

TR T R T T FRETTYT T R

BT B R Y DT

 
 

 

 

ANP PROJECT PROGRESS REPORT

FLUX AT SURFACE OF LOOP

14 UNPERTURBED FLUX

83

FLUX IN MOCK FU
13

8.8in.
TO FUEL

THERMAL-NEUTRON FLUX (neutrons /em? - sec)
o rn

7.8 in. -
2 HOLE END

0 5 10

 

-
ORNL-LR—DWG 6016A

°
o CADMIUM ALLOY CORE
A LITHIUM - BEARING FUSED SALT CCRE

15 20 25 30

DISTANCE FROM BLIND END OF PNEUMATIC TUBE (in.)

Fig. 8.16, Thermal-Neutron Flux Traverses Through Cadmium-, Boron-, and Lithium-Containing Maierialg

in the MTR.

UNCLASSIFIED
ORNL-LR-DWG 50104

 

30(x 10'®) I I
LOOP DESIGN VALUE\

25 1
© R POSITION
® Ri4 POSITION o
1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o
@
@
o~
£
o
~
w
g 20
3
c
§ 15
-
L
g /
= . » e
<t
=
= / LOCOP DESIGN VALUE
¥ -~
= 0
0 5 10 15 20 25 30 35

REACTOR POWER (Mw}

, Fig. 8.17. Thermal-Neutron Flux as a Function
_of Reactor Power at Two Positions in Hole HB-3
of the MTR,

170

are at the same energy as those in Fig. 8.18, Ex-
perimental data for the beryllium chips indicate a
half life of 90 days, which agrees satisfactorily
with a listed'% value of 85 days. Both the gamma-
ray peaks and the energy sums indicate that the
gamma rays are in cascade, as they are shown in
the listed decay scheme.

Another beryllium sample, from the R, D, MacKay
Company, was analyzed in the same manner for
scandium and was found to have about the same
amount present {Fig. 8.20). The beryllium«to-
scandium weight ratio for The Brush Beryllium Co.
sample, as measured by the gamma-ray scintillation
specirometer and the MTR 47 high-pressure ioniza-
tion chamber, was 6.8 x 104, For the R, D, Mac-
Kay Co. sample, it was 7.1 x 104, Therefore, the
scandium present in both samples was below the
limit of chemical detection stated above. |

 

YNuclear Data (compiled by K. Way et al), NBS-49
(Sept. 1, 1950), p 41,

»

 

 

 

 

 
PERIOD ENDING SEPTEMBER 10, 1955

-y

. .. UNCLASSHIED o , . UNCLASSIFIED
" ORNL-LR-DWG 7921A ORNL-LR-DWG 79224

  
    

 

3
3x10

 

40 ~
~ £
E &
: ;
5> 5 =
= @
w
& &
= z
1z = ;
@ &
<I
s s
. s 2 2
£ g 10 3
UL Ll
L >
= =
g ° <
- o &
2
L 10 . . 0 02 04 06 08 1.0 £.2
1 O 04 02 03 04 05 06 07 08 09 10 41 42 13 ENERGY {Mev)
* | ' ' ENERGY (Mev} ' '
" Fig. 8.18. Gamma-Ray Spectrum from Irradiated Fig. 8.19. Gamma-Ray Spectrum from Irradiated
i Beryllium. Beryllium obtained from The Brush S
& .

Beryllium Co.

UNCLASSIFIED
4 ORNL—LR—DWG 7923A
4X10

 

“05 06 07 08 09 10 11 i2
ENERGY (Mevy T o

-

be - 7- - i e Fig. 8..20. Garfima-.Rdy Spectrum from lrradiated
| Beryllium, Beryllium obtained from R. D. MacKay
Company.

171

 

   

 

T R T

 
Gt s

 

 

 

ANP PROJECT PROGRESS REPORT

9 ANALYT!CAL CHEMISTRY OF REACTOR MATERIALS

C D Susano

J. C. White

Analyhcal Chem:éh‘y Division

The n- butyl bromlde mefhod for the determmchon
of ‘oxygen in sod;um was modified to ensure the
~elimination of possuble sources of contamination

7 _""‘3dur|ng analysis. The vacuum-distillation method

oo for this de'rermmahon was investigated. Work was
‘ompleted on the volumetric determination of zir-

 

  

- conium in fluoride ‘salts. A modlftcchon of the
4 "“'rappara’rus for’ the ‘determination of uranium metal
'.;v'_;:n fluorlde sahs was mcorporcn‘ed to permn‘ a more
V;l’dpld anclys:s Analyhcal assistance was ren-
. .f_'-"r.‘i:;dered in the s'rudy ‘of using argon to eliminate air
_i;,_from a dry box. !nvestlgahon of the application
. Clof the brommatlon ‘method for the determination of

fjif__:oxygen in zirconium fluoride and its mixtures with

‘>-~ofher fluorlde solts was continued.

DETERMlNATION OF OXYGEN IN SODIUM

‘A. S. Meyer, Jr. W. J. Ross
G. Goldberg
Analytical Chemistry Division

n-Butyl Bromide Method

The n-butyl bromide method' is the standard
method used in the ANP Analytical Chemistry
Laboratory for determining oxygen in sodium. The
original method has, however, been modified re-
cently by two laboratories.2® In the modified
procedures the organic reagents, n-butyl bromide
and hexane, are purified and dried by passing them
through a column packed with silica gel and dia-
tomaceous earth. The purified reagents are stored
over P,O.. A modification of the reaction tube
has also been recommended to reduce atmospheric
contamination. The modified method is reported
to be very sensitive and to be applicable to the
defermmohon of oxygen in sodium in concen-
trations as low as 5 to 10 ppm. An evaluation of

- these modifications was therefore undertaken pre-

paratory to incorporating them in the standard
procedure,
‘The reagents, hexane and n-butyl bromide, were

purified and dried by passing the liquids through

 

L '/_:'l'Cohf

 

1, c. White, W. J. Ross, and R. Rowan, Jr., Anal.

V_K_‘Cbem. 26, 210 (1954).
2thyl Corporation, Baton Rouge, La.

L.. Silverman, North American Aviation, Los Angeles,

 

columns packed with silica gel and dlq’romaceous
earth; they retained about 10 ppm water. They
were then rendered anhydrous by desiccation with
activated Al,O, or ons' The latter has the dis-
advantage that it reacts with the water in the
reagents to form acids which remain in the liquid
phase in sufficient concentration to cause low
results in the oxygen determination. Such an

error can become significant when the reagents

are stored over P, O, for extended periods. Also,
anhydrous reagents become contaminated with
water when exposed to the atmosphere. Wet re-

agents lead to high results in the oxygen determi-

nation.

The n-butyl bromide procedure was further modi- |

fied so that an atmosphere of argon was maintained
over the reagents and reaction mixtures at all

times. In addition, transfer of the reagents from

the storage vessel to the reactor tube was carried
out by applying pressure with dry argon. The
modifications are illustrated in Fig. 9.1. Excel-
lent reproducibility was obtained in the determi-
nation of oxygen in sodium with the modified
apparatus when the samples were taken in glass
tubes. The oxygen content of the majority of the
samples received from ANP facilities was in
excess of 200 ppm. Much lower concentrations,
of the order of 20 to 40 ppm of oxygen, were found

UNCLASSIFIED
ORNL-LR-DWG 8949

 
 

SERUM BOTTLE

   
   

A
RGON N2> % JOINTS

TO
7-BUTYL BROMIDE VACUUM
+ HEXANE

AGITATOR
ALUMINUM OXIDE

 

 

Fig. 9.1. Apparatus for Transferring Reagents
in the Determination of Oxygen in Sodium by the
n-Butyl Bromide Method.

 

-

 

 

 
. [

 

"' “of the scmple cup i

in sodium samples which were subjected to la-
borious purification.

The n-butyl bromide method was also used to

- study the effect of precleaning glass sample tubes
prior to sampling.

The interior of such sample
tubes was found to be effectively cleaned of

' oxygen com‘alnmg lmpurmes by treating with

- H,GO, or by flushing with molten sodium. A
. multibulb tube is required to accomplish the
. cleaning action with sodium. -

Vocuum-Dlshlluhon Method

A method is being ‘tested for the determmctlon

' of_ oxygen in sodium by titration of the Na,O that
‘remains after vacuum distillation of the sodium
.. metal. A distillation apparatus has been con-

structed that is srmllor to one developed by the

Argonne National Loborotory ~The vacuum-

~distillation method of analysis is particularly
‘suited to the sampling of sodium at operating

temperatures of the order of 1200°F and obviates
the necessity for precooling the sodium before
sampling. A schematic diagram of the equipment
is shown in Fig. 9.2.

The sample of molten sodium is introduced into
the evacuated apparatus through a heated transfer
line, which is maintained at the operating tempera-

ture of the system being tested. The metal flows

directly into a calibrated, hemispherical sample

‘cup, which is fitted with a thin, nickel liner. A

volume of sodium sufficient to flush the transfer

‘line and sample cup is first passed through the

apparatus. The excess sodium, which pours over
the edge of the somple cup, is discharged into

_a glass reservoir in which the volume of the flush
~ ‘sodium can e_observed With the pressure “of the“"
- system mmntolned at Tess than 10 g, ‘the somple
~is heated fo 950°F ‘and the me1oH|c sod|um 15%‘:“ .
‘__:"dlstllled to the cooled ‘walls of the apparo'rus
" When the dlsflllu‘l‘lba is complefe ‘the nickel Imer‘"’
: emoved, and the residual
- N020 is dissolved and then tlfrofed\wnth a dilute

 

 

 

 
 

°" 9* dized acid

   

ured When

 

4J R. Humphreys, personal communication, June 24,

1955.

 

    

  
     
 
  
  
    

i'fl"\e sfoble BeF

  

the prehmmory rons

PERIOD ENDING SEPTEMBER 10, 1955

are completed, the apparatus will be attached
directly to a forced-circulation high-temperature-
differential sodium loop, and analyses will be

‘carried out during operation of the loop (see

Sec. 2,

“Experimental Reactor Engineering’’).

VOLUMETRIC DETERMINATION OF ZIRCONIUM
IN FLUORIDE SALTS WITH DISODIUM
DIHYDROGEN ETHYLENE-
DIAMINETETRAACETATE (EDTA)

A. S, Meyer, Jr. D. L. Manning
Analytical Chemistry Division

A volumetric method for the determination of
zirconium in mixtures of fluoride salts was de-
veloped. This method, which was outlined in the
previous repon‘,5 was based on the work of Fritz
and Johnson. ®

The procedure consists in adding an excess of
EDTA to a sulfate solution containing zirconium
and titrating the excess with trivalent iron to a
Tiron  (disodium-1,2-dihydroxybenzene-3,5-disul-
fonate) end point at a pH of 4.8. The color change

at the end point is from yellow to purple. The-

method is applicable to the determination of zir-
conium in the presence of sulfate, tartrate, smatl
amounts of fluoride, and hexavalent uranium. The
yellow color of hexavalent uranium seriously
interferes with the bismuth-thiourea end point
proposed by Fritz and Johnson.®

There are relatively few cationic interferences,
because zirconium and iron form extremely stable
complexes with EDTA. In the back-titration with
trivalent iron the purple color of the iron-Tiron
complex is not obtained until the metal-EDTA
complexes, which are less stable than the iron
- (IIH=EDTA complexes have been dissociated,

.....Fluoride ion concentrations of up to about 0.1 M
. do not appear to interfere with the titrimetric pro-
.- cedure, Higher concentrations of fluoride ion can

be effectively reduced by adding beryllium to the
solution prior to the addition of the EDTA, The
interference is removed through the formation of
- complex

" With a slight modification in the experimental

o 'rocedure, the method is opphcoble to the determi-
nohon of znrcomum m the presence of moderofe

 

. SA 5 Meyer, Jr, cmd D L Monmng, ANP Quar
Prog ‘Rep. June 10, 1955, ORNL-1896, p 177.

6J S. Fritz and M, Johnson, Volumetric Determination
of Zirconium, an EDTA Method Regquiring a Back-
Titration with Bismuth, 1SC-571 (Feb. 1, 1955).

173

P

 

 

 
 

 

 

 

 

 

ANP PROJECT PROGRESS REPORT

. vacuum ——
-‘———————

1000 °F\

C " TUNCLASSIFIED
~..ORNL-LR-DWG 8950

/CALROD HEATER

 

 

12006F\

 

 

 

 

 

 

 

 

 

 

 

 

N

 

 

 

 

Z
SODIUM Z&

TRAP GLASS - METAL
VACUUM FITTING

 

 

CALROD

 

 

 

 

 

 

 

 

TEFLON GASKET

RN Y

GLASS
OVERFLOW RESERVOIR—7 |

7\ NS
L

 

 

 

 

 

     
 

HEATER =25

SODIUM
VALVE
:‘:5:=:1:'-\:-:-:-:-:;:;;ié nnnnann i j-e———— CALROD -HEA'TER
4 SAMPLE CUP WITH
NICKEL LINER
4 A

N f 1 D/A|R~COOLED CoIL

L»// O

 

 

 

 

 

 

 

 

 

| —s—— O -RING FLANGE

 

 

GLASS PIPE
FLANGE

 

 

 

 

 

 

 

 

 

INSERT HEATER {000 °F

\THERMOCOUPLE

 

 

Fig. 9.2. Sodium Distillation Appqratus;

amounts of trivalent iron, divalent nickel, and
trivalent chromium. These metals form stable

 complexes with EDTA, ond therefore they would

ordinarily interfere seriously with the zirconium
determination. The interference can be overcome
by forming the EDTA complex of the zirconium

““and the interfering metals. Fluoride ion is then

SV

added to the solution to selectively dissociate
the zirconium-EDTA complex and form the stable
ZrF,~" ion, thus liberating the EDTA. The
liberated EDTA, which is a measure of the zir-
conium, is titrated with the trivalent iron solution

to a Tiron end point. The stability of the iron_
(I)~EDTA complex and that of the iron-Tiron

 

R R T

 

 

 

 
"

]

oy

o

 

complex are not adversely affected by fluoride
7 ion. By utilizing Mthis'technicjue,' the procedure
¥ ~° can be made Specn‘lc for zirconium in the presence
" of any metal ion that forms a complex with EDTA
and does not dlssoc:cfe after fluorlde lon is added
+ fo the solutlon._ )
- The proposed volumetrlc mefhod has been tested
- in fhe Iaborcfory and found to be’ suhsfcctory The
- procedure resulfs in a considerable saving of time

~in comparison with ’rhe rather lengthy grav:metrlc |

mandelic acid method. |t also appears to be more
ecsnly adap’rable to the routine defermmqhon of

zirconium ’fhan htrlmetrlc methods 1n whlch the‘

end point is detected spectrophotomemcally or
cmperome?rlcaliy The coefflc:enr of variation is

L ”of the order of 1% undep_' ldeal c«:ndmcns in ’rhe
R range 15 to 45 mg of Z|rco'n'um ‘

     

DETERMINATION OF U ANIUMEMETAL-
IN FLUORIDE SALT MlXTURES

A 5 Meyer Jr B. L McDoweH
- Anolyhcul Chemistry anls:on

, The apparatus for ‘the determmuflon ‘of uranium
) me’ra! in mixtures of fluorlde salts by decompo-
© sition of the hydride in an otmosphere of oxygen
“at reduced pressures7 was modified by replacing

copper oxide tube. The furnace used to heat the
. sample was replaced wn‘h a micro- preheater for
. each combustion tube. The heaters can be re-

’;;'j"moved from ‘the wcmlty “of the sample boct to
allow more rap:d coohng of fhe hydride.

   

' modnf;ca’tlon

 

A‘nalytlccl Chemlsfry ;DIVISIOI‘I -

A sfudy of the rate of elimination of u’rmospher:c
" gases from a dry box hcs been carrled out in

  

These |

PERIOD ENDING SEPTEMBER 10, 1955

conjunction with the Heat Transfer and Physical
Properties Section of the Reactor Experimental
Engineering Division, The concentration of
oxygen in the atmosphere of a 21-ft° rectangular
dry box was measured as a function of time while
argon was being passed through the box at a
constant flow rate. Oxygen concentrations in
excess of 1% were measured volumetrically by
absorption of oxygen in pyrogallol. Lower concen-
trations were determined by a modification of the
Winkler technique in which a suspension of
Mn(OH) in an alkaline solution of Kl is equi-
IIb!‘leECl with a measured volume of gas. The
iodine, whlch is equwaienf to the oxygen in the
sample, is liberated upon acidification of the
solution and is determined colorimetrically or by
titration th N025203 o

~When the gases in the dry box are thoroughly
agitated by means of a blower, the oxygen concen-
tration decreases exponenhcily with the volume
of sweep gos in accordance with theoretical
prediction. The oxygen can be eliminated more
rapidly by introducing the more dense argon at
the bottom of a quiescent dry box. Helium, when
introduced at the top of the box, provides a some-
what less effechve flushmg action thcn does

. the smgle combustion ’rube with two combushon o °r9°"

'fubes, either of whlch can be connected to 'rhell

The most effluenf flushmg was effecfed by
injecting argon at the botfom of the dry box without
supplementary agitation. With the highest argon
flow rate available, 25 cfh, the concentration of
oxygen “in the atmosphere of the dry box was
reduced by a factor of 100 by flushing with two
More fhon flve volumes of

volumes of orgon.
" re ’"'Simllar re-

   

 

 

 

Prog. Rep ]une IO 1955, ORNL- ]896 p 174

M. Codell and G. Norwitz, Anal. Chem. 27,
(1955).

1083

175

 

 

 

 

 
BT

 

Lo sbo sl o
Bl .
v S

"ANP PROJECT PROGRESS REPORT

- This 'mefhod mvolves the reaction of a metal
w ‘oxide with bromme vapor, in the presence of
_"_'_’_f‘igraph:te, to form the metal bromide and CO. The
~CO is then oxidized to CO, by reaction with hot
 f Cu0. The amount of CO evolved is a measure
o of fhe oxygen ortgmol[y present in the sample,
,_fThe opparatus for this analysis was described
e prev:ously
The prlmary objective of recent work has been
S fo find the ophmum conditions for the determi-
e nahon of oxygen in ZrF,. The oxygen in this
S A';Z':}salt is present as ZrO or ZrOF2 or, possibly,
el bo’rh compounds
S An ofiempi was made to analyze synthetic
_---'f-mlxtures of Zr02 and ZrF by the bromination
‘method. The recovery of oxygen as CO,, was
f:ffmcomplete in all cases. Analyses of samples of
_‘-'r-_»pure Z:0, by the bromination method also resulted
©in lncomplete recovery of the theorehcal amount
T 'f-j{;of oxygen.
.7 Several experimental conditions were varied in
| lfi',,,"':_:these mveshgotlons The length of the platinum
. “.beats was changed from ] to 3 in., and in some
- experiments the sample wcs contomed on a sheet
- of platinum. The weight of ZrO, was varied from
3 to 50 mg, and the ratio of ZrO, to graphite was

< varied from 1:2 to 1:20. In one experiment, sugar

-'_‘.:_"'-fcharcoal was substituted for the graphite.
In these studles with ZrO2
’ tcuned und the res;duol Zr02, which remalned in
"Z,',the reacflon boat, were determined. The reaction
involved in fhe initial step of the analysis is

ZrO + 28!2 + 2C-——>ZrBr + 2CO

:L';Therefore any ZrO which did not react with fhe-
bromine should be found in the boat. The amounts

   
  
  
 

of CO, obtained from these samples appeared to
_'ceed the amount of oxygen actually removed
from fhe ZrO This has been attributed, in part,
o the presence of moisture in the graphite.
-‘.lz,f‘Furthermore ‘the residual ZrO, found in the boat
"varied | from 50 to 90% of the original sample
("‘:ffwelght for’ perlods of bromination of up to 6 hr
at 950°C. -

Since unsohsfactory results were obtained with
sdmples of ZrO the method and apparatus were
,;,_checked wnh samples of TiO,. Only the amount

 

  
  
  

 

9J C. Wh:fe, J. P. Young, and G. Goldberg, ANP
Quar Prog Rep Mar 10, 1955, ORNL-1864, p 161.

S 10J P. Young and G, Goldber% ANP Quar, Prog.
" ‘Rep ]une 10, 1955. VORNL']396: p 178.

 

 

both the CO ob-

| of TiO, remaining in the plafindm-(beefz'c'ffer;d:': o
| 2-hr penod of bromination was determined in these

studies. Cqmplete removal of the T|02, as the
bromide, from the boat was achieved when the
oxide and graphite, approximately 15 mg each,
were mixed together with a mortar and pestle.

When the two materials were mixed by using a

spatula, quantative removal of TiO, from the
boat was not obtained; however, when fhe mixing
was done with great care, up to 95% of a 10-mg
sample of TiO, could be removed as TiBr4' in 2 hr,

The sample preparation described by Codell
and Norwitz® consists in placing titanium chips,
metal or alloy, between two layers of graphite.
In this case, it is probable that the volatilization
of the metal as a bromide would leave the re-
maining TiO, in intimate contact with the graphite,
without the need of external mixing. It would also
be probable that, in the determination of the oxide
contamination of salts that are volatile at moderate
temperature, there would be no need for tedious
sample preparation. For a determination of oxygen
in a material containing greater than 1% oxide,
it would be advisable to ensure an intimate mixture
of the sample and graphite. The information
gained in the study of the bromination of TiO,
will be applied to the determination of oxygen in
Zr0, and other oxides of interest.

The formation of a precipitate on the surface
of the BG(OH)2 bubbler, which was reported pre-
vnously, was prevented by inserting a thin
platinum tube inside the ignition tube. It was
necessary to use this platinum tube only during
the analysis of samples containing fluorides. |t

"is believed that the contact of volatile fluoride

salts with the hot quartz of the ignition tube re-
sulted in the formation of SiF,. This gas reacted
with the solution of Ba(OH), to precipitate
BaSiF . The platinum liner inserted in the ignition
tube prevented the formation of SiF .

Dei'ermi natieh ef CO
by Means of a Solution of PdCI,

Various methods for determining CO are being

investigated. The purpose of these studies is to
find a sensitive method for a more direct measure
of the CO that is formed in the determination of
oxygen in metal oxides by bromination.

A method for the determination of CO in blood

 

 

 

 

 

 

w

T R R AT T0 R T Y i

 

 
 

 

 

W oay,

  
   
   

is of current Im‘erest This method involves

the followmg reachon

PdCi +CO+H O—-> Pd+CO +2HCI

Carbon monoxu:le is ollowed fo react W|'rh 0.01 megq
of PdCl in 0.001 N HCI; the solution also con-
tains MgCI to flocculate the colloidal palladium
that is formed in the reaction. The net increase
in hydrogen ion concentration, determined by
hfrchng with a base and using an indicator of
bromophenol blue, is a measure of the CO that
was originally present. ‘In the present study, it
was necessary to use a greater amount of PdCl,

(1 meq) and to eliminate HCI in the preparahon

of the reagent. Poiassium chloride proved to be

an effective substitute for HCl in the dissoluhon
of PdCI :
Tifrchons of synthetic mixtures of HCl and

- PdCI2 were performed to find suitable conditions
for a determination of HCl in such solutions,

Since solutions of PdCl, exhibit considerable
buffering action at a pH of 4.5, it was not possible
to determine HCl in a solution of PdCl,,. Several
means of removing the palladium ion were investi-

- gated, and the addition of Kl was found to be

effective for this purpose. Excellent titration
curves ‘were obtcuned when an amount of Ki, in
excess of the amount of F’c{CI2 was added to the

solution containing HCI and PdCl, prior to the:

titration of HCI. The change in pH, during the

~ titration, was observed by means of a pH meter,
‘and plots of pH agcnnst ‘the quantity of base added

were prepared from the data. These plots are
shown in Fig. 9.3. Subsequent studies are planned

to investigate the appllcoblhty ‘of this method to o
'the determination of CO in the off- -gases resultlng L —

frpm fhe brommohon of metallic ox:des. __Reactor Chems"y 99 2838
AN_P SVERVICE LABORATORY o | O '

R W: l|ams;

 

    

     

<W F chru-ghan . o
: A.nalyncai Chem'stry‘D‘Vlslon“;'/—”"”

 

___Iabo'rkcxtory durmg 'rhe past quar’tel were m:x’rures';":i"” T

  

 

PERIOD ENDING SEPTEMBER 10, 1955

of fluoride salts containing some, or all, of the
of the foilowmg cations: Zr, U3+ U4+ Nc Li,
Be, and K. Several alkali-metal and alkali-metal-
alloy samples were also received. These samples
were received either in sealed metal tubes with
20- to 100-ml thick walls or in sealed glass bulbs.
Most of the requests were for the determination
of oxygen, nickel, chromium, and iron.

Other types of materials that were received for
analysis were

RbF-LiF FeF

Ti Metal FeF§

Cq(OH)2 Zrl:4

(NH4)2C03 Zr02

CrF2 ZrCI4

CrF3 Nc:ZrF3

UF4 (NH4)2SnF4

UF3 Scale from loop sections
Nil::2 Residues in cold traps
RbF BeC12

Bi and U mixture

A total of 1439 samples was analyzed, on which |

7195 determinations were made. The backlog
consists of 159 samples. A breakdown of the
work is shown in Table 9.1.

 

”E J. Conway, Mzcrodll.fuszon Analysis and Volu-
metric Error, 3d ed., p 257, Lockwood, London, 1950.

TABLE_ 9.1. SUMMARY OF SERVICE ANALYSES
REPORTED

 

Most of":the samples rece:ved by the serv;ce\_’fi"”

Number of Number of

~Samples Determinations

 

f_Fue: Production 77 521
Expenmental Engkltneermg 522 " - 2764
= Miscellaneous M 72

 

177

 
o

P

 

 

 

ANP PROJECT PROGRESS REPORT

pH.

178

UNGLASSIFIED

ORNL— LR —DOWG 8951

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

12.0
‘.—-".-.—-‘-
o~
11.0 —
SOLUTION 1: ./’
0.998 meq HCI ,o
10.0 1.0 meq PdCl, Il
1.4 meq KI ®
9.0 i
SOLUTION 2:
1.533 meq HCI
8.0 1.0 meq PdCl,
7.0
— SOLUTION A
6.0 ®
5.0
SOLUTION 2 /’/.
4.0 / ' L>I
o L~
3.0 —— —oo?
w
2.0
1.0
o 2 4 6 8 10 12 t4 16 18 20
NaOH ADDED (mi}
Fig. 9.3. Effect of Kl on the Titration of PdCl, and HCI with NaOH.

22

-

 

 

 

 

 

 

 
G. I. Cathers

PILOT PLANT DESIGN

o

A project analysis will be completed soon that

will serve as a basis for establishing an accurate

. schedule for design and construction of the pilot

plant for recovering fused-salt fuels, A new cost

estimate will also be made. Certain delcys in

design and procurement make a construction com-

'3 -~ pletion date near the end of February 1956 more

{ " redlistic than the December 31, 1955, date previ-
1. ously planned.

An engineering flowsheet was issued that is
subject to revisions as needed to stay abreast of
laboratory work. Approximately 65% of the proc-
ess equipment items are on hand or are in some
stage of procurement or fabrication.

g . The dump tank containing the ARE fuel was
& ~ moved uneventfully from Building 7503 to Cell 3,
i | Building 3019, on July 27, 1955, It is no longer
I  planned to force the molten fuel out of the dump
tank by nitrogen pressure, Instead, the dump tank
will be inverted inside a fumace liner, and the
fuel will drain out of the dump tank into the liner
and thence to a heated pressure vessel, where it
will be stored in the molten state until processed

The moHen fuel will be forced from the pressure
"'”’vessei or hold fonk mfo the fluormchon vessel

by nn‘rogen pressure. - S A

e
e

 

T

 

   

Lt
¥y :

 

o "sccmmng other equipment in the cell,

     

E NGINE RING D VE LOPME NTS

Dtrecf-resustcnce heofmg was tes'red becouse of
its simplicity, as a means of preventing plugging
in the transfer lines between the fluorinator and

-

 

 

tve uranium-bearmg salfé -

A :closed-cn'cuat telews:on system wnll be
‘Used as an _‘cud in posmonmg “the cans and for '

PERIOD ENDING SEPTEMBER 10, 1955

? 10. RECOVERY AND REPROCESSING OF REACTOR FUEL
; F. R. Bruce
| ' D. E. Ferguson W, K. Eister H. E. Goeller
[ M. R. Bennett J. T. Long
| F. N. Browder R. P. Milford

S. H. Stainker

Chemical Technology Division

the ARE dump tank or the waste-salt receiver.
With I/A-in.-dia 0.035-in.~thick-wall Inconel trans.
fer lines, a cutrent of 75 amp was sufficient to
keep the salt molten, except at fittings, where
supplemental extemal heating was necessary, It
was recommended that such heating units be built
into the piping in the Fluoride-Volatility Process
Pilot Plant,

A freeze valve (Fig. 10.1) was designed for
closing the molten salt transfer lines leading to
and from the fluorinator, since no reliable mechani-
cal valve is available. The valve operates by the

PHOTO 15053

_a— VENT AND NITROGEN INLET

   
 
 
  
 
  
 
 
 
 
   
   
  
 
  
  

IDISENGAGING |

SALT INLET
DIAPHRAGM

(BACK OF
'DOWNCOMER)

  

  
 
 
  
  
     

Fig. 10.1. Freeze Yalve for Molten Salt.

179

 
Rt i s

 

 

 

 

 

ANP PROJECT PROGRESS REPORT

freezing of a plug of salt in a vented trap in the
line. The salt outlet is barely visible behind the
downcomer in the phofograph. Novel features of
the design are the provision for inertgas blow-

‘back through the inlet pipe and the conical bottom

to minimize holdup within the valve and to lessen
the mechanical strain imposed by expansion of the
salt during melting, After 15 cycles of freezing
and thawing, this valve, when frozen, held against

" a test pressure of 20 psig without [eaking.

" The nature of gas dispersion through a percola-

~ tor type of gas-liquid contactor was studied, and
 the liquid recifculdfion rate was measured as a
' functlon of gas flow rate, gas inlet configuration,
“and percolotor tube length, diameter, and sub-
mergence, The data are being analyzed for use in
~designing a fluormcn‘or for the fluoride-volatility

' .process. o

e -;;-,;j_‘ PROCESS DEVELOPMENT

An Improved procedure for deconfcmlnufmg the

 

UFé product of the fluorination step was de-
veloped (Fig. 10.2). The procedure is based upon
UF  “absorption on NaF at 100°C and desorption
by heating to 400°C, with the product gas passing

NoF ABSORBENT BED
(ABSORPTION OF Ug AT $00°C,
DESORPTION AT 100 TO 400°C

—

 

through a second bed of NaF before collection of

the UF . in a cold trap. The over-all gamma de-
contammcn‘lon factors of greater than 10° that were
obtained are to be compared with the decontamina-
tion factors of only about 104 obtained with the
process in which a single bed at 650°C! or at
100 to 400°C was used. In the single-bed proc-
ess, cross contamination occurred because of the
use of the same lines for collecting the product
and for handling the waste gases containing small
amounts of fission products, and, as a result, the
decontamination factors are much lower in re-used
equipment than in new equipment. Since fission
products never enter the product-collection system
in the two-bed process, decontamination factors of
greater than 10° were obtained in re-used equip-
ment, Preliminary results indicated that the use
of nitrogen as a sweep gas in both the fluorination
and the NaF absorption and desorption steps re-
duced the amount of fluorine required for the
process. o

In the first decontamination studies, a single,

 

1F R. Bruce et al., ANP Quar. Prog. Rep. June 10,
1955, ORNL-1896, p 181.

ORNL—LR~DWG 8952

COLD TRAP FOR VOLATILE FISSION PRODUCTS

/ F, WASTE

COLD TRAP FOR UFg PRODUCT

 

 

 

 

 

 

UF; ABSORPTION STEP
—_—

 

 

 

 

 

F, '
UFg+ Ny +Fp
Bd.f.~200

FLUORINATOR—»
600°C

 

 

ol
ARE FUEL

{~ B mole % UF,

IN NaF = ZrF,}

 

 

 

\Q

 

 

 

 

DY

 

WASTE SALT
© >99% FISSION PRODUCTS
. <0.02% URANIUM

an. 10 2 Fused Salt F|“°"de'v°|°*'|“y Process in Which Two NoF Absorbent Beds Are Usedr o

 

 

e F, WASTE
/ Bd.f =10% T0 105 2

 

UF. DESORPTION STEP
- e

A\

 

 

 

 

 

/|

NaF ABSORBENT BED
{ ABSORPTION OF FISSION
PRODUCTS AT 100 TO 400°C)

  

 

e

 

 
 

e
L2 P

 

 

[ .

| 18-in.-long bed of NaF was used. The UF, efflus

ent from the fluorinator was passed through this
bed at 100°C, and then the flow was cut off while
the bed was heated to 400°C to desorb the UF .
As shown by analyses of the contents of the cold
trap after the absorption step and of the residual
NaF after the desorption step, 50 to 90% of the
volatile ruthenium passed through the NaF at
100°C, and more than 99% of the absorbed ruthe-
nium was not desorbed on heating to 400°C, Es-
sentially afl the niobium was absorbed at 100°C
and was not desorbed at 400°C, which gives a
decontamination factor of about 10% for the absorp-
tion-desorption step. When new tubing and equip-
ment were used, the over-all decontamination fac-
tors for the single-bed process were about 104
(Table 10.1). |

When the gas lines and equipment were re-used,

‘ruthenium that had been deposited in them in pre-

vious runs prevented good product decontamina-
tion, To avoid this contamination, two 9-in,-long
beds were fried. The UF, was absorbed in the
first NaF bed, the unabsorbed fission products

PERIOD ENDING SEPTEMBER 10, 1955

being collected in a cold trap. The line to the
cold trap was closed during the desomption step,
and the product stream was passed through the
second NaF bed for absorption of any fission prod-
ucts desorbed from the first bed or from the walls
of the lines between the first and second beds
(Fig. 10.2). In two test runs, UF, product con-
taining about the same beta and gamma activity as
natural uranium, or less, was obtained, The gross
beta ot gamma decontamination factors for the
whole process were of the order of 105, Because
of the low product activity, calculation of the
various specific decontamination factors was not
practical, The effectiveness of the method, how-
ever, was shown by the distribution of activities in
the two NaF beds and the cold trap used in the
absorption step (Table 10,2),

A yield of only 40% was obtained in the two runs
because of poor temperature control of the 9-in.«
fong beds. In two later runs, 6-in.-long beds with
better insulation and heating control gave yields
of 90 to 100% with the same high decontamination
factors of about 105, The same NaF was used in

TABLE 10.1, DECONTAMINATION OF UF, IN THE SINGLE-BED FUSED-SALT
FLUORIDE-VOLATILITY PROCESS

u F6 in F-N, gas stream from fluorination of NaF-ZrF ,-UF , (gross beta activity
per milligram of U in salt =5 x 10° counts/min) at 600°C; absorbed at 100°C and
desorbed with excess F2 by increasing the temperature from 100 to 400°C

Absorbent:
F2/U mole ratio: ~5
NaF /U weight ratio: ™6

200 ml of 12- to 40-mesh NaF in l-in.-dia bed

< ..PdelfJ»f:fyiéid:" 870

 

 

  
 

" Decontamination Factors

 

 
  

 

 

 

 
  
  

 

' Radioactivity ~ Over-all,
S Absorption™ Desorption™* Including
Sl e o Fluorination
. Gressbeta a2 40 - 12x 0t
.7 Gross gamma a2 310 14 x 10
‘Rugamma 24 46 '7 1100
 ZeNbgamma 10 1600 | 5.9 x 104

 

*Based on activity not absorbed with UF6 on NaF but passed into cold trap.

**Based on activity remaining on NaF after desorption of UF6'

181

 
TABLE 10. 2 DISTRIBUT!ON OF ACTIVITY IN THE TWO-B ED FUSED-SALT FLUOR!DE-VOLATlLlTY PROCESS

UF6 in F2-N2 gas stream from fluorination of NoF-ZrF -U F4 (gross beta activity per m:”lgrom

“ofUin salt =5x 10° counts/mm) at 600°C; absorbed on first bed ot 100°C, with some ochv:ty
permtfiecl to pass into cold trap; desorbed with excess F by increasing the temperature from
'IOO to 400°C with gas passmg from first bed through Second bed to UF cold trap

Absorbent beds 7 100 ml of '|2- to 40-mesh NaF in 1-:n -dlu fubes .
Tota_l Na_F/U weight ratior ™~ 6

 

Percentage of Total Volatilized Activity

 

 

 

 

 

 

 

 

ety o o
S In Cold Trap _ InBed1  InBed2  InCold Trop  InBed 1  In Bed2
Gross beta 51 48 0.8 59 7
Gross gamma 3 97 0.07 7 93 o002
Ru go.mmo o 81 | 18 0.9 86 1';1' | Very . Tow
ZeNbgamma T o4 ~00 0.4 0.8 99 0 02
' . 0.-1 . 2 97 i Very Iowki Co

_'ll'ofc:i rare eofl'h beta 3 97

 

 

rithe second run"a‘s in the flrst ond 'rhus, for the

first time, it was established that activity can be
~prevented from seriously contaminating the product
- UF process lines. Since the NaF/U weight ratio
in each run was 4, the over-all ratio after the
second run was 2,

The uranium loss in the cold trap used in the
absorption operation at 100°C varied from less
than 0.001 to 0.04% in the four runs with the two
beds. Less than 0,1% of the total uranium proc.
essed was found in the NaF in the two runs with
the use of the same NaF beds.

in the last two runs, an equal-volume mixture of
fluorine and nitrogen was used for the fluorination
until about 75% of the uranium had been volati-
lized, pure fluorine being used to remove the last

- of the uranium from the molten salt, The induction

period observed previously? was eliminated, and
uranium losses in the waste salt were only 0.013

and 0.004% with F,/U mole ratios of 3.7 and 5,
respectively. The over-all F,/U mole ratios, in-
cluding the fluorine used for desorbing UF, from
the NaF, were 5.6 and 6.7, which are somewhat
lower than the ratio of 9 used in previous! NaF
decontamination studies. |

Preliminary results also indicated that nitrogen
can be used to replace part of the fluorine used in
the NaF desorption step. A 10-g charge of 12- to
40-mesh NaF in a "/Z-in.-dia nickel tube at 100°C
was saturated with about 9.5 g of UF, and then
raised to 400°C for approximately 30 min to desorb
the UF, while nitrogen was being passed through
the tube at a rate of 200 ml/min. In two trials,
approximately 80% of the UF, was removed, This
method has not yet been used with activity present.

 

2p. E, Ferguson et al.,, ANP Quar. Pog Rep. Manr

 

 

 

 

 

 
%
5
i s
i .
i
4.
<
*
®
~
*
r
.,
t 4

 

Part |l

SHIELDING RESEARCH

 

 

 
 
 

o L

 

 

 
»

 

 

»

11

SHIELD DESIGN

J. B, Dee

-

C. A, Goetz

J. E. Smolen

H. C. Woodsum
Pratt & Whitney Aircraft

A survey of spherically symmetric unit shields
for circulating-fuel reactors was made, and weight
estimates for unit shields with various dose rates
at 50 ft are presented. The chief sources of radia-
tion in a 300-Mw circulating-fuel reactor for the
NJ-1 power plant were determined. A parametric
weight study of the shield weight dependence on
the dimensions of a 300-Mw circulating-fuel reactor
is described,

WEIGHTS OF SPHERICALLY SYMMETRIC UNIT
SHIEL DS FOR CIRCULATING- FUEL REACTORS

Y survey of spherically symmetric unit shields
for circulating-fuel reactors was made for a range
in dose rate of 0.1 to 10 rem/hr at a distance of 50
ft and a range in reactor power of 100 to 300 Mw,
An estimate was obtained for the added weight of
an NaK-to-NaK secondary heat exchanger and its
shielding. This additional weight is quite sensi-
tive to the manner in which the dose rate is divided
between the secondary heat exchanger and the
reactor and to the absolute value for the sodium
activation. Consequently, the estimates must be
interpreted as indicating trends rather than absolute
values.

The reactor dimensions for this survey were
scaled from those glven previously ! for a 300-Mw
reactor having a power density of 2.75 kw/cm ’

“and the secondary heat exchanger dimensions were

scaled from a Pratt & W ittn‘ey Aircraft de5|gn.
’The shreld dlmensmns ‘were

 
 

 

 

p‘snon and the data

 
 
 

 

d shleld 3

      
 
 

 

   

]ANI_’ Quar Prog.
: "p 74 ;

 

Rep Mar ,

L E. P. Blizard and H. Goldstein, OR

1954),
- 3F H. Abernathy et al., Lid Tank Shielding Tests for
the Reflector-Moderated Reactor, ORNL-1616 (Cct. 5,
1954).

 

o determrned by the
mefhods g‘lven by fhe ]953 Summé;r Shle'dlng Con
efhods presén é‘dj in fhe o
A:le Tan‘k Shleldmg“"Fqcllii‘y (LTSF) report on an’ s
v S s On'\a Clrculatlng-erI s

H0wever'

"; 1953, ‘ORNL-ISIS -

alkylbenzene in such a manner as to attenuate the
neutrons as rapidly as the gamma rays and, thus,
to ensure the effectiveness of the lead; this method
is conservative. An analysis of data from the
current L TSF circulating-fuel reflector-moderated-
reactor and shield tests (see Sec. 12) should make
possible a shield arrangement that would result in
a saving of several thousand pounds of thick
gammea-ray shielding for unit shields. Additional
weight savings could be achieved by shield-shaping
according to the particular aircraft application and .
configuration. The source of data for the sodium
activation for these calculations was the current
LTSF tests, teported previously.? Results of this
survey are given in Table 11.1 and Fig. 11.1,

A
ORNL-LR-DWG 61424

=%
o
o
o
5 Lo

 

.-‘:\ 2
E
2
=
o
w0y
—
L3
w 0.5
w
o
a
0.2 M0 o
L OOO\
180
[o X
400 150 200 250 300

REACTOR POWER(MW)

Fig. 1L.1. Weights of Clrcu!ohng-Fuel Reactor

“‘and Shield Assemblies Without NaK-to-NaK Heat

Exchangers.

- $OURCES OF RADIATION IN A 300-Mw
::j:f CSRCULATING-FUEL REACTOR
_\:iThe chief sources of mdlahon for the 300-Mw

         

2Report of e 1953 e szeldm{S'esszon, ed by c:rculafmg-fuel reacfor fcr fhe NJ-T ‘power plant

 

1575 J 11,
73 Yene Ll 4G. T. Chapman, J. B. Dee, and H. C. Woodsum, ANP

Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 194.
SPratt & Whitney Aircraft, NJ-1 Powerplant Character-

istics Summary, PWAC-126 (Mar. 16, 1955).

185

 
s

 

 

 

ANP PROJECT PROGRESS REPORT

TABLE 11.1. DIMENSIONS OF CIRCULATING-FUEL REACTORS AND CORRESPONDING REACTOR — UNIT
SHIELD WEIGHTS FOR VARIOUS DOSE RATES AT 50 ft

 

Reactor Power (Mw)

100 ' 200 300

 

 

 

 

7 Dose Rate at 50 ft (rem/hr) _
0.1 1 10 0.1 1 10 0.1 1 16
All Components -Excepf Shielding and
NaK-to-NaK Heat Exchanger
Thicknesses (in.)
Core 4.09 4.09 4.09 5.18 5.18 5.18 597  5.97 5.97
Beryllium reflector 12,60 1200 12,00 1200 12.00 12.00 12.00 12.00 1200
Fuelfi-fo’-NdK heat exchanger 2.81 2,81 281 4.57 4.57 4.57 572 5.72 5.72
Outside Radii (in.) ' '

 Beryllium island 4,62 4.62 462 5.8 5.88 5.88 6.685 6.685 4.685

- Core’ o 9.04  9.04 9.04 11.38 11,38 11.38 1297 1297 1297
Beryllium reflector 21.37  21.37 21.37 23.70 2370 2370 25.31 2531 2531
Fuel-to-NaK heat exchanger 24.53 24,53 24,53 28.62 2862 28.62 31.38 31.38 31.38

Weights (103 Ib)

’ "-l-?e.ocfor“(b;‘essuw shell and con=

tents including pump headers) 13.4 13.4 134 203 20.3 20.3 27.7  27.7 27.7
Insul ation 0.18 0.13 0.18 0.24 0.24 0.24 0.28 0.28 0.28
CANETYE Syrycture 2.90 2,54 2,18 3.15 2.82 2,50 3.25 3.00 270
‘ Patches/ 2.81 2,57 2,40 47 4,16 3.77 6.66 5.83 5.25
NaK -fo‘-NVGJI‘( Hécr} ‘Exchanger
Thickness (ft) 1,507 1.504 1.50% 2.12% 2.12% 2,128 2.60° 2.60° 2.60°

Weight (10° Ib) 3.01 3,01 3.0 571 5.71 571 8.5 8.5 8.5

Lead Shielding of NaK-to-NaK Heat
Exchanger '

Thickness (in.) 3.49 1.09 © 4,48 2.09 0 4.85 2.55 0.022

Weight (103 1b) 6.55 2,50 0 14.18  5.99 0 21.14 10,28 0.08

Lead and Alkylbenzene Shield

Qutside diameter (ft) 11.65 10,64 9.32 12,58 11,23 992 12,99 11.62 10.31

Thicknesses
Total lead {in.} 10.57 8.71 7.7 1117 9.20 7.52 11.45 9.43 7.85
Total alkylbenzene (ft) 2.884 2.362 1.821 2.788 2,278 1.762 2.739 2,223 1.699

Weights (10° [b)

Total lead 70.99 49.39 36.13 93.55 66.33 48.85 109.55 78.85 59.78
Total atkylbenzene 33,27 25.06 15.91 40,92 28.12 18.08 4391 30.15 1936

Total Weight of Reactor and Lead- 123.55 93.14 70.20 162.87 122.02 93.74 191.35 14581 115.07
Alkylbenzene Shield Assembly (10% Ib)

Total Weight of Reactor, Lead-Alkyl- 126.56 96.15 73.21 168.58 127.73 99.45 199.85 154.31 123.57
benzene Shield Assembly, and Non- '
shielded NaX-to-NaK Heat Exchanger
(103 1b)

Total Weight of Reactor, Lead-Alkyl- 123.65

benzene Shield Assembly, and Shielded
_ NaK-to-Nak Heat Exchanger (]03 ib)

133.11 98.65 73.21 182.76 133.72 99.45 220.99 164.59

 

3 ft wide X 2.9 ft long NaK«to-NaK heat exchanger.
b4.24 ft wide % 3.15 ft long NaK-to-NaK heat exchanger.
€5.2 ft wide X 3.4 ft long NaK-to-NaK heat exchanger.

Ry

 

 

"~

 
5 S i i

et | e weime oonol

 

o,

*

   
   

»

have been determined.

7 The reactor dimensions
used for this study are given in Table 11.2, and
the sources of rcdia’rion are lisfed in Table ” 3.

TABLE 1.2, PARAMETERS OF 300-Mw
CIRCULATING-FUEL REACTOR USED IN
CALCULATION OF SOURCES OF RADIATION

 

Thickness

Raaius

 

Reactor Re‘g!ion' (in.) (in.)
Beryllium island 6.690
Inconel-X cladding 0.01 6.700
Sodium passage 0.1875 6.888
Inconel-X core shell 0.125 7.013
Core fuel region 5,987 13.000
Inconel-X core shell 0.156 13.156
Sodium passage 0.1875 13.343

- Inconel-X cladding 0.01 137.‘353
‘Beryllium reflector ' 11.908 25.261
fir?cone] X cladding -. 0.01 25.271
|8 10 ceramel 0.10 25.371
dnconel «X claddmg 0.01 25.381

Sodium passage - 0.1875 25.568

" Inconel- Xéclcddlng' 0.125 25.693
Heat exch}irnger 7.13 32.823
Inconel-X cladding 0,125 32.948
Inconel-X thermal shield 1.00 33.948
Pressure shell ~ 1,00 34,948
Insulation 0.25 % 35.198
Insulation cladding 0.032 ! 35.230
VAIkylbenz}ene passage 0.375 35.605
lead o 1.00 36.605
A.Ik'ylbenzene passage

36980

 

“byvan mdependent me’rhod in which the max1mum"
number of captures in the core shell was found by
“assuming thermal absorption in the core shells and

 

* “CAPTURE'RATE (captures/cm®-sec)

PERIOD ENDING SEPTEMBER 10, 1955

all thermal fissions in a blackzcore. For each
k2

neutron captured in the core, ¢ * neutrons were

assumed to have entered the Inconel. The total
inconel capture rate was therefore the core capture

rate multiplied by (e ® ~ 1), where ¢ is the Inconel
shell thickness, %, is the neutron absorption cross
section at the neutron temperature, and £ is a
constant with a value between 1 and 2 that depends
upon the anisotropy of the neutron flux., This % is
in agreement with that obtained in recent foil
measurements at the ORNL Critical Experiments
Facility, and, also, with the multigroup results,
which correspond with a value of & = 1.65. The
neutron copture rates in the reactor are shown in
Fig. 11.2,

- The prompt-fission gamma-ray spectrum used is
based on the data of R, L. Gamble (see Table 11.3),
This spectrum is thought to be reasonably correct
in total energy, although deviations by a factor of

2—-01-059-22A

ISLAND C REFLECTOR

O-
™ N

o

4
N

CORE SHELL CORE SHELL (OUTER)
(INNER) T b

10”

 

0 0 20 30 40 50 60 70
RADIUS (cm)

 
 
       

Fig. 11.2, Capture Rufes;};;l‘lv‘lle:utr'c‘b.hs in a
300-Mw  Circulating-Fuel Reactor (Based on

Foxcode Multigroup Calculations).

187

 

psillcoiili i c Rl s S Rl sl s

bl s catiiad ot i

 
ANP PROJECT PROGRESS REPORT

i

 

 

 

i G

 

 

 

 

 

 

TABLE 11.3. SOURCES OF RADJATIONIN A 300-Mw CIRCULATING-FUEL REACTOR :
- . o o | ) Photon i Photon Emission : Neutron Cra;:a;ur‘es'r " Neutrons .
Region Photon Source ' Efiérgy(a) Probability per Second Produced ok
o - T (Mev) per Copfure(a) at 300 Mw per Second 7
Asland ' n,y) in beryllivm 68 075 0075 10'®
» o ~3.4 ' 0.50
Core shell (mnel') ('nf)/)- .ifiwrinrc;fi)el‘(b) = 9 0.72 0.28 x 1018 | | | :
~4 7004 - | ot
Core Prompt-fission E 7.7 e~ 1-03E y4gle) 9.5 x 10'8 23.3x 1018 :
~ gamma rays '
: F'is:;;ioh-prodfiéf B (7.0 e~ 1-2E gp)ldie)
S ‘gamma rays ' o o X
| S ey in U238 E (8 e~ 1-03E gp)(/) 1.6 x 108
Core shell (outer)  (ap)inlnconel 9 072 127 x10'8
B L T - 04 | | e
Reflector ~ (n;y)inberylliom 6.8 0.75 138 x10'8
ey, S . ™34 0.50 -
Heaf exrcil"-aanger | Fissidfi-produ.ct E (7.0 ™ 1.2E ypy(esg)
Dl . gamma rays .
) L L ' :  12(P)
Gamma shield (n,y) in lead 7.38 0.93 6.8 x 1013
. ~6 0.07
Hydrogenous shield V(n",}/) in borated(?) __ .
: alkylbenzene ! ()
(n,y) in hydrogen 2.2 1.00 9.6 x 1013
(n,y) in boron 0.48 0.94 3.7 x 101°

 

ap, Mittleman and R, L. Liedtke, **Gamma Rays from Thermal-Neutron Capture,’”’ Nucleonics 13(5), 5051 (1955).
byalyes for nickel, '
©From data of J. E. Francis and R. L. Gamble, Phys. Semiann. Prog. Rep. Mar. 20, 1955, ORNL-1879, p 20.
4Values in parentheses are assumed, |
®Preliminary estimate of spectrum; only 20% of the fission products reside in core.

/This spectrum was assumed to be similar to the prompt-fission spectrum because it was convenient. The gamma
| p promp P g

rays from radiative capture in uranium have not been observed and their character is unknown, but they are thought to
be multiple rather than a single gamma ray. The spectrum was normalized to the binding energy of the captured

 

 

nevufron,

£30% of the fission products reside in the heat exchanger, headers, and pumps.

bNominal 4y2 in. of lead assumed,

20 mg of boron per cubic centimeter of mixture.

2 or more may exist at some energies.

It should be noted that only about 20% of the
fuel in the circuit is in the core, the remainder
being in the heat exchanger, headers, pumps, and
core end ducts. The fission-product gamma-ray
spectrum is actually unknown. The spectrum used
has only a very slight experimental basis. The
estimated spectrum given here is probably correct

within a factor of 1.6, with respect fo total energy.
Several experiments are being performed for de-
tiving a more definite estimate,®

The capture rates given for the lead and for the
hydrogen in the alkylbenzene are based on a
numerical volume integration of the thermal-neutron

 

SR, W. Peelle, T. A. Love, and F. C. Maienschein,
ANP Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 203.

 

 

 
 

i

+h

-

-

peri ment:(c":I

- sources.

fluxes measured behind a circulating-fuel reactor

mockup in the Lid Tank Shielding Facility (LLTSF).
This mockup contained 4.5 in. of lead and 2%
boron in the water. As yet, the thermal-neutron
fluxes cannot be calculated for the reactor in this
region, Also, the degree of boration to be used
has not yet been determined. Consequently, the
capture data for the lead and alkylbenzene are
based on the assumption that the fraction of the
neutrons captured in the LTSF mockup applies to
the reactor, corrected only for neutron self-shielding
in the core.

SHIELD WEIGHT DEP ENDENCE ON

THE DIMENSIONS OF A 300-Mw

CIRCULATING-FUEL REACTOR
" A parametric weight study that is nearing com-
pletion indicates that lighter shield weights for
circulating-fuel reactors may result from changes
in reactor dtmensmns. The shield weights are
belng determined for a fixed reactor power (300
Mw) and a ‘‘standard shield”’ (described below)
as a function of power density, islond radius,

“reflector thickness, and heat exchanger thickness.

The weights obtained will be used as part of the
input information in a parametric performance
study under way ot Pratt & Whitney Aircraft,

The standard shield consists of a set of specifi-
cations suff|c1en’r|y representative of tactical
bomber shaelds to ensure a realistic reactor op-
timization. Some concessions to e-xpedlency were
made, such as the selection of sea-level altitude
and a recctor-’ro-crew-compartment separation
distance of 64 ft, in order to make direct use of

the neutron dafa recently ob'rqm_ed at 'rhe Tower

Shleldmg Facnllty (TSF)

   

 

'-f'.-?'r'-ff';i|mpor’rcn’r source eglyc':ns. Because of differences
Lo neutron Ieukage between the LTSF mockup
i ‘and the desngn’_ reactor, fluxes from mulhgroup;;

    
 

LTSF were us_

ergies of fhe gammc rays orlglnahng in the various

 

_ ’G. T. Chdpmcn, J. B. Dee, and H. C. Woodsum, ANP
Quar. Prog. Rep. June 10, 1955, ORNL-1896, p 194.

 Tank Shielding Facility'’).

‘créew compartment,

“aceording to preliminary results.
“*“particular case was chosen for which the neutron
- shield weight was close to minimum and for which
o~ close-to-minimum lead thickness was required
"""’on 'rhe crew-compcrfmenf sides,

PERIOD ENDING SEPTEMBER 10, 1955

regions, the various dose-rate components were
treated separately to account for the different
attenuation characteristics in the shielding ma-
terials. In considering air scattering, the dose
rate from neutron captures near the outer lead
surface was treated as a separate isotropic source
that was relatively independent of lead thickness.

A second difference arises from the use of TSF
fast-neutron experimental data for air scattering
and crew-cempcrtmem‘ penetration and the use
of an optimized shape for the neutron reactor
shield and the crew-compartment wall thickness.
A third difference arises from the use of a tapered

shadow shield.

For these calculations, the crew compartment
was assumed to be a right cylinder, with an area
of 35 ft2 on the rear and 225 ft* on the sides.
Plastic and lead were the shield materials used
to give a neutron dose rate of 0.25 rem/hr and a
gamma-ray dose rate of 0.75 rem/hr in the crew
compartment at sea level. The reactor-to-crew-
compartment separation distance used was 64 ft.
The reactor shield, which consisted of lead and
alkylbenzene (330°F), was designed to give a
maximum dose rate of 1000 rep/hr at 50 ft (as-
sumed radiation damage restriction).

The neutron shield was chosen in the following
manner, For a typical aircraft reactor (Table
11.2), the fast-neutron dose rate as a function
of the thickness of the spherical shield was ob-
tained by the application of conventional meth-
ods®? to data from the current LTSF circulating-
fuel-reactor shielding tests (see Sec. 12, ‘‘Lid
For this reactor and
an optimization study was
initiated at the TSF to determine the minimum-

‘from those mude prev:ously in several respects.m“("'kweigh" neutron shield and its shape (see Sec. 14,

One of the dlfferences arises from the use of the_' ~ Tower Shi
S “*% the neutron shield weight appeared to be broad

““Tower Shielding Facility'’). The minimum of

with respect to degree of division of the shield,
Therefore a

For this case

 

Becauee of A'rhe fchffe-rent primary en-""""'

Report of the 1953 Summer Shieldin Sesszon, ed. by
E. P. Blizard and H. Goldstein, ORNi 1575 (June 11,
1954), p 293 ff.

9F. H. Abernathy et. al., Lid Tank Shielding Tests for
the Reflector—Modemted Reactor, ORNL-1616 (Oct. 5,
1954), p 79.

189

 
ANP PROJECT PROGRESS REPORT

the plastic thickness at the rear of the crew com-
partment was 43 cm and at the side, 14 cm, Be-
cause the neutron shielding on the crew compart-
ment was unchanged throughout the parameter

study, the fast-neutron dose rate at 64 ft from

the reactor at the various values of the reactor
angle @ (where 0 is the angle measured with re-
spect to the reactor-to-crew-compartment axis)

- also remained unchanged. A parameter change

in the reactor that resulted in a change in the
fast-neutron dose rate at any angle ¢ was com-

" pensated for by an adjusiment of the neutron

shielding at that angle 6 to keep the fast-neutron
dose rate at 64 ft constant. In the TSF calcula-
tions, the angular distribution of the neutrons
escaping from the reactor shield surface was
assumed to be éos Y in calculating the direct
dose, where ¢ is the angle measured with respect
to the normal to the shield surface, while radial
emission was cssumed for the calculation of the
scattered dose.

With the neutron shielding fixed, the gamma-ray
shielding was taken to consist of four components:
the basic lead shield required to maintain a maxi-
mum dose rate of 1000 rep/hr at 50 ft (an crbn‘rary
radiation damage limitation); the shaped lead
shadow shield adjacent to the basic shield; the
lead on the side and front of the crew compart-
ment; and the lead on the rear of fhe crew come
partment,  The shadow shield was arbitrarily
shaped to permit the dose emission per unit
angle § to increase exponentially to compensate
for the exponential decrease in the probability
of scattered radiation penetrating the crew-
compartment side'® as 6 increased. An approxi-
mation for the effect of the angular distribution

-of the gamma rays escaping from the reactor
- shield surface wds made by assuming a ‘‘dis-
‘ 'cldvanfage angle a (assumed to be 15 deg) when
:cclcula’rmg the scattered radiation penetrating

the crew compartment. This assumption results

in a constant thickness portion of the shadow
‘ shlel_d that subtends d half-angle of 15 deg with
‘respect fo the reactor-to-crew-compartment axis,
__._fo_“owfiéd by a tapered section that decreases

 

'(1) D= D) H(a)

 

it e L g ST

10

The scattering-penetration probabilities used were

those calculated by NDA; Report of the 1953 Summer
"Shielding Session, ed. by E. P. Blizard and H. Gol dstein,
ORNL ]575 (June ” ]954), p 188--205.

 

‘rr o

 

1
a2 LT

190

c's R —SZtHX

linearly in thickness, with the angle 6, until it
reaches zero. The tapered section ends at 6§ = 63
deg for a 3.23-in. shadow shield shaped with a
12.5-deg relaxation angle and a 2.1-cm relaxation
fength in lead for the primary dosé. The basic
lead shield does not have a constant thickness
but decreqses from rear fo front 1o compensate
for the increased gamma—ay attenuation from the
increasing neutron shield thickness along a radial
path., |

The gamma-ray shield division chosen, under
the ‘assumptions outlined above, consisted of
0.1 in. of lead on the crew-compartment sides
and approximately 6 in. of lead, ‘in addition to
the basic shield for the direct dose, which was
divided into a 2.56-in, shield on the crew-compart-
ment rear and a 3.23-in.-thick tapered shadow
shield near the center; the exact d|v15ton of the
basic shield was not sensitive wsth respect to
weight, The shadow shield was in direct con-
tact with lead in most of the cases considered.

. Spacing was. employed, however, for some cases

in which the reflector was thin and for which the
secondary gamma-ray dose contribution was large.

The first step in the weight calculations was
to determine the shield necessary to give the
design dose rates from the rear of the shaped
reactor shield. The design dose rates were
108 rem/hr at 50 ft for fast neutrons (10,8 rep/hr)
and 1000 rem/hr at 50 ft for gamma rays (1000
rep/hr),

Neutron Shield

The neutron-attenuation data were taoken from
fast-neutron dose-rate curves obtained at the
LTSF by using the Glass-Hurst fast-neutron
dosimeters and the Hornyak button. The data
were extrapolated for thicknesses beyond the
range of the LTSF fast-neutron dose-rate instru-
ments by means of thermal-neutron-flux traverses
in the LTSF and fast-neutron dose-rate measure-
ments at the TSF. Neutron dose-rate curves were
available for beryllium reflector moderator regions
8, 12, and 16 in. thick and o 4-in.-thick heat ex-
changer region; the data were corrected for air
gaps, aluminum tank walls, and Boral regions,
The basic neutron-shield thickness was determined

from the expression

-8, (szo"zAlky)‘Alky
e

e '

 

-

 

 

 

 

 

 
 

" D(z) = LTSF fast-neutron dose rate at z

where _
D, = fast-neutron dose rate at 64 ft, 6.6 rep/hr,

for the
particular reflector thickness,

z = shield thickness measured from the surface
of a sphere representing the core of the
reactor, o |

H(z) = ratio of the fast-neutron dose from an

" infinite plane to that from the LTSF source

disk, taken to be (1~ 630/)\(2)2]

o= equivalent surface source strength for the

LTSF, taken to be 6 w x 0.6 (leakage
factor) = 3970 cm?,

= equ:vclent surface source strength for the
~reactor, taken to be

-

s =t /A
pA| 1= — e e
c

ey,

o
c

A_ = reciprocal of the neutron removal cross

section for the fuel NaF-ZrF -UF (50-46-
4 mole %) at a density of 3. 2 g/cm , taken
to be 11.6 cm,

p = power density (w/cm3')',
r_ = outer radius of the core (fuel annulus),
= inner radius of the core (fuel annulus),

r = outer radius of the fofcl réaétor S‘hield,

e _‘_. = correction by effechve removal cross
R sec ons from the ‘common 4-in.-thick heat

 
  

' exc__

25 vol % Ni, 75 vol % NaF at a dens:ty of

   
  

t

yx = reactor heat exchanger thic kness,

VZHX = effective removal cross section for the

) g/cm3 to a 3.5in.-, 65-,,,_ Cor &in.
1h7|_ck heot exchcmger hcvmg a C°mposmon'" y

onel X, 273 vol % fuel, ,
- " “ponents of the LTSF data were available for several

-

PERIOD ENDING SEPTEMBER 10, 1955

reactor heat exchanger (homogeneized),

taken to be 0.0659 em™},

= effective removal cross section for the

2
HX,

mockup heat exchanger of the LTSF experi-
ments (homogeneized), taken to be 0.0869

em”},

S’H o = effective removal cross. secfion for
2° \ater, taken to be 0.0994 cm™7,

EA“( = effective removal cross section for

alkylbenzene 350 at 330°F, taken to be

0.0882 5,
_ ztR cm

e = effective removal cross-section cor-
rection for minor differences between the
LTSF configurations and reactor configura-
tions,

tp = sum of thickness differences,

—(EHzo‘EAlky>tA|ky ‘

e = effective removal cross-
section correction for the substitution of
alkylbenzene for water, '

t = thickness of alkylbenzene at the reactor,

Alky
d = separation distance from reactor center to
rear face of crew compartment, 64 ft.

Gamma-Ray Shield

" The gammeo-ray shield was determined by finding,
separately, the gamma-ray dose rate at 50 ft caused
by (1) gamma rays emitted by the fuel in the core
and by neutron captures in the outer core shell and
in the reflector, (2) gomma rays emitted by fission
products in the heat exchanger, and (3) neutron

7 _._.captures near the lead-alkylbenzene interface. For
nger ‘mockup havmg a composmon of

each reactor shield design these three groups of
contributions were determined separately and then

su rh_mt_ed.

Reactor Gamma-Ray Dose Rate. Dose-rate com-

fead thicknesses for each reflector thickness (8,
12, and 16 in.). The gamma-ray dose rate data
were divided into contributions from the source
plate, capture gamma rays from the beryllium re-
flector, capture gamma rays from the Inconel core

~ shell, and secondcry gamma rays, principally from
" ““neufron captures in the lead, boron, and water near

the lead-water interface. The source correction
was determined by assuming an exponential power
distribution through the natural uranium of the

191

 

 

 
g ]

ANP PROJECT PROGRESS REPORT

source plate and determining the self-shielding

 

factor, s: :
S !
S
: L o e t
- o 'O
s = _ =4,02,
¢
u - — — L]
f (}5 e Eaz e ‘uu(tu t) e Fppipy d
0
o
where

¢, = thermal flux incident on the reactor side of
the source plate,

3,  =total thermal absorption cross section for
natural vranium,

{,, = gamma-ray absorption coefficient, 4 Mev,

pps = gamma-ray absorption coefficient for the
lead over the source plate,

effective source plafe thlckness, 7/4 in.,

~~
I

i

pp = lead thickness, 1/2 in.

 

where

i, = effective mass absorption coefficient for
the core for a gamma-ray energy of 3 Meyv,
t = thickness of fuel annulus,

pt,. = typical core fuel annulus thickness in
gamma-ray mean free paths,

in the natural uranium of the source plate, three-
eights of the thermal-neutron captures take place
in U238 and give radiative capture gamma rays
that are not present in the enriched reactor, Un-
fortunately, the energies and associated probabili-
ties of emission per capture for these gamma rays
are not known. For this study, a total energy of
6.4 Mev per U238 capture was assumed to be
emitted in a distribution corresponding to the
fission gamma-ray distribution,  Similarly, for
radiative capture in U235, a total energy in gamma
rays of 7.6 Mev per nonfission capture was as-
sumed. The fraction of gamma-ray dose attributable
to U233, t, in the source plate was then taken to

be:

E,(235) 3,(235) + E (235) = _(235)

 

-—

2 7 E(235) 5,(235) + E_(235) 3_(235) + E_(239) T _(238)

 

 

Since only about one-fifth of the fission products
are in the core, the core dose rate should be

= 0.57 ,

where

E (238) = total photon energy per capture in

 

multiplied by the fraction of energy from the core Y238
while fuel is circulating, ¢,:
E _(235) % (235) + Ef(235) Ef(235) + fp(235) Ef(235)

= 0.867 ,

 

"7 TE_(235) £, (235) + E,(235) %,(235) + E, ,(235) %, (235)

where

E _(235) = totdl photon energy per capture in
U235
7

Er(235) = total photon energy per U237 fission,

(235) —photon energy from fission products
per U235 ftss:on,

EC(235) = nonfission capture cross section for

U235’
' Ef(235) = fission cross section in U235,

The reactor gomma-rdy core leakage factor, I,
was esnmofed to be:

ol L ooars,

 

2% _(238) = capture cross section for U238,

The total scaling factor to be applied to the core

source for the same power per unit area of source
surface is the product

f=sxt, xIxty, =09,

The scaling factor for the capture gamma rays
from the reflector was obtained as the ratio of the
fraction of neutrons captured in that region in the
reactor (as determined by application of Eyewash

code neutron cross section to Foxcode multigroup -
fluxes) to the fraction of neutrons captured in the.

same region in the LTSF mockup (as determined
by numerical integration of LTSF gold-foil flux

 

 

 

 

 

 
DAl i i, LA i AR
K&_Lm
' -l

S A bl u abie g i, kit i

 

oy

M,

el

o

w

measurements). This ratio was 0.0592/0.0145 or
4.00. The contribution from the island was calcu-
lated to increase this to 4.05,

 The fraction of neutrons captured in the Inconel
core shell in the reactor was estimated in the same
manner as that from the reflector and was checked
against foil data from the Critical Experiment
Facility. The fraction captured in the LTSF mock-
up was also determined by integration of gold-foil
‘measurements. The scaling factor for the outer
core shell contribution is the ratio 0.0545/0.0161
or 3.39. Similar corrections applied to the inner
core shell raised the scalmg factor for the core
shells to 3.62. ,

- For each reflector thickness (8, 12, and 16 in.)
and for each lead fhlckness (3, 4.5, and 6 in.),

" these correction factors were applied to the com-
 ponents determined from the LTSF data. The

‘variation of the reflector component of the total
gamma-ray dose rate with reflector thickness, as
obtained from the LTSF dm‘a, was observed to be
in agreement wn‘h the variation for a reactor, as
determined in PWA-NDA multigroup calculations,
and therefore it was assumed that nonc of the
scaling factors changed with reflector thickness,
_ The sum of the dose-rate contributions from the
source plate, the capture gamma rays from the
beryllium reflector, and the capture gamma rays
from the Inconel core shell, that is, Dy, (a,z2),
for each reactor design was transformed to the
corresponding dose from an infinite plane by

 

o

  
 

B(ut) E ,(u2) N(E) ME)

 

 

,u..-— effechve mass cbsorpflon coefflcrem‘ for
the shielding materials for a gamma-ray
energy of 6.8 Mev,

PERIOD ENDING SEPTEMBER 10, 1955

E,(uz) = exponential integral whose argument is
the number of mean free paths from the
source to the detector,

The primary dose at a distance d from the reactor
was then determined by the expression:

D, (d)
rr PR/4n'r3_

cC S

 

— H{a,=) Dp i fla,z) cicqeq
d? PLT/mz2

]
It

reactor power, 3 x 108 w,

1.7 = LTSF source plate power, 6 w,

c, = change in attenuation due to replacing the
4-in.-thick test heat exchanger mockup
with a particular reactor heat exchanger
composition and thickness (based on
exponential attenuation at 6.8 Mev),

= change in attenuation due to addition of
small claddings, correct pressure shell
thickness and composition, etc. (based
on exponential attenuation at 6.8 Mev),

= change in attenuation due to substitution
of alkylbenzene-350 at 330°F, assuming
that . = 0.8 py 0! based on electron
density.

Heat Exchanger Dose Rate. The dose rate from
the heat exchanger was calculated by using the
following expression:

 

T

HX, i L /ME) AE) ~t ./ AME)

R
’Hxi Tux

o

dE ,

   

2C(E)

= volume fraction of heat exchanger
that contains fuel,

= volume of fuel in total circulating

) system,

U"‘:""‘=:?oui's ide rodlus of heu’r exchanger,
= inside radius of heat exchanger,

fission product gamma rays of

energy E emitted per unit energy

per fission (taken to be 7.0
e~ 1+2E) (see Table 11.3),

193

 

T -

 
 

 

. »‘.E‘ ki

 

ANP PROJECT PROGRESS REFORT

C(E) = number of phofons‘ of energy E
" equal to 1 rem/hr, 1T
E}'(pi) = exponenhal integral whose argu-

~ment is the shield thickness in
. mean free paths,

B(uz) = the dose buildup factor,

‘tHx = heat exchanger thickness,

ME)

reciprocal of gamma absorption
“coefficient for a mixture of the
R mcn‘erla Is in the heat exchcnger,

3] X ]0]0 = number of f:ss:ons per second
- per watt,
The above 1nfegral was evaluated numerically

\.:-,over a range ‘of thicknesses for the heat exchanger,
- lead, cmd water, and it appeared to be valid over

the range required to separate the self-shielding

- factor from the remainder of the integral. An
" average " self-shielding factor was derived which

depended only upon the heat exchanger thickness
and which was in agreement with the more exact
integration to within 7% over a range of lead
thicknesses from 0 to ¢ in. and a range in water
thicknesses from 40 to 100 cm. Thus, to simplify
the calculation procedure, the integral was broken
up into the following product:

where

tp, = reflector thickness,
tpy = lead thickness,

Eatky = dlkylbenzene thickness,
Top = radius to outside of lead,

-~ H(a’,t) = lead surface neutron source to
infinite plane neutron source trans-
formation calculated by ratios of
exponential integrals, as before
(from LTSF flux traverses, a’ is
dpproximately 45 cm),

H(a,zpb) = lead surface neutron source to infi-

“nite plane neutron source frans-
formation, taken to be
. -84/z
/(1 - e Pb)
t;, = thickness of coohng luyers in the

lead basic shield,

’\l = relaxation thickness of cooling
layers for secondary dose (deter-
mined to be 5.33 ¢m from correcting

LTSF data for spaced lead by the

 

 

AMEY | T =

 

r

HX HX
o

Q

The first expression represents an average self-
shielding factor as a function only of the heat
exchanger thickness., Evaluation of the second
factor for a series of lead thicknesses results
in a curve of gamma-ray dose rate as a function
of lead thickness for an infinite plane source
of fission products,

Secondary Dose Rate. The secondary dose rate
(from captures in the lead, boron, and hydrogen
near the lead-water interface) was determined by
scaling the neutron dose rate in the source region
(lead-alkylbenzene interface) as follows:

"‘,:.‘Ds(d) = Dg_; rltgertppitap,) Hla ,ztA“W)—---*—-——d2 Hla,zp,,) =

194

HX; 1, /MEY  ME) [ =t/ ME)
: HX + ()(e HX __1)

* B(pz) E,(ut) N(E)
- dE .
2C(E)

 

effective removal cross sections for
alkylbenzene),

= ratio of reactor to LTSF equivalent
LT surface source sitrength for neu-
trons, '

 

 

”Calculated from an expression in ORNL. CF-51-10-
98 [F. L. Culler, EBR Fuel Element Design {Oct. 15,
8951)] and from absorption coefficients in NBS- 1003
G. R. White, X-Ray Attenuation Coefficients from 10
kev to 100 Mev (May 13, 1952)1

 

"ob's e %R _;,/Al"' |

 

b LT

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

        
   
 
          

_: 2 ' - _ Lol : PERIOD ENDING SEPTEMBER 10, 1955
1 3 7 4"-::%.71;:: ' weigl‘lf Deiermiridtidms ’Cre"fi"compartfilehf plastic 9,900
The weights determined include the reactor, Crew compartment rear lead 3,270 h
{ - pumps, heat exchcngers, reactor shield, crew- Crew compartment side and fronf. lead 1,530 b
compartment shiel_d, .shcdow shield, structural Total Crew Compartment - 16,700
- ‘weight, and pc:’rch weight, For example: Toral System 70,000
'Reactor No 3232 ' ‘ d
" Power density _ 4.125 kw/cm> Variations of this total system weight are shown
" Reflector fhlckness o 12 in. in Figs. 11.3, 114, 11.5, and 11.6 with respect to
" Tsland radiu's [ S T variations in island Vrcxdius, power density, heat
e Heqf exchqnger fhlgkness T 6.50n. exchanger thickness, and reflector thickness.
7 ; Welght (lb) 7 .
" Pressure shell and con'l'enfs]z 22,970 12} cludes  all layers of materials, in addition to
. ‘ - _ : o .- pumps and drives, control rod and island support, i
* : y Alkylbenzene L o _ - 6,650 sodium~to-NaK heat exchanger, manifolds, decking, h
' ' Basic lead shell = - 14,740 ducting, and expansion tanks. i
Patch v_ve'iéhf S 3,230 ' | E
) Structural weight o o 1,910 7 s
Shadow shield 7 3,800 85 2-01-059—-24
" Total Reactor cmd Reactor Sh:eld Welghf | 5.3;,300
7' 80 e 2—01—059—23 R |
I, ] I ; .
. : . je}
o POWER. DENSITY = 4.125 kw/cm® % \
o HEAT EXCHANGER THICKNESS = 6.5in. 2 _ ,
B . . S ° . ‘ b:
o REFLECTOR THICKNESS =12 in. = 80 -y
! : g
& o : ¥ HEAT EXCHANGER THICKNESS = 6.5 in.
* % . / E , REFLECTOR THICKNESS =12 in.
5 L & ISLAND RADIUS = 4.0 in. ;
= " P
£ ) 3 \
O o g P
o o T o
R . L a; :
9 S 2
z o . g i
6 70 — Ly -
& S o 70
Tk o O
< ' 5
L 1»—-‘_-/ % '
Q
G
@
st ... 0 1 2 3 4 5 6 i
i e oy s . POWER DENSITY (kw/cm3) '
{ “ISUAND RADIUS (in) : o , i
- - Fig. 1.3, Total Weight vs Island Radius for Fig. 11.4. Total Weight vs Power Density for
e ~ 300-Mw  Circulating-Fuel  Reflector-Moderated =~ 300-Mw  Circulating-Fuel = Reflector-Moderated
Reactor. Reactor.
195

 

 

 
ANP PROJECT PROGRESS REPORT

.
75 2-04-059-25

 

2-01-059-26

)

 

80

| _ T T

=3

- REACTOR, REACTOR SHIELD; AND CREW COMPARTMENT WEIGHT (b x 10

 

 

 

/

i pda &
~
w

65 Z

" POWER DENSITY = 4.125 kw/cm®
REFLECTOR THICKNESS =12 in.
ISLAND RADIUS = 4 In.

 

 

 

 

 

 

 

 

 

 

 

 

o
o

60

 

3 a4 5 6 7 8 9

 

 

 

POWER DENSITY =4 .125 kw/cm3
HEAT EXCHANGER THICKNESS =6.5in.
ISLAND RADIUS = 4.01n.

REACTOR, REACTOR SHIELD, AND CREW COMPARTMENT WEIGHT (ib x ‘TO_'3)
: ~
o

HEAT EXCHANGER THICKNESS (in)

 Fig. 115. Total Weight vs Heat Exchanger
Thickness for 300-Mw Circulating-Fuel Reflector-

Moderated Reactor. 60

 

 

 

 

 

6 8 10 2 14 16 18
REFLECTOR THICKNESS (in.)

 

Fig. 11.6. Total Weight vs Reflector Thickness
for 300-Mw Circulating-Fuel Reflector-Moderated

Reactor.

 

196

 

 
e e

.

 

%

*

PERIOD ENDING SEPTEMBER 10, 1955

12, LID TANK SHIELDI'NG«F3AC1LITY

G. T. Chapman

J. M, Miller

~ Applied Nuclear Physics Division

W._ J. McCool

H. C. Woodsum

Pratt & Whitney Aircraft

- The static source tests of the second series of

the circulating-fuel reflector-moderated-reactor and
shield (RMR-shield) mockup experiments at the Lid

Tank Shielding Facility (LTSF) have been com-

pleted and the data are being analyzed. Tests with
the dynamic source are in progress,
REFLECTOR-MODERATED-REACTOR
AND SH!ELD MOCK UP TESTS

Fur'fher variations were made i in the mockups of

| fhe RMR- shleld for 'rhe final measurements in the

stm‘lc source tests, ! Gommq-roy and neutron

measurements: were made in the water beyond the

mockups, and sodium activation measurements

‘were made within the heat exchanger regions of
the mockups.

Gummu- Ruy and Neutron Measmements
- Beyond the Mockups

Neutron and gamma-ray measurements were made
in water beyond the RMR-shield mockups to de-
termine the effect of placing an intermediate or
high atomic weight material immediately behind
the beryllium reflector, varying the thickness of
the reflector, and d:stribufmg the lead gommo-roy
shield in borated water,

The effecf_of placing an lnfermedmfe or hlgh
ihdterml |mmed:a1rely behind the
_'freflec’ror fegion was studied by’ cdohng a 3-|n.-fh|ckh
~ slab of blsmi_q“h

 

 

 

 

 

'IThe first measuyrements were reporfed by G. T.
Chapman, J. B. Dee, and H. C, Woodsum, ANP Quar.
Prog. Rep. June 10, 1955, ORNL-1896, p 194,

 

o u nick slab of ‘copper be- S
- hind c:n_fl8-m2lfh|c|< slab e}f berylhum.' An onalysrs{" R

 
 
 
 
 
 
  
 
 

 

   
 
 
 

%,\but unforfunate y‘, there was'_’f“;-;
' used to allow apprec:oble sel}f-;_v

_obsorphon of capfure gamma rays; fhe fcst neutron

dose rate was the same as that for an equivalent
thickness of beryllium, Although the 2-in.-thick
layer of copper would not be sufficient to effect
an appreciable weight saving, there might be
enough self-absorption of capture gamma rays in a
4-in.-thick layer to show a valuable weight saving.

The effect on the total gomma-ray dose rate of
varying the beryllium thickness (8, 12, and 16 in.)
was not appreciable. There was greater attenuc-
tion of the source gamma rays by the larger berylli-
um thicknesses, but more capture gamma rays
resulted from the beryllium, Fast-neutron dose
rates were not very different behind the various
thicknesses.

A study of the effects of distributing the lead
gamma-ray shield in borated water was also made
in order to obtain the information needed for an
optimization of the placement of the lead, This
study showed that there would probably be no
weight saving as a result of distributing the lead
rather than placing it in one piece, for lead thick-
nesses up to 5 in., but that there might be an ap-
preciable weight saving as a result of distributing
the lead beyond the first 5-in. layer. The secondary
gamma-ray dose rate produced in the lead and
borated-water shield fell off at the same.rate as
the thermal-neutron flux and thus was apparently

" caused by thermal-neutron captures in the shield.

-+ Sodium Activation in the Heat
Exchanger Region of the Mockups

Sodtum activation tests were performed to de-

ll termlne the achvcmon of the coolant in the heat
exchanger region as a function of the heat ex-
" ) chcmger thickness, the boron curtain thickness and

d:sfrlbuhon, and the reflector thackness. The ef-

| fect_on the achvahon of p!acmg a copper gamma-

ray' shield ‘immediately behmd the beryllium
reflector was also studied,

The sofurated specnflc ochvmes in the heat
exchcmger regions were measured in the manner
described previously! for a total of 17 mockups,
5 of which are shown in Fig. 12.1. If only two
heat exchanger tanks were used in a test, a

197

 

 

 
 

 

 

ANP PROJECT PROGRESS REPORT

2-01-057-66-160

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 ~
\l
™
\\
5 \=
3N\,
\ll
> N,
—_ &
% e e
. 'g 4 \a \
g0 e N
& ™ N2
= oA A
2 \\
-
x5 N
3 |_® REPRESENTS MEASUREMENT _ &\?
5 APPROXIMATELY IN CENTER
s —  OF HEAT EXCHANGER TANK —--=8=m
i _ . .
; |
&2
&
5 \°
K 3
% 10 [T_ CONFIGURATION MOCKUP SYMBOLS ]
(] - .
— 2~in.-THICK Nof TANK
— ¥ -in.-THICK BORAL
S SHEET s
2-in.- THICK COPPER |
SLAB
.
% MOCKUP 5 HAD A {2-in.- THICK BERYLLIUM REFLECTOR ;
2 MOCKUPS 1 THROUGH 4 HAD 8-in.- THICK BERYLLIUM
REFLECTORS
sl
25 30 35 40 45 50 55

2, DISTANCE FROM SQURCE {cm)

Fig. 12.1. Sodium Activation in the Heat Ex-
changer Region of the RMR-Shield Mockups as

a Function of Distance from the Source.

straight line was drawn through the two points

because of the lack of more information. Addi-

tional information probably would show these
curves to be slightly concave downward, as in the

198

cases where three heat exchanger tanks were used,
The five tests shown in Fig. 12,1 are representa-
tive of the results obtained to date. A more com-
plete analysis of all tests is being prepared.

If test 2 is used as a reference, comparison of
the activations in the different tests shows the
effect of adding or removing different components
of the configuration, In tests 1 through 4, 8 in, of
beryllium was used to simulate the reflector,
while in test 5, 12 in. of beryllium was used.

Comparing test 1 (4-in.-thick exchanger) with
test 2 (6-in.-thick heat exchanger) shows that
there is greater relative activation in test 2; this
results from the reduced resonance-neutron escape
probability in the thicker heat exchanger. The
escape probability may be increased simply by

distributing the boron curtain through the heat

exchanger region, [n test 4, a distributed curtain
containing less curtain material reduced the total
activation to about 25% below that found in test 2.

An increase in the reflector thickness from 8 in.
(test 2) to 12 in. (test 5) effected a decrease in
the sodium activation of about 80%. This was due
to two effects: an inverse-square spreading and a
reduction by the extra 4 in. of beryllium of the
number of neutrons at and slightly above the sodium
resonance energy of 3 kev, »

fn test 3, o gamma-ray shield of copper was
placed between the beryllium reflector and the first
boron curtain. The sodium activation was increased
by a factor of roughly 6, in comparison with the
activation found in test 1 with no copper present
in the reflector, The copper apparently changes
the fast-neutron spectrum so that there are relatively
more neutrons at or just above the 3-kev resonance
level of sodium. The boron curtain is only gray to
neutrons of this energy, and, hence, some of these
neutrons diffuse through the boron curtain into the
heat exchanger, where they are slowed down to the
3-kev resonance level by the fluorine.

 

 

T T

 

 
 

o ’range of decay hmes.

 

CTL V. Blosser
G. M. Estabrook
J.D.

P
.M, Henry

A 7. Fufl'erer

PERIOD ENDING SEPTEMBER 10, 1955

e W 'BULK SHIELDING FAGILITY
S s R, Ma:enschem -

E. B. Johnson

T. A, Love

F. J. Muckenthaler
R. W. Peelle

W. Zobel

Apphed Nuclear Phys:cs Division

K. M. Johnson

Prafl' & Whn‘ney Anrcraft

The angular—dlstr:buhon measure=ments of fdst

neufrons have been extended to include measure-

. “‘ments’in a plain-water medium. The resulhng data
" have been correlafed with the TSF shield Op‘l‘!ml-
o "zahon studies, The first portion of the expeti-

*“menfs designed for defermmmg the ‘spectrum of

7,f_”flsmon-producf gamma rays was comple'red for one

' "li,’ANGULAR DISTRIBUTION OF FA T NEUTRONS

s . 'THROUGH WATER

) T V Blosser, Appllecl Nuclear Phys&cs Division
M E Valermo, NACA Cleveland

: A seties of shleld ophmlza'non sfud:es is bemg
~‘made at the TSF, and for such o rational deter-
~ mination of the neutron shielding lhlcknesses re-

qu;red at the reactor and at the cre'w compartment |

- ofa nuclear—powered aircraft, it is necessary to
; know the angular distribution of the fast-neutron

[ dose at fhe reactor shield surface. Some prellml- _

ndry méasirements have therefore been ‘made at

r {f" i’rhe Bulk §_h|e|d|ng Fac:llfy (BSF) to obtaln anh

   
   

T f{jrmdlcaho/n of the angular dlsfrlbuhon of the fast:

"background’ from neutrons orlg:nahng in the

 

line. Some uncertainties exist in the interpretation
¥ _ ~of the measurements because of the unknown

water along ‘rhe sides of the collimator and scat-

teted by the collimator into the Homycdk button,
This background is expected to result in a broader
angular distribution than actually exists,

The arrangement, in the pool, of the air collima-
tor, the fast-neutron detector, and the lead detec
tor housing, relative to the reactor, is shown in
Fig. 13.1. The active lattice of the reactor is a
5 by 6 fuel element array with two fuel elements
missing from the back (south) corners of the
lattice. The air collimator is a hollow cylindrical
aluminum tube 168 em long, 5.08 e¢m ID, with a
0.16-cm wall, The position of the end of the colli«
mator toward the reactor is given by coordinates
(z,x), and the collimator angle, with respect to
the reactor center line, is designated by the angle
a, as shown in Fig. 13.1.- Measurements were
taken only in a horizontal plane at the mid-plane
of the reactor.

The angular distributions were measured at the
following locations in the pool water shield:
= 70 crri, x = 0;

70 cm, x ).

“,7‘ N
]

Thefwoangles, d] and a2, at which a pro;ec- ‘
tion of the axis of the collimator just intersects
the two edges of the north face of the reactor are

199

 

I

 
o i S

 

i s . g . . o e s

 

‘ EY -,

  

   
 

v b e e e

ANP PROJECT PROGRESS REPORT

ORNL-LR-DWG 8553

REACTOR .
(LOADING NO. 33)

 

~=-19,6 cm =

 

 

 

a, s,

     
  

7.6 cm OF LEAD

 

 

5.08-cm |D

HORNYAK SBUTTON

H, PHOTOMULTIPLIER

DETECTOR HOUSING
AND SHIELD

Fig. 13.1. Experimental Arrangement for Meas-
urements of the Angular Distribution of the Fast-

Neutron Dose in the Water Shield of the BSF

Reactor.

indicated i:ri Fi'g.' 13 2. For the rdn'gétdf the qngles
lli[between OL] ‘and’ a,, the collimator sees some por-
- tion of the active lattice. It may be seen in Fig.
_ "3’13 2 that in thls range of angles the variation in
- .the dose rate per unit solid angle is small, Be-

“of the relatlvely poor statistics in these

o r;'mecsurémenfs, the exact nature of the variation is
- “not defined. For angles outside the range o',1 and

dq, the pr0|ected axis of the collimator does not

infersect any portion of the active lattice. In this
. .region, the measured dose rate is principally from
o ii_neutrons which have made at least a single colli- -
~ sion in fhe water sh[eld and therefore, as the col-

limator swings away fi'om the reactor face, the

‘dose rate per unit solid angle drops rapidly.

200

ORNL-LR-DWG 8954

WIDTH OF REACTOR
. 39.2cm

Q, Q,

CORRECTED FOR BACKGROUND

BACKGROUND

DOSE {mrep/hr-w- steradian)

o!
n

 

-80 -60 -4Q ~20 c 20 40
: Q (deq)

Fig. 13.2. Angular Distribution of F'ds"t‘r Neutrons
(Z=70cm,x=0), ) : e B o

The results obtained for z =70 em, x = +19.6 cm

‘are presented in Fig. 13.3, In this case, the pro-
‘jected axis of the collimator at @ = 0 just inter-
" “sects the northwest comer of the active lattice.

Agam only a relatively small variation in dose

rate per unit solid angle is indicated over the

range of angles where the collimator sees the
active lattice, and there is a relatively fast varia-
tion as the collimator swings away from the active
lattice. < S ,

The results for 2 = 5 cm, x = 0 are given in
Fig. 13.4. Over the entire range of angles covered
in these measurements, the collimator saw some
portion of the active lattice, The variation over

the entire range of angles is relahvely small and™

is consistent with the results shown in F:gs. 13.2
and 13.3. However, the error m:ght be very icrge
for this 5.cm measurement because of the lack of
complete collimation. -

 

 

 

 

 

 
PERIOD ENDING SEPTEMBER 10, 1955

' SR
ORNL-LR-DWG 8955 105 ORNL-LR-DWG 8956

 

 

WIDTH OF
1 REACTOR
T 39 2c¢m
{. Q,
15
¥ ORRECTED FOR BACKGROUN

BACKGROUND- BACKGROUND NEGLIGIBLE

" DOSE: (mrep /hr-w- steradian), c
OOSE (mrep/hr-w- steradian)

 

-80 -60 -40 -20 -10 20 40
Q {(deg)

 

 

_ Flg. 33 3 Angular Dlstrlbuhon of Fosf Neutrons
’ (= 70 cm x'- +'|9 6 cm). : Fig. 13.4. Angular Cistribution of Fast Neufrons
L D -~ (z=5¢em, x=0).

    

An mtegrahon was made ‘of the dose rafe per

it solid 'r bta fhetfad erat di- L . .
um so : ongle o o "n o0 I 22 .,.e m : gross fission products for short times after fission

‘was discussed previously.] A large share of the
gamma-ray dose in a CIrculahng-fuel reactor Ol'lgl-

   
   
    
   
 
   

 

 

£«

'7,-.,«1 T o T

ar ioadlng by

 

3 >

  

 
    
   

iminary résu]ts are presefited her

“The ugémma-roy decay rate'qs a fuhchon of hme_'

 

E. C Cqmpbe“ Ph.);sws DIVISI;)H '

The |mportonce of information about the decay ]R W Peel[e ot al., ANP Quar. ng Rep June 10,
) chqrqcfe_rlshcs and photon energy spectrum of the 1955, ORNL-1896, p 203.

-

 

 

  

 
Lok

 

 

 

 

 

 

 

ANP PROJECT PROGRESS REPORT

ma-rdy Qbeci'romefer,2 'rhe qssocmted e!ecfronlc' |
instrumentation 'described previously,3 especially
: constructed timing devices, and the fcst pneumatic

probe assembly designed by E. C. Cumpbell
Small, enriched-uranium samples that weighed

7 about 7 mg were irradiated in the ORNL Grdphn‘ef
" Reactor for times rangmg from 1 to 8 sec. From
10 to 50 samples were required to obfain a single

decay curve, Each sample was withdrawn pneu-

- matically in about 0.4 sec and stopped, automati-
, 'colly, in front of a 10-in. lead collimator that led
.- to the multiple<crystal spectrometer. For this
-~ application the lead spectrometer shield was sup-
" plemented by a T-ft-thick layer of LiF and parcffm
7 in order to reduce the neutron-induced gamma-ray
o bcckground ~ This shield makes the background |

' counting rate enhrely negllglble.
 The electronic instruments cssoc:qted with ’rhe
= mulhple-crystal specfrometer were arranged so
o fhat the counhng rote assocmted with a predeter-
' mined photon energy interval was available at any

given time after fission. A time analyzer recorded
this counting rate as a function of time in 40 time

" channels. These time channels were larger for

the longer times after fission so that the number of
counts received in the various time channels
would not vary too widely. In the phase of the
experiment discussed here, the first time channel
was about 1.6 sec wide, and the last was about
10 sec wide,

Photon energy calibrations were made by using
Hg2093, Na22, Zn65, F20, 7,90m ThC'', and N16
sources, Calibrations of the time analyzer were
obtained by using a 60-cycle pulser.

The efficiency of the spectrometer as a function
of energy must be known in order to interpret the
data in terms of photons per second leaving the

vranium sample, In the energy region up to 1.3

Mev, the efflmency was determined by using the
ubsolu)‘e source strength of Hg203, Na?2, Cs137,
Nb95, and Zn6> calibration sources. These abso-
lute source strengths were obtained with the cali-

brated high-pressure ion chamber of the ORNL

Radioisotopes Control Laboratory.

 

2F C. Maienschein, Multz le-Crystal Gamma-Ray
Spectrometer, ORNL-'|'|42 (Aprn 14, 1952).

3T, A. Love, R. W. Peelle, and F. C. Malenschem
Electronic Instrumentation for a Multiple-Crystal Gam-
 ‘ma-Ray Scintillation Spectrometer, ORNL-1929 (to be
<7~ published).

 

 

T%’"”&nclyms of fhe 'aotq- was '-s'frdi"gl'if'foi'wo\rc-i :

~except f for the problems of absolufe efflmency of

fissions occurring in the scmples. For this pre-
llmmary analysis, large uncerfomhes were per-'

"“mitted in these two factors.

Al’rhough the efficiencies at eherg'ies up to

1.3 Mev were measured for fhls unolysrs, the

efficiencies at higher energles were found largely
by comparison with previously ‘obtained perform«

“ance data, [t is therefore to be dssumed' that a

further analysis of these data would gtve a shghfly_ '

“different result. A further dlfficul’ry arises from

the wide vonoflon of fhe Specfrometer effic:ency'

VW|’rhm a single energy group. The average efi‘l-_ |

ciency was used here, but the spectrum shape
within a given energy group should be used to

_ :obtam a welghted average effncnency. 7
izati n_ of the results de-

 

The absolute normcl' ,
pends upon the number of fissions in the samples,
which, in turn, depends upon several doubtfu!
quantities. For the l-sec bombardments, the time
spent by the samples in the reactor is in doubt by
at least 10%. The thermal flux and the macro-
scopic cross sections of the samples are also in
doubt. The effective thermal flux available to the
sample was measured by using gold foils and a
cadmium-difference technique.*  Thermal cross
sections were used for both the gold and U235,

The basic results of the dqta analysis are shown
in Fig. 13.5. Six energy groups were studied, but
the region between 1.6 and 2.3 Mev was studied
with both the Compton (two-crystal) and pair
(three-crystal) spectrometers. [t was expected
that these curves might not agree in a preliminary
analysis, because it is in the region studied that
the efficiencies of either type of spectrometer vary
most rapidly with photon energy. The times after

fission were measured from the center of the lrra-, -

diation time interval. In the pair spectrometer

runs, it was necessary fo use an irradiation time
of 8 sec, and therefore sa'rurahon effecfs pre|ud|ce

the first few points on these curves.

Two cross plots of Flg. 13.5 that give the': S
photon energy specfrum as measured 20 and 150

sec after fission are presented in Fig, 13.6. The
results obtained by integrating the curves of
Fig. 13.5 to obtain the total energy release per -

 

4John Moteff pnvote communlcahon

 

 

 

 

 

  
B i Moo

 

 

!
|
i
|

[
Wy

 

"

e

 

PERIOD ENDING SEPTEMBER 10, 1955

fission between 10 and 150 sec are given in (Mev) per fission in this interval must, for the
Table 13.1. Because of the uncertainties men-  present, carry an estimated probable error of about
tioned, the total photons per fission and the energy +25%.

UNCLASSIFIED
ORNL—LR—DWG 8957

-1

   

  

  
   
   
   
 
   
   
 
 
   
 

10
5
2
® COMPTON SPECTROMETER
A PAIR SPECTROMETER
—2
10
5
-
. Q
o
2
5 2 0.28 TO 0.5% Mev
v
Z 054 TO 112 Mev
QO
2 o~
g 1.12 70 1.62 Mev
3
=
o
b S 162 TO 2.3 Mev
x 1.62 TO 2.3 Mev
%
O
w
o
2
2.3 7O 3.5 Mev
1074

3.5 TO 5.0 Mev

  
  
 
 

20 \ 220
S CTIME AFTER FISSION 1sec) 77 7 e oo

 

   
 

 

n of Time After Fission for Six

 

 

203

 

 

 

 
ANP PROJECT PROGRESS REPORT

UNCLASSIFIED
“ ORNL-LR-DWG 8958 T e

 

- COMPTON SPECTROMETER
—m— =~ = PAJR SPECTROMETER

 

b Akl

 

DECAY RATE ( pho’rons/Mev . sec fission}

  

0 10 20 30 40 50 6.0
PHOTON ENERGY (Mev) '

Fig. 13.6. Histogram of the Fission-Product Photon Energy Spectrum for 20 and 150 sec After Fission.

TABLE 13.1. MEASURED VALUES OF PHOTON INTENSITY PER FISSION AND TOTAL ENERGY
RELEASE PER FISSION INTEGRATED BETWEEN 10 AND 150 sec AFTER FISSION

 

 

Energy Range Photons per Energy per Energy Range Photons per Energy per

 

 

 

(Mev) - Fission Fission (Mev) (Mev) Fission Fission (Mev)
. o Cc;m;;fon :Spéé.fromefer - Pair Spectrometer
- 0.28t0 6.51- | | | 0.25 0.10 1.62 to 2.3 0.13 0.26
0-.'51716 1:12 | 0.45 037 2.3 to 3.5 0.069 0.20
- 112 f0 1.62 B | 0.22 0.31 3.5 to 5.0 0.038 0.16
162t 23 | 0.19 10.38 Total 0,28 to 5.0 1.19 1.46
o L AvOs A 032

 

     

 
i

-

1

-
e

 

oo

 

   
   

 

PERIOD ENDING SEPTEMBER 10, 1955

g 14, TOWER SHIELDING FACILITY
C. E. Clifford

F. L. Keller

F. N. Watson

Applied Nuclear Physics Division

J. E. Van Hoomissen, Boeing Airplane Co,
M. F. Valerino, NACA, Cleveland

In a previous report! an analysis of the fast-
neutron measurements of the differential shielding
experiment at the TSF was presented. In this
analysis the probability of the air scattering of
neutrons into the various sides of the crew-
compartment shield was obtained as a function
of the angle 8 which a fast-neutron beam from the
reactor shield surface makes with respect to the
axis joining the reactor and the crew compartment,

It also gave the probability of neutron penetration
" through the shield into the crew compartment as a

function of §. These probabilities have now been
used in the development of a procedure for opti-

" mizing the neutron shielding of a dwnded shield

for an aircraft,

'A further mveshgahon of the GE-ANP reactor
and crew shield mockups was also made. Measure-
ments of gamma-ray doses inside the crew-
compartment mockup were completed, and, at
present, a study of the distributions of gamma-ray
intensities around the reactor shield is bemg made.

PROCEDURE FOR USING TSF DATA FOR
THE OPTIMIZATION OF A DIVIDED
NEUTRON SHIELD

'M F \Valerlno ' o

   
 

" would have the minimum welght that w wcs cons:sfenf:';f:' |
“with_ all the desrgn parameters._ As a first step in
_‘fhe des:gn of a completely ophmlzed ‘neutron shield,
ocedure hus been developed to de?ermme, for:f'f‘

" onical s shel!s, as shown in Flg. 14. ]W'L“The ve”ex.-:.ug:;-,,;.;..: S
i‘z;;'-\':of each cone IS token to be clf the center of the Y o

 

 

M. F. Valerino, ANP Quar. Prog, Rep. June 10,
1955 ORNL-1896, p 206.

reactor, The radius ¢ in Fig. 14.1 is the radius of
the reactor, heat exchanger, gamma shield, etc.,
and thus the outer radius is the 'sum of radius a
and the neutron shield thickness T ., The crew
shield is assumed to have a rear thlckness T and
a side thickness T_, For simplicity, and because
of the small sacrifice in shield weight involved,
the front thickness T{ is taken to be equal to the
side thickness T _,

The procedure consists, first, in expressing both
the total weight of the neutron shield and the
dose rate at the center of the crew compartment
as functions of T , T, and T ; second, of using
the method of Lagrange multipliers to obtain the
equations which T , T, and T_ must satisfy in
order that the weight be a minimum for a specified
total dose rate; and third, of developing an iterative
procedure for the solution of these equations.

Calculation of Shield Weight

The weight, W_, of the neutron shielding ma-
terial included in the nth conical shell at the
reactor is given by

-

 

 

        

. 2mpg
nV - (COS Gn"-—- cos Gnl'l) [(Tn + CZ)3 - a3]
A(6,)
iy 3 ~ a3 ,

e where pp is the density of the shielding material

to at the reactor; the angles 6, -, 6+, and g, for the

nth conlcql shell are |dent|f|ed in Fig. 14.1. The

,fofol welght of shleldmg ‘material at the reactor is
thus

n=1,2e00,N.

The model assumed for the cylindrically shaped
crew compartment is also shown in Fig., 14.1. The

205

T T T Y IR e

 

 

 
 

 

 

 

ANP PROJECT PROGRESS REPORT

   

REACTOR SHIELD

2-01—056—-7-D~165

CREW COMPARTMENT SHIELD

L

 

oy

 

 

 

 

 

 

 

 

 

 

 

 

THREE-DIMENSIONAL VIEW OF
AN CONICAL SHELL

S
I~

o

o

 

 

Fig. 14.1. Divided Neutron Shield Model Used for Optimization Calculation. |

weight, 'WC_S', of such a crew shield is given by

TP

 

@) W, = {Df_(Tr +T,)

* [(Dc + 2Ts)2 - Dg] Lc} !

where P.s is the density of the shielding material
used at the crew compartment; D_ and L_ are
[defi'ned_ in Fig. 14.1.  The total weight of the
neutron shielding material is thus

(4) o Wiotal = Wp + Weg o

206

Calculation of Dose Rate

Let D j(a,4) be the direct-beam dose rate a‘r":'a
distance { ft from the unshielded (T, = 0) sphere

of radius @, The dose rate at a distance ffi_fl‘fo[‘q o
uniform shielding thickness T, is then given by

..fT" (1/A)dt
(5) Dad)e ~° '

where A, the direct-beam relaxation length, is «
function of t (dt is the increment of T ), The rate
at which the dose is scattered to the side of the

 

T T TR (R O VRTIO

 

 
AN, SR A . e

wr

 

"

o,

 

€ wflbox, cnd {5 is the ‘‘focusing factor’ for the

crew compartment from a conical shell is given by!

. 8.’
(6) fc“’ " p’ps., (6) d{cos 6) ,

058 »e
n

where D7 is the dose rate at the surface of a
hypothetical unit sphere at the reactor and P, . (6)
is the probability of scattering into the crew com-
partment side from a conical-shell beam, as de-

termined from the TSF differential experiments,

In the present case

T
..f 7 /A )dt
(7)-, D5 = 42 Jalye ~F .

Thus, if PS.. (8) is considered to be constant
over a given conical shell, the integration gives
the rate at which the dose is scattered to the side
of the crew compartment from the nth conical
sBeH as

(8) (cos 0,- — cos §,+) Dd(a,fl) X

T
—f i (I/Ad)dt
x € ° /E })zlt:le(6 ) .

The contribution to the dose rate at the center of
the crew compartment by neutrons which leave the
nth conical shell and scatter into the side of the
crew compartment is given by

(9 D:'Side = (cos 0, — cos 0,,+) x
T
_f " (1/A)de
X Dd(a,fi)e ° X

L S
)fi?:,de(e )fS P o
whlch is a funchon of 9

k% (dx s fhe increment of T ), is the relaxcmon
- length for the sccftered ‘dose at the side of the

AV::""wher_e )ls

  

   

112

e- scattered radiation, The “‘focusing factor

décounts for ’rhe radmhon Focusmg in the center of
the crew compartmenf cnd depend: on the angularv

 
 

 

; "‘2_J. E. Fculkner, Foc:usz'ng of Radiation in a Cylin-
f'irz'ca;l Crew Compartment, ORNL CF-54-8-100 (Aug. 18,
954).

Tn, and

PERIOD ENDING SEPTEMBER 10, 1955

distribution of the radiation, at the inside surface
of the crew-compartment side shield.

An expression similar to Eq. 9 holds for the
contribution to the dose rate at the center of
the crew compartment by neutrons which leave the
nth conical shefl at the reactor and scatter into
the rear of the crew compartment. This expression

is

(]0) Dzrreur = (COS Qn; — COS 672”) %

T
_f n (I/Ad)dz

)(l Dd(a,'fl) e 0 X
T
..f T (/A)dy
0
x AP}, (6,) 1. e :

where A7, which is a function of §,, T , and y (dy
is the increment of T ), is the relaxation length for
the scattered dose at the rear of the crew com-
partment, The factor /7 accounts for a decrease
in the dose rate at the center of the crew compart-
ment caused by the angular distribution of the
neutrons leaving the inside surface of the shield
at the rear of the crew compartment.

Finally, the dose entering as a direct beam from
the reactor through the rear of the crew compart-
ment shield must be considered. Let the angular
distribution of dose from an element of area at
the surface of the reactor shield be of the form
cos™ i, where i denotes the angle between the
normal to the surface element and a given direction.
If it is assumed that the separation distance be-
tween the reactor and crew shield is large com-
pared with the outer radius of the reactor shield
and that the nth conical shell is in the hemisphere
toward the crew compartment, it may be shown
that the contribution of the nth shell to the direct-
beam dose arriving at the rear of the crew com-
partment is given by

(”) (éosmfl 0, - cosm” Bn”) X

.
._f n (1/A y)dt
x Dyad)e O .

If the shell is in the hemisphere away from the

~crew compartment, the contribution of the shell to

the direct-beam dose is zero. If the nth conical
shell is in the hemisphere toward the reactor, the
rate at which the direct-beam dose leaves this

207

 
 

ANP PROJECT PROGRESS REPORT

shell and arrives at the center of the crew compartment is given by |

B ;f (1/A pdt _f (1/)\ ydy

(12 Di'rear =. (cosm-fl Gn' - cos m+'l 9 ) d(a /fl) e f"d e ~ 0 ‘

where A7, the relaxation length for ’rhe darect beam

dose at the rear of the crew compartment, is a

 

 

T, TS, cnd C (L.ag.rcn;tg'e ;nultiplier):

aw aD

 

 

 

 

 

Calculahon of Mlmmum Shleld Welghf
for Spec:hed Dose

- The th:cknessesT T , and T are considered as
parameters whose vqlues are fo be determined so
thatW, .. isa minimum when D ., is a specified
dose rate, By the method of Lagrange multipliers,
this amounts to solving the following (N + 3) simul-
taneous equations for the N values of T, and for

208

function of T,2 and y. The factor f7; accounts for total 1 total -
the decrease in dose rate of the direct beam at the 3T + T o =0,»n=1,2,...,N,
center of the crew compartment caused by the n n
angular distribution of the direct-beam neutrons ow oD
leaving the inside surface of the rear of the crew ..__l?_f.i + _]_ __total =0,
compartment. | the nth conical shell is in the (14) dT, C dT,
hemisphere away from the crew compartment, then . D .
D‘iv'e‘" = 0. The rate at which the total dose total +_l total 0
arrives at the center of the crew compartment is aT', o aT ’
therefore
D,mm, C (specified constant) ,
(]3) Dfoml For the calculation it is convenient to attach a
subscript to the { involved in each equation, but
N it must be remembered that, for an exact solution,
z z (DSeside pS.rear Dd,reur) all the s must be the same. When this is done
-1 " ” " and the differentiation indicated in Eq. 14 is
carried out, the following (N + 2) equations for the
£'s are obtained:
: fota i/aT Di,Side + D;s‘l,reuf + D;zz’,rear ‘
(15) L = = , n=1,2,.00, N,
W 10t/ 9T, AA00) (T, + a)?
(16) € Droral T '3 0
- ’
° anofa I/aTs fotc: I/aT n=|\ )\f}
. | aDtotal/aTr 1 N Dz,rear Di,rear
I =l ) . .
' awtpto I/aTr tota l/aT n=1 )t; )t;

The complexfiy of the sysfem of equohons de-

rived above requires an iterative mode of solution,

and a procedure was desired that converged rapidly

enough for the solution to be accomplished in a

reasonable length of time by hand calculation,
A satisfactory procedure was devised and is de-
scribed below in a detailed, stepwise fashion.

Step a. Equations 16 and 17, with the condition |

fir =L =L, are combined to relate £ to the speci-

 

 

 

 

 

 

 
[

o M S\ b el aans

N

 

fied dose rate, D__. . (given by Eq. 13), and to
appropriately averaged relaxation lengths. In-
cluded in this relationship are the shield-weight
partial derivatives with respect to T, and T_,
which are known functions for a given crew-
compartment size, These steps are shown below
in equation form.
If the following definitions are made,

N »
E Dz's‘de//\i

 

 

1 “
_ = N '
AS ,s id
s z Dz side
' n=1
(18)
N
S, ,rear /) 7
-I E Dn /)\s
n=1
= ’
N N
A7 ,
s z:l)irecr
n=}

then Eqs. 16 and 17 may be combined to give

 

PERIOD ENDING SEPTEMBER 10, 1955

Equation 21 gives L in terms of quantities which
are either specified or determined from the previous
iteration, For the first iteration, fairly accurate
estimates can be made of the values of the quan-
tities in Eq. 21. Since D?‘f"“' << D, .oy and
AL/AG ~ 1, X, will be approximately equal to 1.
Also, GWfoml/aTr is a constant and oW . |/8TS
varies only slowly with T_ for a reasonable crew-
compartment size. The relaxation lengths AY and
Al can be chosen to be equal to some average
values from the TSF data to within about +25%
accuracy. Hence, in the first iteration, a reason-
ably good estimate of { is obtained.

Step b. Equation 17 divided by Eq. 16 and com-
bined with the condition Er = ES gives a relation-
ship between T, and T_ that involves appropri-
ately averaged relaxation lengths and three basic
parameters. One parameter involves the relative
rate of shield weight change with respect fo the
rear of the side shield thicknesses at the crew
compartment. The second parameter is the ratio
of the total air-scattered dose into the rear to that
into the side of the crew compartment. The third
parameter is closely related to the ratio of the
total direct-beam dose at the rear to the total air-
scattered dose at the side of the crew compartment.

In detail, the relationship discussed above is
obtained in the following manner. If Eq. 17 is
divided by Eq. 16 and the definitions for AY and A
given by Eq. 18 are used, there results

N .
E [sz,reur + (A;‘/I\dr)Di,rear]
n=] )

 

S anofal
(19) AS ——— [
o s
T awtotul o
+ A'.S aTr r = Dtotul Xd '
(22) _ 'Cr 1 awtotol/aTs Aj‘-
S L Mo/ T,

 

   

 

e D el Xd
(21) £ = — e’ 2 )
Aj (aWtofal/aTs) + )\: (aWtotcl/aTr)

 

N
z Ds,side
n

Lon=1 -

(cos 6+ — cos §,+) t2ps.  (6,) 1,

S

(cos §,+ — cos @, ) 42ps (6,) fi

rear

= (cosmfl en, - Co$m+l.6n,,) fdr ,

aWtofal/aTs F

5

aWtotcxl/aTr ;\—

r
5

 

209

 
 

 

 

 

 

"ANP PROJECT PROGRESS REPORT

!, and )\.f by

ond defme new quonhhes AS, A
f (V)ld)dz -j; T (1/AS)dx
Z B, e e ‘
n=1
: R '
N 7_f " (1/A,)dt
L 5,e
n
n=1 _

- —T //\S

rT | T
_j; ”(l/)td)dt -_]; ’(I/A;)dy

 

 

_ N
. o = E Cne e
! . S T - / r =
(24) . T, )\s _ n=1 '
- o e - ‘N ..f " (/A )de
- : 0 d
, E ,Cne
n=1]
o -f (I/Ad)dt ..f (1/A7)dy
' (J, /Al
‘-T/\' E] oA :

" (1/A )t

-

N _
L U,/Ap)e
n=1

Equcmons 9, 10, cnd 12 are substituted into Eq. 22 and then, by using the defmmons given in Eqs. 23

and 24, Eq. 22 can be rewritten as

 

.. [w ..fT” (17 )dt
'[ - ,{.3—.. E Cn € 0 /‘———r—
(25) — e-Ts/ S ik T/ s
. | IZVJ . e-foT” (1/7A)dt
n=] "

N - " (1/A )t

AL U, /ADe 0 fon
' -, )\;

n=1
e

T
N -f n (l/hd)dt
Y B,e 0
n=1

 

 

 

 

 

 
nfi‘ .

*

L

 

"

| (27)

Equation 25 relates T, and T_ in terms of the
various averaged relaxation lengths and the three
pardmeters whose physical significances are dis-
cussed above. The first parameter discussed above
is the quantity a in Eq. 25. The second parame-
ter is the left bracketed term at the right side
of Eq. 25, and the third parameter is the right
bracketed term. In a given iteration the parameters
and averaged relaxation lengths are evaluated on
the basis of the information obtained in the pre-
vious iteration. In the first iteration, AS is taken
to be the same constant as that used for )ts in
Step a, etc. Also, for the first iteration, “the
reactor shield is taken to be of an unspecified

' umform thickness (T = a constant for all »), in

whuch case the bracke’red terms in Eq. 25 reduce to

 

N __ N
- ‘}"ZI Cn A; 2: ]n
n=1 d n=1
a .
N " __ N
L B, r; LB,
n=] n=1

Hence, in a given iteration, a plot can be made of
T, vs T, which, upon convergence to the required
solution, is the relation which T and T_ must
satisfy in the vicinity of the solution to obtain
equal values of L. For convenience, this relation-

shlp is de5|gnated by

(26) T, = Q(T) .

between T_ and the various T,'s. [f use is made
of Egs. 9, 10, and 12 and of the definitions given
by - Eq.723 ‘then Eq. 15 can be rearranged as

'follows -

 
    

P 2 —_I; " (1/A)dt
d(“ ) e

 

oMt

PERIOD ENDING SEPTEMBER 10, 1955

In Eq. 27, Cn is set equal to L, as previously de-
termined in Step a, Eq. 21. By using the values
of A that are consistent with the results of the
previous iteration, the right side of Eq. 27 is cal-
culated for each n (n = 1, 2, ..., N) for various
assigned values of T_. From a plot of the left
side of Eq. 27 vs T, the values of T, correspond-
ingto each assigned value of T _ are then obtained.
Hence, plots can be made of T vs T _ for each n,

Step d. For various assumed values of T, values
of T  are obtained from the plots made in Step c.
These values are then inserted into Eq. 16, with
use being made of Eq. 9 and the definitions given
by Eg. 23 to compute the C for each T, A plot is
then made of .,C vs T_. The value of T, for
which the correSpondmg CS is equal to the value
of L determined in Step a, Eq. 21, is the required
T for. the given iteration. The corresponding
values of the T and T ,’s, for the given iteration,
are then obtained from the plots made in Steps b
and ¢c. By using the above-described procedure,
the various shield thicknesses T, T, and T, for
n=1,2, ..., Nare determined in a given iteration.
The entire procedure is repeated for the following
iteration, and the iferations are continued until
convergence is obtained,

Application of TSF Data

The following fundamental quantities must be
known as functions of the pertinent variables
involved in order to carry out the neutron shield
optimization calculations:

1. the fast-neutron-dose air-scattering probabili-

- ties, ©2PS. . (6) and 42P3,_ (6),

feay

Sfe'fi c. Relctlonshlpsare ‘also to be obtdi'néah 2. the reloxohon lengths A g A Al s and /\S

3. the focusing factor {$ and the geome'rrlcul atten-
vation factors and /%

 

 

T
5
_f (1 /)\i)dx

B e "0 + Cn

(T )
_f s (1/A])dy
e 0 + ], e 0

21

 
 

 

 

e

 

" ANP PROJECT PROGRESS REPORT

4, the angulur dlstrlbutlon of the fcst—neufron dose
at the surface of the reactor primary shield.
lnformcmon concermng these fundamental quanti-
ties was obtained in Phase | of the TSF differ-
enhcl shleldmg experlmenfs. However, there are
some gaps that require extension of the ex:s’rmg
data for the present calculations, In some cases,

__Ilml'red but appropriate, data were available to
"mdlcate a reasonqble means for extension. Also,

the extension was “guided by qualitative theoretical
considerations of ‘the attenuation processes in-
volved. For some cases, no reliable extension of

~ the data was p0551ble, and these limitations in the
'1i'use of ‘the data will be indicated. It should be

‘noted that the reqUIred data can be obtained by

, furfher experlmenfs and dnclyses cn‘ the TSF of

  

2-0t~056~ 7 ~A—165

 

0.02

iwete)

£2%p

0.005

0.002

0,001 —
0 30 60 90 120 150 180
8, CONICAL-SHELL BEAM ANGLE (deg)

Fig. 142, P;‘ob'abamy'of Fast-Neutron Dose

o Scaflermg from Comcal Shell Beam to Crew-
o ‘Compurfment Side as Obtained from TSF Experi-
ffl'_"“"?‘ (Tn =45 em, 1= 64 #).

the attenuation processes involved.
Plots of {2PS 4o(6) and £2ps

Coarl0) as functions
of 6 are given in Flgs. 14.2 and 14.3, respectively.

~ These plots apply for a reactor-to-crew-compartment

separation distance, 1, of 64 ft and for sea-level
altitude — the conditions of the experiments. At
other separation distances and alh’fudes, it is
necessary to apply corrections; corrections indi-
cated by single air-scattering theory will probably
be adequate for the present calculations, Rigor-
ously, the air-scattering probabilities in Figs, 14.2
and 14.3 apply only for fast-neutrons filtered
through 45 c¢m of water at the reactor shield, that

is, for T, = 45 em. However, limited data obtained

for T, =15 cm (for which the fast-neutron spectrum
is expected to be quite different from that for
T, = 45 cm) indicate that the effect of the spectrum

2-041-056 ~7T-A~467

0.5

Q.2

8)

rear

o4

P

Q.05

0.02

 

Q.01 ——
0 30 60 20 120 150 180

~ 8,CONICAL SHELL BEAM ANGLE (deg) )

 

Fig. 14.3. Probablhty of Fast Neutron Dose "
Scattering from Conical-Shell Beam to Rear of

Crew Compartment as Obtained from TSF Experl-
ments (T = 45 cm, 1= 64 f1).

 

 

 

 

 
 

e

s i £

e

*

i

W

A

v»x '

 

Ca i

i ‘

7.\, DIRECT- BEAM RELAXATION LENGTH. {cm)

shift is relatively small. Hence, the air-scattering
probabilities plotted in Figs. 14.2 and 14.3 were
taken, in the present calculations, to be inde-

pendent of the energy spectrum of the neutrons

emanating from the reactor shield (that is, inde-
pendent of T ). Also, values of *fiszea (6) were
measured only in the range of angles, 6 = 90 to
180 deg, wherein the direct- beum dose at the rear
of the simulated crew compartment was small com-
pared with the air-scattered dose. The values of
@Pfe (6) in Fig. 14.3 are extrapolated values in
the range of angles from 0 to 90 deg, the extrapola-
tion being guided by the variation of £2ps 6
measured for this range of angles.

“The TSF measurements show that the relaxation
length for the direct-beam fast-neutron dose is a
function of the total water thickness between the
reactor and the dose detector at the rear of the

crew compartment; that is, the relaxation length

is independenf of whether the water is at the
reactor or at the rear of the crew compartment,
A plot of the direct-beam relaxation length vs the
total water thickness is given in Fig. 14.4, and
from this plot the direct-beam relaxation length,
A, ot any point in the reactor water shield can be
obtained. In addition, this plot gives the direct-
beam relaxation length A} at the water shield at

2-C1-056—-7-A428

 

O TSFDATA |
'® BSF DATA CORRECTED FOR GEOMETRY

 

 

 

 

 

 

 

0 VN _4,50._
TOTAL WATER THICKNESS BETWEEN REACTOR FACE AND DETECTOR (Cm)

 

   

for Fast-Neutron Dose.

. Dose Scattered to the Crew-Compartment Side

  
 
 
 

.Fig.. 14.4. Direct-Beam Relaxation Length ()t)

PERIOD ENDING SEPTEMBER 10, 1955

the rear of the crew compartment for direct-beam
radiation emanating from the nth conical shell of the
reactor shield; the water thickness to be used in
the plot is the shell thickness, T, plus the dis-
tance in the water at the rear of fhe crew-compart-
ment shield at which the relaxation length, A7, is
desired. If plastic, instead of water, is used as the
crew-compartment shielding material, the values
of AJ taken from Fig. 14.4 should be divided by
1.14 (based on removal cross sections).

In the TSF experiments, the scattered-dose re-
laxation lengths, AS, at the side of the crew-com-
partment shield were obtained as functions of
water thickness at the side for various values of
the angle 6. In these measurements, the fast
neutrons were filtered through 45 e¢cm of water at
the reactor, that is, T, = 45 cm. Plots of A3(45,x)
vs x for various angles O are presented in Fig.
14.5, where A3(45,x) is the A measured at water
thickness x into the side of the crew-compartment
shield for fast neutrons filtered through 45 cm of
water (T = 45 cm). For the neutron shield calcu-
lations, it is necessary to know the variation of
A, with reactor shield thickness, 7 . A limited
amount of data obtained for T = 15 cm, when
compared with the data for T = 45 em, indicates
that A7 varies with T, in the same ratio as does
the direct-beam relaxation length, A, This re-
fationship is reasonable in view of the small

2-01-056-7-A-168

 

70 |

g=0 MO 30 4%

 

6.0

g
)\s {cm)
i ]
a0
W
B
»
%
©
O
%
£

——
/
/

 

50

 

 

 

 

 

 

 

4.0 — -
0 5 10 15 20 25

7, WATER THICKNESS AT SIDE OF CREW COMPARTMENT (cm)

Fig. 14.5. Relaxation Length for Fast-Neutron

Shneld from a Comcul Shell Beam as _a Function

of Water Thlckness at the Crew-Compartment Side
for Various Values of the Angle & (T
£ = 64 ).

= 45 cm,

213

 
 

 

 

 

 

 

ANP PROJECT PROGRESS REPORT

neutron energy degradation that results from a
neutron collision with air nuclei. The results of
some dose measuremen’rs at the center of a crew-

compartment mockup at = 90 deg further indicate

the general validity of this relation. For a total
neutron attenuation through the crew-compartment
side shield of the order of ¢, the neutron attenua-

tion calculated for Tn =15 cm, on the basis of the

foregoing relationship, and the A2 values for T =45
cm checked with the measured attenuation to within
30%. Hence, the foregomg relationship, written in
equcmon form,

Lo A A.S(45 x)
7(28) Y (Tn,x) = ,—_,"d(_45—_3 AAT, + x)

 

was used in 'rhe ccrlculaflons. In Eq. 28 )\S(Tn,x)
is the AS at water depth x into the crew-compartment

R Slde shleld for those fast neutrons emanating from

the rec\cfor shleld comccl shell of thickness T _;
AT, + x) is the Ad at total water thickness equal

L to T+ x,
The relaxation Iengths, Al for the air-scatfered

dose at the crew-compar’rmenf rear shield were
 medsured for T
~of 6 from 90 to ]80 deg. The A values were es-
senha”y constant over the mecsured range of 0
and T At the mid-point of the measured range
of Tr which was approximately 15 cm, the A7
value was 7.0 ecm. It was assumed in the calcu-
lations “that this Al value also applied in the
range 6 = 0 to 90 deg In the same manner as
described for A$, the effect of T on Al was ac-
counted for by the relationship

d(T + 15)

S v T

For plastic insrecd of water at the crew shield,
both A% and A7 would be corrected in the same
manner as described for )\é.

An important limitation exists, however, in the use

of the air-scattering probabilities and relaxation

lengths p'r'esenfed here. In the TSF experiments,
' scofiered -dose measurements were not taken for
- TS or T, smaller than about 5 cm of water. Hence,
“fo obtain the dose for T orT, less than 5 cm,
the data were extrapolated to TS =0orT, =0.

The values of {ZPS (6) and *fizPs (8) as well

'ff‘,;_,f’_.?as the re]axahon lengfhs )\S qnd )\’ for T, or

= 45 em over a range of values

T, less than 5 cm are, hence, extrapolated values.

For T_ and T, greater than 5 cm, the use of the
plots, of course, leads to correct results (within
~the limitations previously discussed). However,

for T, ‘and T, less than 5 cm, the plots are only

as rellable as the extrapolation. Indications are

that for T_or T less than 5 c¢m, the relaxation
fengths AS and A7 change appreciably; hence it
does not appear advisable to use the plots for

crew shield thicknesses less than 5 ecm of water.

The concept of the focusmg factor f; and the
method of interpretation of the TSF experiments

to obtain an experimental value of f were dis-

cussed previously by J. E. Van Hoom:ssen.3 Van
Hoomissen obtained an average value of < of

5.3 over the range of angle 0 f-rom 0 to 180 deg,
" the variation of fs with 6 was small and well

within the stchshcal accuracy of the data. Com-
parison with the analytical results of Faulkner?

indicates that the angular distribution of the dose
at the inside surface of the crew-compartment side
shield is approximately cos®. In the TSF experi-
ments from which the value of /° of 5.3 was ob-
tained, the side shield thickness for the crew-
compartment mockup was about 38 cm of water.
Inasmuch as the angular distribution of the dose
at the inside surface of the crew-compartment
sides would be expected to vary with side shield
thickness, the focusing factor /7 is also expected
to vary “With side shield thickness; however,
Faulkner's results show that /% is not very sen-
sitive to angular distribution (/3 is 4.0 for cos
distribution and 6.5 for cos® distribution), In
the calculations, % was taken to be constant at
5.3. This value of /7 should be well within 25%
of the correct value, at least for side shield thick-
nesses greater than about two relaxation lengths
wherein the extremely slanted neutrons are atten-
uated.

To obtain the geometrical attenuation focfors "
. _
fI and f], the angular distribution of dose at the

inside surfoce of the crew-compqrtment rear shield

was taken to be cos® for the scattered dose and
cos® for the direct-beam dose. Integration over

the inside surface of the crew-compartment cirs
cular rear shield for ’rhe dose at the center of \‘he

 

3.]. E. Van Hoomissen, ANP Quar, Prog. Rep. June
10, 1955, ORNL-1896, p 217.

4). E. Faulkner, Focusing of Radiation in a Cylin-
drical Crew Compartment, ORNL CF-54-8-100 {Aug. 18,

1954).

 

 

 

 
ot AR e

#

-y,

 

"

 

¥

 

: cn‘ ‘a distance of 64 ft from the reactor as a func.
" s.&of the water shleid cn‘ theflj o

cylindrical crew compartment (length of cylinder
assumed to be equal to diameter) gives fI = 0.35

and 7 = 0.50.

_ The angular distribution of the fosf-neutron dose
emanating from the reactor primary shield surface
was taken, in<the calculations, to be given by
the function cos® ¢(m = 5 in Eq. 11). Prelimi-
nary calculations mdlccfe that the change in
shield thickness distribution would be small if a
more peaked distribution, cos'® i, were used in
the cclculahons.

RESULTS OF A SAMPLE SHIELD
OPTIMIZATION CALCULATION

. - F.L.Keller
A neutron-shield optimization calculation for a
300-Mw, circulating-fuel, reflector-moderated re-
actor was carried out at the TSF by using the
above-described procedure and the following con-
ditions:

Roadius of pressure shell 97 cm

' “Reactor shleldmg materlaf Water
Reucfor-to-crew-compartment 64 ft

. separutlon distance, 1,

Crew-compartment Jiameféf, D_ 6.6 ft
.-Crew-comparfn"ent lengfh L 10.8 ft
"_Crew-compartmenf shleldlng Plcshc

: mcn‘enal o ' _ {p = 0.94 g/cm )

Neufron doée'rdte at center of

0.25 rem/hr

trew compartment

Altitude ' Sea level

    

 

_ A plot of the direct-beam fast-neutron dose rate

 

 

'the relaxohon lengths obfalned from 'rhe TSF curve
“of N, Vs total water 'rhmkness (Flg 14.4) overw'
the range of the LTSF measurements._ Thus, the
TSF data were used to extend the LTSF curve

  

  
   

 

PERIOD ENDING SEPTEMBER 10, 1955

2-01-056-7-A-169

FAST-NEUTRON DOSE RATE (rem/hr)

 

0 10 20 30 40 50 €0 7 8 90
WATER SHIELD THICKNESS {cm)

Fig. 14.6. Direct-Beam Fast-Neutron Dose Rate
64 ft from a 300-Mw, Circulating-Fuel, Reflector-
Moderated Reactor as a Function of the Thickness
of the Water Shield at the Reactor.

: ;:beyond the measured range. This curve was used
to ob’rcm the values of

c:fun"h'

"o .
Dd( 2 -_]; " (1/A)at
ad) e -

for "'rhej optimization.

“"The reactor shield was divided into six conical
~ shells (N = 8) for the calculation. The values
;of Gnrcmd 9 »» for each of these shells are given
"in Table 14. ] and the valves of the T's and L's
resulting from each of three iterations, as well
"as the shield weights from the third iteration,

215

 

¥ T greTTr

 
i A 5 i it ., il i O i B o i e |

 

" 'ANP PROJECT PROGRESS REPORT

 

TABLE 14.1, OPTIMIZATION OF NEUTRON SHIELDING FOR A 300-Mw, .
'CIRCULATING-FUEL, REFL ECTOR-MODERATED REACTOR | -

 

First lteration

Shielding -.L agrange
Thrckness Multiplier

Second lteration Third Tteration

Shielding L agrange Shielding L.dg'ra;-gé' "\“c’e'ight
Thickness Multiplier Thickness Multiplier of Shield

 

(em)  (x109) (em) (x 108 (em) (x10%)  (tons)
Rur;g;;f'réactdf‘ u.n-gkie‘ 0 (deg) |
01015 | 42.5 14.2 40.2 14.9 43.4 168 115
Steas 397 137 35.9 16.0 39.3 170 096
4ste7s | 33.4 9.2 26.4 14.8 28.6 83 1
758005 216 8.2 20.0 13.5 23 189 o089
05165 232 3.4 10.6 9.5 104 178 053
6510180 21.4 2.9 8.3 93 78 ‘18 o: 002
Rear of crew 47.6 6.3 49.2 8.3 43.8 11‘."9’ 1.47
‘ com'parfmentr _ ,
“Side and front of 17 8.6 19 12.1 168 180 482
crew compartment _ ' o
Total weight | 9f65

 

K R G o cu . i R O il i

 

are presented. The dose rates in the crew com-
partment, as obtained from the third iteration,
were:

péresr _ 0,029 rem/hr
DS,I'BO? — 0.002
ps:side _ 0,205

D = 0.236 rem/hr

total

From the table it may be seen that the iterative
procedure results in a reasonably rapid conver-
gence of the L’'s toward a constant value. A
better choice for the average relaxation lengths
used in the first iteration would have led, of

course, to a much more rapid agreement between

the Efs. The average relaxation lengths used
in the first iteration were A = 53 cm, AT = 6.0
cm, and A} = 8.5 cm. The results of the third

“iteration md:co’re that the correct averages are
_opproxtmotely )\S = 4,3 cm, Al = 54 em, and

= 8.7 cm. A fourth iteration would bring the

) :_'Q s into still closer agreement; however, the
o 'wetgh'r would not be affected appreciably. Also,
 the input data are not sufficiently accurate to
':"h—‘\give real meanmg to the small changes in the

 

 

thicknesses which would be required to give
exact agreement between the [’s.

It is desirable to place a limit on the dose rqtev‘
in the vicinity of the reactor shield from the view

point of neutron activation of aircraft components.
A reasonable limit is that the dose rate of 64 ft
from the reactor should not exceed 125 rem/hr,
From the curve of Fig. 14.6 it is seen that this
shield optimization gives a dose rate of approxi-
mately 600 rem/hr at 64 f from the rear of the
reactor. Thus, some correction should be made
to this shield. The limit on dose rate from the

rear of the reactor shield could be included as 7 -

another condition in the ophmlzcmon calculation;

however, it would introduce a second Lagrange

multiplier and seriously complicate the iterafive

procedure. A simpler procedure is, of course,»

merely to add sufficient shielding to the rear
conical shells to reduce the dose rate to the de-

sired level. While this procedure does not make

precisely optimum use of the added we:ghf the

fact that only 11 cm of additional water is re-
quired on the rear to reduce the dose rate to the

desired level indicates that the fe.s-_ults__wiil_"fi'ot
be far from optimum. If the thicknesses of the

two rear conical shells of the calculated shield

 

 

 

 
i

ufi'

2 ¥

Y

 

were changed to 20 em (T5 =T

 

= 20 cm), the
weight of the shield would be increased by 0.57
ton, The final shield weight would thus be 10,22
tons. - B '

It may also be seen from Table 14.1 that the

“optimum shield has a ratio of crew-compartment
‘shield weight to reactor shield weight of approxi-
. mately 1.6, If the caleulation were carried out

for a larger crew compartment, this ratio would
decrease, and a larger fraction of the weight would

be shifted to the reactor shield. Thus, the larger

the crew compartment, the more nearly the limit
on dose rate at the rear of the reactor is satisfied
by the calculated optimum shield. Conversely,
if the crew compartment were smaller, weight

_ '_VIWO'UIVd-Vbe_ shifted from the ‘rea'cfor shield, and the
limit on the dose rate from the rear of the reactor
would be more seriously violated by the calcu-

lated optimum shield.

MEASUREMENTS IN THE GE-ANP
CREW-COMPARTMENT MOCKUP

F. N. Watson
‘Measurements have been made, with an anthra-
cene crystal, of the gamma-ray dose rates inside

PERIOD ENDING SEPTEMBER 10, 1955

the GE-ANP crew-compartment mockup (Fig. 14.7)
by using the TSF reactor inside the GE-ANP RI
shield mockup® as a source.

Dose-rate measurements (Fig. 14.8) along the
y axis (fore to aft) of the crew-compartment mock-
up indicate a relatively large contribution from
the sides of the mockup, with only a small amount
of radiation penetrating the shadow shield at the
rear of the mockup. Data for the curve labeled
““Compartment D Empty”’ were obtained when
compartment D of the reactor shield was filled
with air.

The gomma-ray dose rates in the front half of
the crew compartment mockup were measured with
the relative position of the crystal and preamplifier
reversed from that shown in Fig, 14.7. Shielding
the detector by the preamplifier and the Lucite
spacer resulted in a 10% reduction in dose when
data were taken in the front portion of the crew
compartment,

Gamma-ray dose rates were also measured along
the x axis (left to right) for various values of y

(Figs. 14.9, 14.10, and 14.11). Again there was

 

®J. E. Van Hoomissen, ANP Quar. Prog. Rep. June
10, 1955, ORNL-1896, p 218, Fig. 14.25.

2-01-056-3-T24-81

 

(THIS PLATE LEFT OUT
‘. ,

ALUMINUM
//

 

 

 

 

 

|

 

 

 

 

 

 

 

 

 

 

 

 

~LueITE |
= g ¢

| REAR

 

 

 

 

 

-:|~71/4~n. e ]

 

 

PREAMPLIFIER

. T - - 5
_ =l _removaBLE PLUG FOR 55
] .
. DETECTOR INSERTION CRYSTAL o
— | ; B
' .oo*o’

 

LEAD—

 

 

 

 

 

 

AT T ALuMINGM

7 ALL DIMENSIONS ARE IN INCHES

 

 

 

v-—-*:”/s

 

 

~ Fig. 14.7. Schematic Diagram of GE-ANP Crew-Compartment Mockup.

217

 

 

 
 

 

 

ANP PROJECT PROGRESS REPORT

an indication that the majority of the gamma rays  Investigation of the GE-ANP R1 shield mockup is

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

entered from the sides of the crew-compartment continuing with the measurement of gamma-ray
mockup. The dose rates from center to side did dose rates in air 4 to 30 ft from the shield at
not, however, vary in excess of approximately 30%. various values of the angle 6.
10~° 2.—0‘.!—'656—3-67 69,71-156A
.
COMPARTMENT D EMPTY
|| 1 —FACE OF LEAD SHIELD / ] FRONT OF COMPARTMENTH—""]
i A -
9 o L 3
—_ — d/--"""""""--'!"—--_---1 /
% 5 / /
z ] / ® \
£ / "~~~ COMPARTMENT D FuLL—"|
E / /]
E / _ A
b / ACCESS PLUG —+——]
o / /
wJ
w
o
a /
X / h=195 ft
& 4 | d= 64 ft
<
=
I 2
o » ‘
CENTER OF
SCINTILLATION
CRYSTAL AT:
x=0
«pee s ¥= VARIABLE
z=0
107’
80 o2 04 #6 128 140 152 164 {76 188 200 212 224 236 248 260

¥, DISTANCE FROM REAR FACE OF CREW COMVPARTMENT TO DETECTOR CENTER {c¢m)

Fig. 14.8. Gamma-Ray Dose Rates Along y Axis of Crew Compartment.

218

 

 

 
o Camonxan oo o ieas,
o

TR

“’x .

PERIOD ENDING SEPTEMBER 10, 1955

2-01 -056-3‘-72, 73-152A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wy

& kel

5

 

lfi‘:.‘y'

x, HORIZONTAL DISTANCE FROM ¢ AXIS TO DETECTCR CENTER (cm)

Fig. 14.9. Gamma-Ray Dose Rates Along x Axis
of Crew Compartment (y = 100, 125, ond 155 cm;

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

z = 0). 230 cm; z = 0).
SRR
2-0{—-056-3—-68~155A
1078 i l : i i
| ALL CURVES TAKEN WITH REACTOR SHIELD
COMPARTMENT D EMPTY,
]
y =460 cm
/_,-—'—'_"'_::: ~ O
——T
Pt o y =150 cm i S
P l e
[ e e T e 0 -
£ 5 T y =130 cm e
.
=
L
S =
> ~ v
& : \\.,___ y=100cm /
g S —— »
N
o
D -
g
<
g
& \ !
- F b =195 ft
2 d = 64 ft
; CENTER OF SCINTILLATION CRYSTAL AT:
' " x = VARIABLE
" y'="AS NOTED
2=0
------- . ‘0—7 —— — - - —_— N
.40 S300 =p0 o 0 Ti00 20 30 40

 

x, HORIZONTAL DléTANCE FROM ¢ AXIS TO DETECTOR CENTER (em)

Fig. 14.11. Gamma-Ray Dose Rates Along x Axis
of Crew Compartment (y = 100, 130, 150, and
160 cm; z = 0).

 

- 46 2-01-056-3-70-154A 10'6 ’ l ’ I \ } r ‘ l
107
, , ‘ ALL CURVES TAKEN WITH REACTOR
. ALL CURVES TAKEN WITH SHIELD COMPARTMENT D FILLED
© REACTCR SHIELD COMPARTMENT D WITH WATER
- p—- FILLED WITH WATER
— T = e,
P T L e T
. - L ",’_ _,_.—-‘_--_ -—..A,___bh
3 T | e s é/“" T ri— \‘
S g ',/'\ . 3z 5 71~ T~
5 7 y =155 em ] : 1 N
S ~ £
E — - . b e E
u \ W
E ¥y =125 e¢m =
@ L
W w
-9 @ h =195 ft
- _ 8 d= 64 ft
~ =100 cm Y ‘
% T~ \\>/ % CENTER OF SCINTILLATION CRYSTAL AT
CF —— e x X = VARIABLE
< < o y=198.8cm
= 2
52 ; h =195 ft £ e y=230cm
g d =864 ft 3 A y=1489 cm
_ A y=17
CENTER OF SCINTILLATION s =170 cm
CRYSTAL AT: z=0
x = VARIABLE
¥ = AS NOTED
Z=0
- ’ ‘ g’
1077 ~40 =30 20 -0 0 10 20 30 40
-4Q0 =30 -20 -0 0 10 20 30 40 x, HORIZONTAL DISTANGE FROM o AXIS TO DETECTOR CENTER {cm)

Fig. 14.10. Gamma-Ray Dose Rates Along x Axis
of Crew Compartment (y = 170, 189, 198.8, and

219

 
Jois.

i

e e

 

PrARRR TS

 

 

 

 

 

 
 

i i s i et 3 iy ik A BRI BB i s y: il b, e il e . i e S i i o e e . i dara iy il A i e oo, d ) A e L

~ Aol ik i i —y . —— i e ; - i i "

 

OFFICIAL USE ONLY

AIRCRAFT REACTOR ENGINEERlNG DIVISION
THE OAK RIDGE NATIONAL LABORATORY

SEPTEMBER 1, 1955

 

 

§. J. CROMER, DIRECTOR
H. MCFATRIDGE, SEC.

RD
ARE

 

 

 

 

ASSISTANT TO DIRECTOR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

R. § CARLSMITH ARE
R. E. THOMPSON ARE
REACTOR PHYSICS REACTOR CONSTRUCTION POWER PLANT ENGINEERING EXPERIMENTAL ENGINEERING ENGINEERING DESIGN 7503 AREA CONSTRUCTION
A, M. PERRY ENR E.’S. BETTIS ARE A. P. FRAAS ARE H. W. SAVAGE ARE H. C. GRAY PWA . G. PIPE ARE
M. WILSON, SEC. ARE M. OVERTON ARE P. HARMAN, SEC. ARE D. ALEXANDER, SEC. ARE J. ZASLER ARE S, M. JANSCH SEC. ARE
L. E. FERGUSON, SEC. ARE
H. W. BERTMNI ARE FLUIE MECHANECS AND ART ADMINISTRATION R. CORDOVA ARE
A, FORBES ARE REACTOR DESIGN V. 5. KELLEGHAN ARE
W. E. KINNEY ARE G.D. WHITHAN ARE W. L. SCOTT ARE J. A, OLSON PWA F, R, MCQUILKIN ARE
M. TSAGARIS ARE A. M. SMITH ARE D. C. BORDEN PWA J. M. MILLS, JR. ARE
W. J. STELZMAN ARE HYDRODYNAMICS AND THERMODYNAMICS R.C. DANIELS ARE G. C. ROBINSON ARE
FILL AND DRASN SYSTEM ¥, T. FURGERSON ARE J. Y, ESTABROOK ARE R. D. STULTING ARE
T ARy e W. A, FRY PWA C. F. WEST ARE
L. A. MANN ARE - L W. E. LEUTLOFF PWA .
- A M. E. LACKEY ARE €. MANTELL A
G. SAMUELS ARE DEVELOPMENT TEST SERVICES W o P
HIGH-TEMPERATURE CRITICAL M. W. HINES, COMPUTER ARE . - 1. NELSON WA
EXPERIMENT J, J. TUDOR, DRAFTSMAM  ARE E.R.DY PHA . B. MCDONALD ARE '
W, C. TUNNELL ARE ; ' L. RUSSELL SEC. ARE D. P. HARRIS, SEC. ARE PUMP DESIGN %
6. W, PEACH ARE APPLIED MECHANICS AND STRESS PUMPS A- N. MONTCOMERY ARE W. G, COBB ave . | ;
ANALYSIS A, €. GRINDELL ARE TEST FACILITIES ASSEMBLY 1. T. DUDLEY ARE- |- '
ENGINEERING TEST UNIT {ETU) R. B. MEGHREBLIAN ARE " R. CURRY PWA COORDINATION \; ; l;-;RLI;ISS ,i\gé. g !
L.E. ANDERSON PWA S, M. DECAMP ARE E. STORTO ARE
M. BENDER ARE :
F.AF USAT W. L. SNAPP PWA J. L. CROWLEY ARE e ey e
INSTRUMENTATION AND CONTROLS B.L. GREENSTREET ARE 3. J. SIMON PWA W. H. KELLEY ARE
o werLe s B WARD ARE W, J. MUENZER ARE L soN e :
E. R. MANN Ic W. 1. CALYON, DRAFTSMAN  ARE H. C. YOUNG PWA H. C. SANDERSON PWA V. G GEORGE ARE
R.G. AFFEL IC ’
CORROSION LOOPS ‘
M. C. BECKER P HEAT TRANSEER HEAT EXCHANGERS AND 5
G. H, BERGER IC SPECIAL EQUIPMENT C. P, COUGHLEN ARE ELECTRICAL DESIGN
J. M, EASTMAN B R. D. SCHULTHEISS ARE R. E. MACPHERSCN ARE R. A. DREISBACH PWA T. L. HUD! ARE
C. F, HOLLOWAY Ic é TOSRAs QSE J. C. AMOS ARE P. G. SMITH ARE A.H. ANDERSON ARE
R, F. HYLAND iIc M. H. COOPER PWaA J. KERR ARE
1. W, KREWSON Ic H. C. HOPKINS PWA L. H. DEVLIN PWA ELECTRIC SERVICES B.C. GARRETT PWA
W. R, MILLER* Ic H. WIARD cv L. R. ENSTICE PWA E. M. LEES ARE J. 0. NICHOLSON * ARE
S. C. SHUFQRD PWA M. M. YAROSH ARE F. FERRIGNO PWA D. L. CLARK ARE
C. B. THOMPSON MH 1. W. KINGSLEY ARE W. F. FERGUSON- ARE GENERAL DESIGN
E. YINCENS IC SPECIAL PROBLEMS R. D. PEAK PWA C. E. MURPHY ARE
C. 5. WALKER Ic A. L. SOUTHERN ARE A. A ABBATIELLO ARE
G. €. GUERRANT Ic ¥ 2 COTTRELL fne J. G. TURNER PWA TEST OPERATIONS J. M. CORNWELL PWA
C' s' BURTNETTE USAF J. 4 MILICH PWA H. J. KICKINSON PWA
-3 E. F. JURGIELEWICZ PWA
C. J. PRICE USAF PROCUREMENT COORDINATION TECHNICIAN GROUP 3T, MEADOR PWA
W. F. BOUDREAU ARE R. HELTON ARE W. A, SYLVESTER PWA
CONSULTANTS J. PARKER, 3EC. ARE J. S. ADDISON ARE N. E. WHITNEY PWA
T. ARNWINE ARE G. R. HICKS ARE
A. H. FOX, UNION COLLEGE GENERAL EQUIPMENT 6. 5. CHILTON ARE C. A, MILLS ARE
W. LOWEN, UNION COLLEGE W. R. OSBORN ARE J. M. COBURN ARE C. F. SALES ARE
R. L. MAXWELL, UNIVERSITY OF F. M. LEWIS ARE J. R. CROLEY ARE
TENNESSEE s g-g\-MS\LSREY QEE W, CLFNN!NGHAM ARE CHECKING
G. F. WISLICENUS, PENNSYLVANIA STATE - ¥ R. E. DIAL ARE :
COLLEGE D. L. GRAY ARE LR DUCKWORTH ARE J.R. LARRABEE PWA
$_ flAsxsgg igg W. H. DUCKWORTH ARE G. G, MICHELSON ARE
H. W. K. R, FINNELL ARE
$.R. ASHTON ARE H. FOUST ARE RECGRDS AND PRINTING
C. A, CULPEPPER ARE ¥. D. GHORMLEY ARE s, J, FOSTER ARE .
G. C. JENKINS ARE C. J. GREEN ARE 3. 1 PLATZ ARE :
E. 8. PERRIN ARE T. L. GREGORY ARE
F. M. GRIZZELL ARE
WELDED EQUIPMENT R A HAMRICK ARE
R. B. CLARKE ARE P. P, HAYDON ARE
B. E. BLACK ARE C. G. HENLEY ARE
E. B. EDWARDS ARE B. L. JOHNSON ARE
H. ¥. HOOVER ARE J.R. LOVE ARE
E. A. JAGGERS ARE D.E. MCCARTY ARE
T. A. KING ARE G. E. MILLS ARE
J. W. TEAGUE ARE B. H. MONTGOMERY ARE
C. C. NANCE ARE
SPECIAL EQUIPMENT 3 T PARSONS ARE
W. D. GOOCH ARE H. E. PENLAND ARE
L. W.LOVE PWA M. A. REDDEN ARE
T. K. WALTERS ARE R. REID ARE
D, R, WESTFALL ARE F. J. SCHAFER ARE
J. R. SHUGART ARE
IN-PILE LOOP TESTS €. E. STEVENSON ARE
D. B. TRAUGER ARE A. G. TOWNS ARE
C. A, WALLACE ARE
5. EUBANKS, SEC. ARE B C WILLIAMS ARE
C. C.BOLTA PWA -
L.'P. CARPENTER ARE
J. A, CONLIN ARE
C. W, CUNNINGHAM ARE CONSULTANTS
ot AT o J. F. BAILEY, UNIVERSITY OF TENNESSEE
i R. L. MAXWELL, UNIVERSITY OF
TENNESSEE
OFFICE SERVICES W. K. STAIR, UNIVERSITY OF TENNESSEE
J. P, LANE ARE

 

 

 

 

 

 

 

 

 

 

T AT TR

 

221

 

 

 

 

 

 

 

  
e i

 

 

 

 

 

OFFICIAL USE ONLY

i

THE AIRCRAFT NUCLEAR PROP;ULSION PROJECT
AT .,
THE OAK RIDGE NATIONAL Lfl)‘BORATORY

SEPTEMBER 1, 1955

i

 

 

SUPPORTING RESEARCH

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W, H, JORDAN
A, 3, MILLER i
%
STAFF ASSISTANT STAFF ASSISTANT STAFF ASSISTANT STAFF ASSISTANT STAFF ASSISTANT STAFF ASSISTANT
W. K. ERGEN ARE E. P. BLIZARD AP W. R. GRIMES MC W. D. MANLY M D. 5. BILLINGTON s H. F. POPPENDIEK REE
REACTOR PHYSICS SHIELDING CHEMISTRY ANALYTICAL CHEMISTRY |METALLURGY METALLOGRAPHY RADIATION DAMAGE HEAT TRANSEER AND PHYSICAL
1 PROPERTIES RESEARCH
A. SIMON* AP W, R. GRIMES MC M. T. KELLEY* AC W. D. MANLY | M R. ). GRAY* M D. 5. BILLINGTON* 5
SHIELDING RESEARCH D. R. CUNEO NC C. D. SUSANO* AC W. H. BRIDGES* M M. J. FELDMAN I3 H. F. POPPENDIEK REE
D. E. CALDWELL, SEC. MC 8. E. YOUNG,* SEC. AC J. THOMAS, SEC. M G. W. KEILHOLTZ s T. K. CARLSMITH, SEC. REE
C. L. BRADSHAW" MP E. P. BLIZARD* AP J. C. WHITE AC R. 5. CROUSE M J. B. TRICE 5
R. R. COVEYOU* AP R. RICKMAN, SEC. AP E. M. ZARZECKL,* SEC. AC T. M. KEGLEY M J. €. WILSON ss
L. DRESNER* AP $. AUSLENDER PWA SPECIAL PROBLEMS GEMERAL CORROSION D. F. STONEBURNER M HEAT TRANSFER
M. D. GIVEN* MP M. E. LAVERNE AP ] E.R. BOYD M
1. H. MARABLE* AP £. H. MURRAY AP R. F. NEWTON RD E. E. HOFEMAN M B. F. DAY M MTR LIAISON C. M. COPENHAVER REE
RESEARCH
M. L. NELSON* AP R, B. MURRAY AP R. CARLANDER PWA B J REECE " N. D. GREENE REE
PHASE EQUILIBRIUM STUDIES A.S. MEYER, JR. AC W. H. COOK " H. V. KLAUS 5 H. W. HOFFMAN REE
SHIELD DESIGN G. GOLDBERG AC D. H. JANSEN M F. €, LYNCH REE
CONSULTANT ¢ i ':AA%L?\?jrCHER fi B. L. MCDOWELL AC J. E. POPE M HOT LAB FACILITY G. L. MULLER PWA
J. B, DEE PWA e W. J. ROSS AC L. R, TROTTER M L. D. PALMER REE
J. A. NOHEL, GEORGIA TECH. B. H. CLAMPITT MC -d: : M. J. FELDMAN 55
C. A, GOETZ PWA H. DAVIS PWA C. R. WILLIAMS AC : ELLIS J. L. WANTLAND REE
K. PENNY AP - DYNAMIC CORROSION C. 55 R. M. BURNETT REE
J, SMOLEN PWA R. E. MINTURN USAF CERAMIC RESEARCH E. J. MANTHOS 58
. DEVELOPMENT J. LONES REE
D. K. TRUBEY AP R. E. MOORE e J. H. DeVvAl M . . B. PARSLEY 55 R. L. MILLER REE
M€, WOODSUM PWA R. J. SHEIL MC 1. P. YOUNG AC E. A, KOVACEVICH M J. M. WARDE® M R. N. RAMSEY 55 G . WINN REE
D. ZUCKER M M. A, MARLER AC A. TOBCADA M A, HOBBS,* SEC. M E. 5. SCHWARTZ s5 ’
CRITICAL EXPERIMENTS LID TANK SHIELDING FA G. D, BRADY M
1L CILITY PHYSICAL CHEMISTRY SERVICE k. W, EVANS " RADIATION METALLURGY
A. D. CALLIHAN® AP R. W. PEELLE AP ‘ C. E. CURTIS* " . PHYSICAL PROPERTIES
M. L. RUEFF,* SEC., AP G. T, CHAPMAN AP F. F. BLANKENSHIP MC W. F. VAUGHN AC FABRICATION J. R, JOHNSON* M J. C. WILSON 5 5.1, COHEN REE
W, J. McCOOL, PWA M. BLANDER MC R.F. APPLE ) AC [ A J. TAYLOR* M C. D. BAUMANN S5 v b oeRs REE
3. M. MILLER AP S. CANTOR MC D. E. CARPENTER AC J. K, COOR M 1. A, GRIFFIN® M W. E. BRUNDAGE s5 . G‘ C. BLALOCK REE
J. 5. CRUDELE PWA E. BECKHAM AP R. E. CLEARY PWA R. C. BRYANT AC M. R. D"AMORE PWA W, W, DAVY(S sS S.J ‘CLAIBORNE REE
T ELLIS AP 3 FRANCIS AP H. A, FRIEDMAN MC L. R, HALL AC H. INOUYE M N. E. HINKLE s 4 Sones REE
W. J. FADER PWA H. JARYS I S, LANGER MC L. E. IDOM AC J. P. PAGE M CONSULTANTS A. 5. OLSON S5 -
V. 6. HARNESS* AP J.R. TAYLOR AP R. P. METCALF MC A. H. MATTHEWS AC T. K. ROCHE M R. A. WEEKS* 8§
J. 1 LYNN AP . W, WAMPLER AP R.E. THOMA MC C. E. PRATHER AC R. W. JOHNSON M G. M, BUTLER, CARBORUNDUM COMPANY 1, C. ZUKAS 55
E. R. ROKRER* AP A D WILSON ac !D M. T. CORY, GRAPHITE SPECIALTIES T. PRICE PWA
E. V. SANDIN PWA FUEL PREPARATION RESEARCH C. M. WILSON AC WELDING AND BRAZING CORPORATION
D. SCOTT, JR. ARE BULK SHIELDING FACILITY ! H. INSLEY RADIATION CHEMISTRY
5. SNYDER PWA F. €. MAIENSCHEIN AP G. M. WATSON st P. PATRIACA M T. N, MGVAY, UNIVERSITY OF ALABAMA FUEL REPROCESSING
R, M. SPENCER USAF C. BOUNDS, SEC. AP C. M. BLCOD MC R. E. CUAUSING ] 7. 5. SHEYLIN, OHIO STATE UMIVERSITY G. W. KEILHOLTZ 58 .
D. V. P, WILLIAMS* AP T.V. BLOSSER AP F. L. DALEY MC R.L.H ‘ESTAND Pwa H. THURNAUER, MINNESOTA MINING AND W. E. BROWNING 58 F. R, BRUCE cT
¢ D-ZERBY* AP G. M. ESTABROOK AP F. W. MILES . MC G. M. SLAUGHTER M MANUFACTURING COMPANY D. E. GUSS USAF
e LD FLYNN AP G. E.{CONNOR M H. L. HEMPHILL $5
A T. FUTTERER PWA CORROSION STUDIES SPECTROGRAPHIC ANALYSIS B. McDOWELL M M. F. OSBORNE 35 CHEMICAL DEYELOPMENT
M. P. HAYDON AP L . C. E. SHUBERT M H. E. ROBERTSON 55
- F. KERTESZ MC 4. R MCNALLY st D. E. FERGUSON* cT
K. M. HENRY AP B ORRIS* AND OTHERS by R. G, SHOOSTER M M. T. ROBINSON 55
£ B, JOHNSON AP H. J. BUTTRAM MC - A. NORR L. C. WILLIAMS M C. C. WEBSTER 55 G. I. CATHERS cT
T. A LOVE AP F. A, KNOX MC METALLURGY CONSULTANTS W. R. WILLIS 55 M. R. BENNETT CT
W 20BEL 5 R. E. MEADOWS MC PHYSICAL CHEMISTRY OF LIQUID METALS R. M. DUFF cr
" D. J. KIRBY AP G. F. SCHENCK PWA ! N. CABRERA, UNIVERSITY OF VIRGINIA FLUX MEASUREMENTS
J. SELLERS iC J. M. DIDLAKE MC G. P. SHITH M N. J. GRANT, MASSACHUSETTS UNIT OPERATIONS
R. M. SIMMONS AP €. R. BCGSTON " INSTITUTE OF TECHNOLOGY A B TRICE 5 W, K. EISTER* cT
D, SHIDDIE \c PRODUCT{ON OF PURIFIED FUELS MASS SPECTROMETRY J. V. CATHCART M J. L. GREGG, CORNELL UNIVERSITY J. KRAUSE PWA )T, LONG cr
F L] . H, W. LEAVENWORTH PWA W. D, JORDAN, UNIVERSITY OF ALABAMA M STAINKER &1
. G. $TOUT ic G. J. NESSLE MC C. R. BALDOCK S| G. F. PETERSEN M E.F. NIPPES, RENSSELAER ENGINEERING PROPERTIES 5. H.S
H. WEAVER AP 3. P, BLAKELY MC ) R.SITES si . G, JONES, JR. cr
M. E. STEIDLITZ M POLYTECHNIC INSTITUTE 0. SISMAN® WA
€. R, CROFT MC R. W.ANDERSON M W. F. SAVAGE, RENSSELAER - 58 4. 5 WATSON T
TOWER SHIELDING FACILITY F. A. DOSS MC L : R. M. CARROLL 55
J. E. EORGAN L L'FA - M POLYTECHNIC INSTITUTE J. G. MORGAN 55 DESIGR
C. E. CLIFFORD AP N, V. SMITH i l G. SISTARE, HANDY & HARMON M. T. MORGAN 5
E. McBEE, SEC. AP ) TROITT s MECHANICAL PROPERTIES P. C. SHARRAH, UNIVERSITY OF H. E. GOELLER* CcT
J. L. HULL AF ¥. 7. WARD MC D. A. DOUGLAS M ARKANSAS R. P. MILFORD cT
F. L. KELLER AP . R. K. BAGWELL MC C. R, KENNEDY PWA F. G. TATNALL, BALDWIN-LIMA-HAMILTON F. N. BROWDER cT
F. W. SANDERS AP J‘ P ’ EUBANKS MC J : RI'WE R IR M E. C. WRIGHT, UNIVERSITY OF ALABAMA
N BAC B, F. HITCK MC J. W. WODDS M PILOT PLANT
D RSROLL « W, JENNINGS MC K. W. BOLING W H. K. JACKSON™ eT
-D. G. A. PALMER MC 4. T. GAST M W. H. CARR* T
1. D. CONNER AP B. C. THOMAS MC 4. B, HUDSON M *
J.N. HONEY AP ‘ W. H. LEWIS cT
- M R. G. WILEY MC V. G. LANE M
G. G. UNDERWOOD ic B. MchABB, JR. M METALLURGY CONTRACTORS
R. E. WRIGHT AP CHEMICAL EQUILIBRIA g- g TL%‘:‘;%S' IR. ’; BATTELLE MEMORIAL INSTITUTE
. K. BRUSH BERYLLIUM COMPANY
L. G. OYERHOLSER NC C. W, YIALKER M FERROTHERM
CONSULTANT E. E. KETCHEN MC
M B. PANISH He | GLENN L. MARTIN COMPANY
H. A.-BETHE, CORNELL UNIVERSITY Vb REDMAN e NONDESTRUCTIVE TESTING METAL HYDRIDES, INC.
D NEW ENGLAND MATERIALS TESTING
i B. J. STURM M R. B. OLIVER M
LABORATORY
. L. E. TOPOL. MC J. W, ALLE M
CONTRACTORS ‘ R W. MeCLUNG “ RENSSELAER POLYTECHNIC INSTITUTE
METAL HYDRIDES, INC. SUPPORTING STUDIES K. W. REBER M fj:f’\fg;gmuoaf 'F&anéshslze
NUCLEAR DEVELOPMENT ASSOCIATES, : W, J. HASON W
P. A. AGRON c
| J. E. SUTHERLAND <
CONSULTANTS
H. INSLEY
T. N. MCVAY, UNIVERSITY OF ALABAMA
! CONTRACTORS
5 AMES LABORATORY
, BATTELLE MEMORIAL INSTITUTE
! CARTER LABORATORIES
f MOUND LABORATORY
| A. R. NICHOLS, SAN DIEGO STATE
: COLLEGE
UNIVERSITY OF ARKANSAS
*PART TIME

 

 

 

 

 

222

 

 

 

 

 

 
i

e,

<

-

g

i .

i oncill ik ol e oy i i 055, Gk sein. A s i R

 

 

 

OFFICIAL USE ONLY

THE AIRCRAFT NUCLEAR PROPULSION PROJECT

AT
THE OAK RIDGE NATIONAL LABORATORY

SEPTEMBER 1, 1355

 

 

 

 

 

 

 

ANP PROJECT DIRECTOR W. H. JORDAN RD
CO-DIRECTOR S. J. CROMER RD
ASSISTANT DIRECTOR A. J. MILLER RD
D. HILYER, SEC. RD
PRATT & WHITNEY AIRCRAFT REPORTS
E. R. DYTKO, ASST. PROJ. ENG. PWA . A. W, SAVOLAINEN ARE
A. GIANGREGORIO, ADM. ASST. PWA
0. A, LIVINGSTON, SEC. ARE LITERATURE SEARCHES
H. C. GRAY, DESIGN ENGINEER PWA : A. L. DAVIS ARE

 

 

 

 

 

 

 

 

 

 

S. J. CROMER, DIRECTOR
H. McFATRIDGE, SEC.

 

AIRCRAFT REACTOR ENGINEERING DIVISION

RD
ARE

 

 

 

ANP STEERING COMMITTEE

W. H. JORDAN, CHAIRMAN

E. 5. BETTIS

D. 5. BILLINGTON
E. P. BLIZARD

G. E. BOYD

S. J. CROMER

 

 

 

ASSISTANT TO DIRECTOR
R. S, CARLSMITH

. ERGEN
. FRAAS
. GRIMES

ARE . LARSON

 

 

 

 

 

EXPERIMENTAL ENGINEERING

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

H. W. SAVAGE ARE
POWER PLANT ENGINEERING

A. P, FRAAS ARE

ENGINEERING DESIGN
H. C. GRAY PWA
REACTOR CONSTRUCTION
E.S.BETTIS ARE
REACTOR PHYSICS

A. M. PERRY ENR
7503 AREA CONSTRUCTION

W. G. PIPER ARE

 

 

. MANLY

. MILLER

. POPPENDIEK
. SAVAGE

. SHIPLEY

. SWARTOUT

. WEINBERG

PrmITI>EAErE
TPUET-OMBTOX

 

 

 

 

NOTE: THIS CHART SHOWS ONLY THE LINES OF TECHNICAL COORDINATICN OF THE ANP PROJECT, THE
VARIOUS INDIVIDUALS AND GROUPS OF PEOPLE LISTED IN THIS AND THE FOLLOWING CHARTS ARE EN-
GAGED EITHER WHOLLY OR PART TIME ON RESEARCH AND DESIGN WHICH IS COORDINATED FOR THE
BENEFIT OF THE ANP PROJECT IN THE MANNER INDICATED ON THE CHART. EACH GROUP, HOWEVER, -
1S ALSO RESPONSIBLE TO ITS DIVISION DIRECTOR FOR THE DETAILED PROGRESS OF THE RESEARCH -
AND FOR ADMINISTRATIVE MATTERS. '

L
THE KEY TO THE ABBREVIATIONS USED IS GIVEN BELOW.

AC  ANALYTICAL CHEMISTRY DIVISION —ORNL

AP APPLIED NUCLEAR PHYSICS DIVISION — ORNL

ARE  AIRCRAFT REACTOR ENGINEERING DIVISION — ORNL

BAC BOEING AIRPLANE COMPANY

BP  BENDIX PRODUCTS, DIVISION OF BENDIX AVIATION CORPORATION
C CHEMISTRY DIVISION ~ ORNL

CT  CHEMICAL TECHNOLOGY DIVISION - ORNL

CY  CONSOLIDATED VUL TEE AIRCRAFT CORPORATION

ENR ELECTRONUCLEAR RESEARCH DIVISION — ORNL

tC INSTRUMENTATION AND CONTROLS DIVISION ~ ORNL

M METALLURGY DIVISION ~ ORNL

MC  MATERIALS CHEMISTRY DIVISION ~ ORNL

MH  MINNEAPOLIS-HONEYWELL REGULATOR COMPANY

MP  MATHEMATICS PANEL — ORNL

PWA  PRATT & WHITNEY AIRCRAFT, DIVISION OF UAC

RD  RESEARCH DIRECTOR’S DEPARTMENT — ORNL

REE REACTOR EXPERIMENTAL ENGINEERING DIVISION ~ ORNL
St STABLE 1SOTOPE RESEARCH AND PRODUCTION DIVISION — ORNL
$s SOLID STATE DIVISION ~ ORNL

USAF UNITED STATES AIR FORCE

*PART TIME

 

 

 

 

-

 

 

- SUPPORTING RESEARCH

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W. H. JORDAN
A. J. MILLER
METALLURGY _
W. D. MANLY, STAFF ASSISTANT M
CHEMISTRY -
W. R. GRIMES, STAFF ASSISTANT MC
. SHIELDING :
E. P. BLIZARD, STAFF ASSISTANT ' AP

 

 

 

 

 

 

RADIATION DAMAGE
D. 5. BILLINGTON, STAFF ASSISTANT . |

 

 

 

 

   

W

 

HEAT TRANSFER AND PHYSICAL PROPERTIES
H. F. POPPENDIEK, STAFF ASSISTANT '~ REE

 

 

 

 

 

REACTOR PHYSICS
W. K. ERGEN, STAFF ASSISTANT ARE

 

 

 

 

 

 

FUEL REPROCESSING
F. R. BRUCE* c1

 

 

 

 

 

 

T Pt

TR T Y TN TR

 

 

223

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

- LA
L . ;
. " L L
[ 1
: C s
v .
Nl .
;
;
§
;

 

 

 

 

 

 

 

 

 

 

 

Vi

 

 

 

 

 

 

 

 

 

 

1

 

 

 

 

 

 

 

 

 

 

. Coe
{ : L
L . v
i . ;
)
‘.
. Lo .
' .
. i .
: o Ll
. .
e . . e

 

 

    

e